EPA/630/R-01/005
April 2002
Summary Report of the Peer Consultation Workshop
on the Draft Framework for Cumulative Risk Assessment
U.S. Environmental Protection Agency
Washington, DC
August 22-23, 2001
Risk Assessment Forum
U.S. Environmental Protection Agency
Washington, DC 20460
-------
NOTICE
This document has been reviewed in accordance with U.S. Environmental Protection Agency
(EPA) policy and approved for publication. Mention of trade names or commercial products
does not constitute endorsement or recommendation for use.
This report was prepared by Versar, Inc, an EPA contractor (Contract No. 68-C-99-238, Task
Order No. 50), as a summary of the discussion held at the Peer Consultation Workshop on the
Draft Framework for Cumulative Risk Assessment (August 22-23, 2001). This report captures
the main points and highlights of the meeting. It is not a complete record of all details discussed,
nor does it embellish, interpret, or enlarge upon matters that were incomplete or unclear.
Statements represent the individual views of each workshop participant.
-------
CONTENTS
Page No.
EXECUTIVE SUMMARY i
1.0 INTRODUCTION 1-1
1.1 Workshop Purpose 1-1
1.2 Workshop Participants 1-1
1.3 Discussion Topics 1-1
1.4 Agenda 1-2
1.5 Workshop Summary Report 1-3
2.0 SUMMARY OF OPENING REMARKS 2-1
2.1 Welcome 2-1
2.2 Background on Risk Assessment Forum and Framework Development 2-1
2.3 Presentation of the Draft Framework for Cumulative Risk Assessment 2-2
2.4 Discussion on Opening Presentations 2-4
2.5 Introduction of Chair and Overview of Discussion Topics 2-6
3.0 DISCUSSION SESSION I - OVER-ARCHING ISSUES 3-1
3.1 Summary of Comments and Suggestions 3-1
3.2 Detailed Discussion 3-3
4.0 DISCUSSION SESSION H - PROCESS ISSUES 4-1
4.1 Summary of Comments and Suggestions 4-1
4.1.1 Stakeholder Involvement 4-1
4.1.2 Use of Cumulative Risk Assessment in Decision Making 4-3
4.2 Detailed Discussion 4-4
5.0 DISCUSSION SESSION IE - TECHNICAL ISSUES 5-1
5.1 Vulnerability 5-2
5.2 Combining Chemical and Non-Chemical Stressors 5-3
5.3 Combining Different Types of Risk 5-4
5.4 Uncertainty/Variability Issue 5-5
5.5 Approaches to Cumulative Risk Assessment 5-6
5.6 Biomarkers 5-7
5.7 Distinguishing Between Cumulative Risks and Cumulative Impacts 5-8
5.8 Epidemiology Studies 5-9
5.9 Issue of Completeness (Quantitative/Qualitative) 5-10
6.0 OBSERVER COMMENTS AND DISCUSSION 6-1
-------
LIST OF APPENDICES
APPENDIX A LIST OF PARTICIPANTS
APPENDIX B LIST OF OB SERVERS
APPENDIX C DISCUSSION TOPICS
APPENDIX D AGENDA
APPENDIX E PRESENTER OVERHEADS
APPENDIX F FRAMEWORK CONCEPTUAL DIAGRAMS (FROM DOUGLAS
CRAWFORD BROWN)
APPENDIX G POST-MEETING WRITTEN COMMENTS
APPENDIX H DRAFT FRAMEWORK FOR CUMULA TIVE RISK ASSESSMENT
-------
EXECUTIVE SUMMARY
The Peer Consultation Workshop on the Draft Framework for Cumulative Risk Assessment was
held on August 22 and 23, 2001, in Arlington, VA. The purpose of the meeting was to provide
interested stakeholders with an opportunity to see an early draft of the Framework and to obtain
early input from these stakeholders that will assist EPA with further development of the
Framework document.
After opening presentations by EPA staff providing background information on the draft
Framework, the participants engaged in discussions organized around three topic categories: (1)
over-arching issues, (2) process issues, and (3) technical issues. Most of the remainder of the
first day was devoted to such over-arching issues as the scope and organization of the draft
Framework report. Discussion on process issues began late in the afternoon of the first day and
included considerable discussion of stakeholder involvement in the cumulative risk assessment
process. The technical discussion, on about ten topics, was completed on the second day. Time
was set aside on both days of the workshop for comments to be made by observers.
Summary of Comments on Over-arching Issues
Many experts stated that the draft Framework represents a good start and it is expected to
be a useful document (after a lot of revisions) that will help to promote consistency in
cumulative risk assessment.
• Several individuals commended EPA for writing a concise document that maintains a
broad perspective, which will allow users flexibility in planning and implementing
cumulative risk assessments.
Some thought that the document is too generic and needs to better present the essence of
a "framework" and highlight the unique aspects of cumulative risk assessment,
particularly tying in environmental health issues, vulnerability of populations, and related
issues.
• Many individuals commented that the document needs editing/reorganization because the
current draft is unbalanced, where some topics receive too much detail, while other issues
deserve more information.
• The document would benefit from diagrams to illustrate the concepts and flow of
information within the three steps: Planning and Problem Formulation, Analysis, and
Interpretation.
-------
• The document should include a better representation of the scientific literature by citing
additional papers published in peer reviewed journals.
Summary of Comments on Process Issues
Discussion on process issues mostly focused on stakeholder issues, with some
commentary on the use of cumulative risk assessment in decision making. Most of the
stakeholder involvement comments concentrated on the level of stakeholder discussion in
the document, how to get stakeholders involved, and the role of stakeholders in
cumulative risk assessment.
• Comments on the draft Framework's stakeholder involvement discussion were generally
positive with some suggestions on possible changes. Several participants thought that the
discussion was too general and should be reduced to focus on those issues that are
specific to cumulative risk assessment.
• Extensive comments were provided on how to promote stakeholder involvement in
cumulative risk assessment projects, particularly with respect to providing a level playing
field for communities and other groups that may not have technical resources available to
them. In particular, it was noted that significant stakeholder involvement will require not
only a "place at the table" for these groups, but access to resources needed to participate
in what can be dauntingly scientific processes.
Several people suggested that the document provide a clearer picture of the potential role
of stakeholders throughout all three phases of the cumulative risk assessment process.
The document should clarify that stakeholders may be involved in the management team
throughout the life of an assessment, which may often require considerable time
commitments.
Several individuals raised the issue of EPA developing trust with stakeholders, through
up front planning and discussion of the expectations and possible outcomes from an
assessment. Specifically, it is important at the beginning phases of a project to clearly
articulate the objectives and discuss the possible end results, including uncertainties and
the required level of confidence in the conclusions.
Discussion on the use of cumulative risk assessment in the larger context of decision
making mostly focused on adding narrative to the document on when it might be
appropriate to use cumulative risk assessment and how to maximize the utility of the
assessment in the larger decision making process. The document could provide further
detail on the kinds of questions that may be addressed legitimately by a cumulative risk
assessment, and the kinds of questions for which such an assessment is not suited. This
discussion could also mention the ways in which the quality of data needed in the
assessment might affect the kinds of questions that can be addressed. These are not
11
-------
purely policy issues, but ones that involve an understanding of the quality of the science
underlying an assessment.
Suggestions were made to present examples of the types of decisions that cumulative risk
assessment can support. Several experts noted that this tool appears to be particularly
applicable to permitting and siting issues where it can be used in a proactive manner to
determine potential risks from new facilities or activities.
Summary of Comments on Technical Issues
Several individuals thought that the draft Framework document needs to address
qualitative risk to the same extent as quantitative risk. They felt an increased emphasis
on qualitative risk assessment is important since it allows for the evaluation of a broader
range of stressors.
• There was general agreement that combining different types of stressors (e.g., chemical
and non-chemical) can be problematic. Various approaches were discussed to address
this issue including developing a matrix and using comparative risk assessment.
Although there was no general consensus regarding appropriate methodology, it was
recommended that the document should maintain a broad scope and include all stressors
even though certain stressors cannot be evaluated quantitatively by existing methods.
An individual noted that the document should emphasize that some kinds of uncertainty
analyses cannot be done quantitatively. The current section in the document should be
expanded to show that the limitations of the analysis include more than just statistical
uncertainties, broadening the concept of uncertainty from just quantitative uncertainty to
actual level of confidence in the assessment.
It was suggested that the technical sections of the document be modified so that they are
more specific to cumulative risk assessment. For example, when discussing vulnerability,
it would be helpful to mention the synergy or antagonism that can occur when several
different stressors are combined. Similarly, when combining chemical and non-chemical
stressors, the reader should be informed that interactions may result and influence the
level of the effect(s).
• Many individuals commented throughout the technical discussion on the importance of
defining the scope of issues to be addressed early on in the cumulative risk assessment
process to help determine the types of analyses that can be done and take into
consideration the resulting level of uncertainty.
in
-------
1.0 INTRODUCTION
1.1 Workshop Purpose
The Peer Consultation Workshop on the Draft Frameworkfor Cumulative Risk Assessment was
held on August 22 and 23, 2001, in Arlington, VA. The workshop was sponsored by the U.S.
Environmental Protection Agency's (EPA's) Risk Assessment Forum and was facilitated by
Versar, Inc. The purpose of the meeting was to provide interested stakeholders with an
opportunity to see an early draft of the Framework and to obtain early input from these
stakeholders that will assist EPA with further development of the Framework document.
1.2 Workshop Participants
A group of 12 experts, from different disciplines and types of organizations, was assembled by
Versar for this peer consultation workshop. These individuals have experience related to
cumulative risk assessment from a variety of perspectives: academia, consulting, industry,
environmental groups, and community activist groups. In addition, the experts were selected
such that the following topic areas would be covered at the meeting: aggregate exposure, risk
assessment for chemical mixtures, accident and transportation risks, epidemiology, community-
based risk assessments, socioeconomic issues, uncertainty analysis, and other topics of
importance in cumulative risk assessment. The list of experts is presented in Appendix A. In
addition to these individuals, about 25 observers attended the workshop. The list of observers is
presented in Appendix B.
1.3 Discussion Topics
A list of discussion topics, presented in Appendix C, was distributed to the participants prior to
the meeting to help stimulate dialogue on technical issues related to the draft Framework
document. These discussion topics were categorized as (1) over-arching issues, (2) process
1-1
-------
issues, and (3) technical issues and represented those topics which have been under deliberation
within the EPA Risk Assessment Forum Technical Panel. This peer consultation, as opposed to
a peer review, was held to obtain input from scientists earlier in the document development
process so the draft Framework can be expanded and improved as it moves forward. The
discussion topics listed were a starting point for the dialogue and participants were encouraged to
raise other issues or topics. However, regulatory policy issues were beyond the scope of the
meeting.
1.4 Agenda
The workshop agenda is presented in Appendix D. The meeting began with opening remarks
including an overview of the agenda for the two-day meeting, presentations from EPA staff on
the background of the draft Framework document, and review of the objectives of the meeting by
the chair. Consistent with the more informal nature of a peer consultation, the agenda was
purposely designed to be flexible to accommodate discussion of new ideas as they arose. The
agenda was organized around the three main discussion topic categories (1) over-arching issues,
(2) process issues, and (3) technical issues. After the morning presentations, most of the
remainder of the first day was devoted to such over-arching issues as the scope and organization
of the draft Framework report. Discussion on process issues began late in the afternoon of the
first day and included considerable discussion of stakeholder involvement in the cumulative risk
assessment process. The technical discussion, on about ten topics, was completed on the second
day. This discussion included such topics as approaches to cumulative risk assessment,
uncertainty/variability analysis, epidemiology, and combining different types of risk. Time was
set aside on both days of the workshop for comments to be made by observers.
1-2
-------
1.5 Workshop Summary Report
This report summarizes the workshop presentations and discussion, with appendices that provide
handouts, other materials used in presentations, and post-meeting written comments from the
participants. The report is organized as follows:
Section 2 of this report summarizes the opening presentations. Overheads used by the
presenters are provided in Appendix E.
Sections 3, 4, and 5 provide summaries of the comments and suggestions of the
participants, with supporting dialogue, organized by the three main discussion topic
categories.
The appendices to this report present the handouts from the meeting (e.g., agenda, list of
discussion topics, slides, and other materials) as well as post-meeting written comments
from the participants.
1-3
-------
2.0 SUMMARY OF OPENING REMARKS
2.1 Welcome
David Bottimore of Versar, Inc, opened the meeting by welcoming participants and observers.
He presented an overview of the agenda, introduced the participants, and described the goals and
intended outcome of the workshop. During his opening remarks, he emphasized that the
workshop was a peer consultation, not a peer review, and was intended to promote dialogue and
provide input on technical issues associated with the draft Framework document earlier in the
process than would be done in a peer review. Mr. Bottimore talked about the effort to assemble a
group of experts with diverse backgrounds and expertise to contribute new ideas to the draft
Framework report. He also noted that in addition to the main discussion sessions, time would be
set aside each day for comments from observers. His opening remarks were concluded by going
around the room for introductions of the participants and observers.
2.2 Background on Risk Assessment Forum and Framework Development
Bill Wood, Executive Director of EPA's Risk Assessment Forum (RAF), provided a short
overview of the RAF and its activities working toward issuing cumulative risk assessment
guidelines. He stated that this peer consultation will be helpful to the Cumulative Risk
Assessment Technical Panel writing the Framework., particularly obtaining input on community-
based risk assessment issues. His presentation began with background on the RAF. He
explained that the RAF's mission is to promote consensus across the agency and apply the best-
available science in risk assessments. The RAF, and technical panels assembled to address
particular issues, are composed of scientists from the Agency's program, regional, and research
offices. The RAF provides guidance to EPA risk assessors through three principal types of
products (1) agency guidelines, (2) guidance documents, and (3) technical papers. Major
2-1
-------
previous efforts included the cancer risk assessment guidelines and the ecological risk
assessment guidelines.
Development of the Framework began in response to a request from the EPA Science Policy
Council that the RAF begin developing guidance on cumulative risk assessment. The draft
Framework document has been developed by a Technical Panel composed of scientists from
about ten Agency program offices and regions. They have been working over the last two years,
consulting twice with the Executive Committee of the Science Advisory Board (SAB) and
recently meeting with other Federal and State government scientists, to obtain early input for the
Framework. Bill Wood emphasized that developing a framework is the first step in the overall
process of developing guidelines on cumulative risk assessment. Publication of the final
Framework in FY 2002 is anticipated, contingent on the outcome of peer review (perhaps by the
SAB). Future guideline development efforts will include activities to identify and develop case
studies and issue papers on various cumulative risk assessment topics. EPA anticipates issuing
proposed guidelines within the next several years. Bill Wood noted that this process is similar to
that used by other Agency efforts, most notably the ecological risk assessment guidelines, which
were started in 1991 and finalized in 1998.
The Framework document is intended to provide a flexible structure, capable of evolving with
experience, that captures the basic elements of cumulative risk assessment. It should serve as a
basis for developing guidelines by defining key terms and principles to promote a common
language and furthering development of the approaches. He stressed that the Framework is not
intended to present substantive technical guidance, but rather provide a general overview of
issues, topics, and approaches that can be considered in conducting a cumulative risk assessment.
2.3 Presentation of the Draft Framework for Cumulative Risk Assessment
Mike Callahan, from EPA Region 6 and Chair of the RAF's Cumulative Risk Assessment
Technical Panel, provided a more detailed presentation on the Framework document. He opened
2-2
-------
his talk by reflecting that cumulative risk assessment is a wonderfully challenging and exciting
scientific topic. The Panel has been working diligently on the Framework and it is helpful to get
technical input from other scientists, such as those assembled for this peer consultation. The
Framework document is intended to present options and possible approaches to cumulative risk
assessment and is not a protocol or a guideline on how to perform such an assessment. EPA will
be developing the guidelines over the next several years, building on the Framework and other
subsequent efforts. Cumulative risk assessment is not a replacement for traditional risk
assessment techniques. Rather, it addresses specific issues growing out of the need to focus on
populations or communities, rather than on sources of pollution. He noted that several EPA
offices are already doing cumulative risk assessments, such as the Office of Pesticide Programs
(under the Food Quality Protection Act) and the Office of Air (as part of the National Air Toxics
Assessment - NAT A). As in Bill Wood's presentation, Mike Callahan emphasized that the RAF
effort is focused on the scientific issues, not policy or regulatory issues.
The Framework for Cumulative Risk Assessment presents information that can be considered in
planning and performing such an assessment. Included in the document are working definitions
and descriptions of the overall process, including the three main phases: (1) problem formulation
(planning and scoping), (2) analysis, and (3) interpretation. Cumulative risk assessment is not
appropriate for every task but it is a tool available to help risk assessors and decision makers.
The process includes both analytical and deliberative aspects that are useful in answering
questions, particularly in circumstances involving multiple stressors and multiple chemicals. It
is important to recognize what a cumulative risk assessment can and cannot do, so people have
realistic expectations. This is also something that groups performing assessments should discuss
during the front end of a project, so all stakeholders come to a common understanding that a
cumulative risk assessment can only answer certain types of questions. The science behind such
efforts is still being developed and many of the challenges in cumulative risk assessment arise
from limitations in methods to aggregate and/or compare risks from different chemicals and
types of stressors. The Framework describes methods that the Agency is aware of, recognizing
the strengths and limitations of the different approaches. Mike Callahan's presentation
2-3
-------
introduced many of the scientific issues related to combining different types of risks, concepts of
vulnerability and susceptibility, and evaluating uncertainty in cumulative risk assessments so the
results will be more meaningful to decision makers.
Mike Callahan concluded his presentation by noting that EPA is already performing cumulative
risk assessments in certain situations and needs to continue to improve methods in the future. He
reiterated that guidelines are four to six years away, but the Framework will help Agency risk
assessors and programs in the near term to apply more consistent techniques in cumulative risk
assessments. The Framework should be completed in early 2002 following peer review, perhaps
by the SAB. Mike Callahan again stated that policy issues are being addressed on a separate
track by the Science Policy Council, and will not be part of the final Framework document. He
noted that case studies and issue papers will be developed following completion of the
Framework as part of the guidelines development process, which will start in 2002. He thanked
everyone for participating in the peer consultation and looked forward to hearing the experts'
thoughts and ideas for improving the Framework.
2.4 Discussion on Opening Presentations
During the opening presentations by Bill Wood and Mike Callahan, several participants raised
issues and asked questions about the draft Framework. Norris McDonald asked Bill Wood about
the extent of industry involvement in the preparation of the draft Framework document. Bill
Wood commented that this is an open process and comments from all sectors of the public would
be welcome here at this meeting, by written comments in response to the Federal Register notice,
or at future peer review meetings. Later, it was clarified that the current draft of the Framework
document was developed by EPA scientists with some input from other Federal and State
government scientists obtained through recent meetings. This peer consultation meeting
provides the first opportunity for broader public input on the Framework document. Amy Kyle,
at the end of Bill Wood's talk, encouraged EPA to write the Framework so that cumulative risk
assessments can be as broad and inclusive as possible, because current risk assessments tend to
2-4
-------
have a narrow focus. She hoped that the document could address this issue and advocate that
assessments consider as many factors as possible. Douglas Crawford-Brown raised an issue
about subpopulations and their vulnerability. He said that in addition to looking at stressors,
another way of evaluating cumulative risks would be to consider the population and what makes
it more vulnerable to risks.
Ken Bogen raised the issue about EPA's legal mandates and whether or not the Framework
should reflect the role of cumulative risk assessment under particular legislation/regulatory
programs. He questioned whether or not cumulative risk assessment would be applied in
situations where an "increased risk" might result from a particular policy or new source. Mike
Callahan answered that cumulative risk assessment is not constrained by EPA's mandates and
that it is a tool that can be used to solve problems and answer questions that are much broader
than those specifically addressed by legislation.
Following Mike Callahan's presentation, Ken Bogen commented on the importance of doing
uncertainty analysis when doing cumulative risk assessments. He noted that aggregating upper
bound risk estimates will result in answers that are overly conservative and not meaningful.
Margo Schwab said that part of the problem with this discussion is the meaning of a quantitative
assessment when the context of the decision being made is not known. It is important to know
what problem is being addressed and what answer is needed before trying to evaluate uncertainty
in a quantitative manner. She concluded by saying that it is important to define up front the
issues of concern and the decisions that need to be made so that the analytical tools selected are
appropriate. Once this is done, it would be appropriate to discuss the ability to evaluate risks in a
qualitative manner, and the resulting uncertainties. Douglas Crawford-Brown added that this
issue is not so much a question of the reliability of the risk estimates, but rather the adequacy of
the safety margin. However the assessment is performed, it will be necessary to make some
statement of the degree of margin of safety, and this is precluded if standard regulatory
approaches to calculation of quantities such as hazard indices are combined. Ken Sexton
2-5
-------
commented that more discussion on uncertainty was planned for the technical discussion session
on the second day.
2.5 Introduction of Chair and Overview of Discussion Topics
Ken Sexton, from the University of Minnesota, was the chair for the peer consultation and served
as facilitator. He started by introducing the three discussion categories for the two-day
workshop. The goal of the peer consultation meeting was restated to emphasize that diverse,
technical input was sought from each participant. There would be no attempt to achieve
consensus through this meeting. Rather, the discussion should bring out the diverse perspectives
of individual experts in the group. He introduced some ground rules and guides to keep the
discussion focused on technical issues related to the draft Framework document. He talked
about the post meeting activities to prepare a workshop report that summarizes the discussion
and made the request that each participant prepare written comments after the meeting that will
be appended to the workshop report.
2-6
-------
3.0 DISCUSSION SESSION I - OVER-ARCHING ISSUES
Discussion on the draft Framework report started with over-arching issues. The experts were
asked to consider the following topics:
Comment on whether the draft Framework adequately captures, describes, and reasonably
organizes the key issues for cumulative risk assessment.
What additional issues, if any, should be covered?
Comment on whether some of the topics presented in the draft Framework are too generic
for the intended purpose.
Comment on whether the draft Framework adequately conveys the state of the science
with respect to cumulative risk assessment and the areas that are in need of further
research and development.
Presented below are summaries of the main comments and suggestions as well as detailed
dialogue from the over-arching issues discussion session. Presented first are the main points
from the session (Section 3.1) followed by more detailed discussion (Section 3.2) which led up to
the final suggestions.
3.1 Summary of Comments and Suggestions
Overall, many experts stated that the draft Framework represents a good start and it is expected
to be a useful document that will help to promote consistency in cumulative risk assessment.
Several individuals commended EPA for writing a concise document that maintains a broad
perspective, and allows users flexibility in planning and implementing cumulative risk
assessments. Some people thought that the document is too generic and needs to better present
the essence of a "framework" and highlight the unique aspects of cumulative risk assessment,
particularly tying in environmental health, vulnerability of populations, risk acceptability criteria
and other topics. Several people felt that the lack of a coherent "framework" to the discussions
3-1
-------
in the document limits its ability to serve as an organizing tool. As written, the draft document
does not present a coherent way for the reader to understand the major issues, how they arise
within the context of a cumulative risk assessment, and why they are important. As such, many
individuals commented that the document needs editing/reorganization because the current draft
is uneven and unbalanced, where some topics receive too much detail, while other issues deserve
more information. For instance, several experts found Section 4 (Interpretation) to be more
useful than Section 3 (Analysis), which was much longer. Revisions are needed to provide better
balance and a clearer hierarchy of the information presented. One individual commented that the
appendix on future research needs is very good, but that the document should introduce this issue
earlier in the body of the report.
The document would benefit from additional examples and diagrams to illustrate the concepts
and flow of information within the three steps: Planning and Problem Formulation, Analysis, and
Interpretation. Particularly, adding a series of diagrams would help the reader to understand the
relationships among the sections of the document that describe the three steps of cumulative risk
assessment. Providing expanded diagrams for the problem formulation phase might also help to
link disparate elements such as physical, chemical, biological, behavioral, and economic
considerations. A proposed "framework" diagram was developed during the workshop
(presented on page 3-6).
Comments were provided on the importance of up front scoping before any analysis begins.
From this exercise, it might be clearer which elements can be analyzed quantitatively and which
will have to be described in more qualitative terms. One suggestion was made to add a list of
questions that might be asked within each step of the process to depict the types of decisions that
need to be made within each step. Appendix F presents flow diagrams developed by Douglas
Crawford-Brown as an example of key questions and decisions to be made within each step of
the conceptual framework.
3-2
-------
Other suggested changes to the document included providing a better representation of the
scientific literature through citations of papers published in peer reviewed journals. Several
individuals provided specific examples of papers that could be cited in the Framework. Also,
some people felt that the document is redundant with previous reports. The Framework should
provide a clearer perspective of the links between this document and previous documents cited,
such as the EPA documents and the National Academy of Sciences/National Research Council
(NAS/NRC) reports listed in the reference section of the draft Framework. Similarly, a comment
was made that the Framework should acknowledge cumulative risk assessment activities in all
parts of the Agency. One expert commented about the intended audience of the Framework and
recommended that stakeholders participate in the process to finalize the Framework, which could
help build trust in the development and implementation of cumulative risk assessments.
Changes such as these should help to bring better balance and a clearer presentation of the salient
issues of cumulative risk assessment.
3.2 Detailed Discussion
Discussion on over-arching issues began with the chair, Ken Sexton, going around the room
asking participants if they were comfortable with the plan to have each person make a short
statement about their general comments on the draft Framework document, followed by more
detailed discussion. After this initial discussion, each person provided a short summary of issues
of concern, which led to more detailed discussion later in the session. Several individuals raised
new issues to be added to the list of over-arching topics for discussion during the session.
Some of the early over-arching comments addressed concerns that the document did not present a
clear picture of the overall framework for cumulative risk assessment and an integrated picture of
the relationships among the individual steps. One participant voiced concerns about the extent of
stakeholder involvement, both in the preparation of the Framework as well as later in the
implementation of a cumulative risk assessment. As this discussion unfolded it became clearer
3-3
-------
that stakeholder involvement (to be addressed in more detail under process issues) would be
adequately addressed. Another participant voiced the concern that the document does not
address immediate (acute) impacts from situations such as transportation accidents or chemical
spills. He felt that the document would be more useful to the public if it explicitly addressed
these types of circumstances where questions arise over siting of facilities and the need to
evaluate the risks from those types of incidents.
Discussion on over-arching issues continued by having each participant provide an overview of
their comments. Elizabeth Boa thought that the Framework is needed because of the many
different approaches being used by different EPA offices. She encouraged EPA to make sure
that the definition of cumulative risk assessment is consistent with other EPA programs and that
the Framework addresses the variety of cumulative risk assessments that are being performed by
the Agency, particularly those under FQPA. Jim Butler felt that the document should be as
flexible as possible, however, he noted that there appeared to be inconsistencies in
definitions/usage in different parts of the document. He suggested that the document be
reviewed for consistency. Margo Schwab felt that there was an uneven level of detail in the
document which posed a challenge in understanding how the various types of stressors (e.g.,
chemical, physical, political, etc.) would be integrated. She felt that the essence of cumulative
risk assessment was not apparent and that the document should be more transparent with respect
to the paradigm that is being used. She also noted that the Framework document should do a
better job of integrating information from supporting documents (e.g., previous EPA reports as
well as reports by the NAS/NRC).
Douglas Crawford-Brown proclaimed that the document is a good start at introducing very
complex issues, but felt that "there is not a framework yet" that gives the reader a clear picture of
the overall process of cumulative risk assessment. He also raised the issue that cumulative risk
assessment needs to address margin of safety for multiple chemicals. Amy Kyle agreed with
Elizabeth Boa's earlier comment that there seems to be a conflict between the FQPA definition
of cumulative risk assessment and the one used in the draft Framework document. She
3-4
-------
acknowledged the issue of scale and the need for a more global approach, such as that described
in the draft Framework document.
Norris McDonald remarked that the document is concise and does a good job of introducing
many of the science issues. He inquired about the target audience for the document and
questioned whether policy issues can be separated from technical issues in practice. He also
revisited and expanded on the issue of whether EPA is going beyond their statutory mandates.
Ken Bogen thought that it might be useful for the document to list current legislative mandates
and how and where cumulative risk assessment might be applied. Mike Callahan clarified that
the Framework is for EPA staff, but acknowledged that people outside of EPA might use the
document as a reference. He also noted that cumulative risk assessment is a tool and the
Framework document is intended to provide information that can be used in a variety of
scenarios, many of which are not explicitly part of the Agency's regulatory programs. Several
participants noted that the Framework and cumulative risk assessment would be applicable for
issues that might not traditionally be part of EPA's scope but which are increasingly the focus of
attention for assessments, such as evaluating air pollution from traffic in local neighborhoods.
Subsequent discussion took place on whether the Framework should be limited to issues that are
explicitly addressed in EPA's legislative mandates, with two participants voicing the opinion that
the document goes beyond EPA's authority.
Steve Olin commented that the Framework is a useful and groundbreaking document, however,
it needs more focus in some parts, while maintaining a broad, over-arching perspective. He also
advocated adding diagrams to help illustrate the methodologies described. Jim Butler reiterated
that the Framework is useful in providing a big picture, but commented that it may attempt to try
to be too much to too many people. He contemplated that the document could be reworked to be
a series of issue papers with multiple frameworks.
Discussion continued with Jennifer Sass revisiting the issue of separating policy from science
when the overriding need for cumulative risk assessment is to make decisions using the best
3-5
-------
available scientific information. Ken Bogen followed up by revisiting the earlier comments for
the document to be clearer on when agency programs should/could do cumulative risk
assessments. Several participants noted that this is a policy issue and beyond the scope of this
meeting. Mike Callahan reiterated that the Framework is a technical document that describes
cumulative risk assessment as a tool that is available to solve problems and answer questions. As
such, it can be used at the discretion of Agency scientists in a variety of scenarios. He stated that
the Science Policy Council is charged with addressing policy issues and noted that the RAF
Technical Panel may be reluctant to include lists of legislative mandates that may limit the
applicability of the Framework or make it become obsolete.
In the afternoon session, discussion on over-arching issues resumed with Ken Sexton illustrating
the overall framework in a flow diagram (below). He proposed that this diagram captures the
overall context and relationships among the steps of a cumulative risk assessment, the key
participants in each step, the type of activity (analytical or deliberative), and the technical issues
involved in each step.
Proposed "Framawork* for Cumulative Risk Assessment (Sextan 8/22/01}
MM**
HfeHK
9jpHWMW BPHpil
Interpretation
Characterisation)
Ibh
IMBI
Analytical and D*llb*ratlv*
A.
S. Problem Form,
Conceptual Model
Analysis Plan
C. Passible Outcomes
A, CumwIcrfivB Inct. Risk
S, Risk Factors
C. itairoirfcBrs
D, Qirwr
A, fciskbt.
EL Risk Description
C. Uncertainty
3-6
-------
Margo Schwab voiced the opinion that the framework for cumulative risk assessment is a layer
above Ken Sexton's diagram. She described cumulative risk assessment as the result of a
function that incorporates such disparate issues as epidemiology/health statistics, chemicals,
health access and many other factors. One aspect that distinguishes cumulative risk assessment
from the traditional risk paradigm is the increased emphasis on the problem formulation step and
the deliberative nature of conducting the assessment. Douglas Crawford-Brown commented that
the framework illustrated by Ken Sexton was correct, but that it was not specific to cumulative
risk assessment. Later in the meeting he distributed a flow diagram that he felt better captured
the problem formulation - driven aspects of cumulative risk assessment (Appendix F).
Norris McDonald returned to the issue of target audience for the document and how stakeholders
will perceive EPA's role in implementing cumulative risk assessments. Ken Sexton answered
that it is difficult for the document to be targeted to anyone except EPA, but acknowledged that
EPA documents such as these often become the de facto standard operating procedures used
throughout the scientific community.
Ken Sexton asked again if the framework that he illustrated seemed to capture the essence of the
cumulative risk assessment process. Several individuals provided comments that the diagram
was helpful, but that it was still too generic and unclear how it was unique to cumulative risk
assessment. Ken Sexton and Douglas Crawford-Brown also noted that none of the steps are
purely analytical or deliberative, but that each step encompassed both aspects. Amy Kyle, citing
the 1996 Reducing Risk Report, emphasized that analysis takes place in each step of the process
and that an iterative approach should be used. Several individuals noted that issues such as
uncertainty and vulnerability actually cut across all three phases and the document should reflect
that these issues should not be limited to the interpretation phase. Margo Schwab again raised
the issue of how cumulative risk assessment is different from the traditional four-step paradigm
and contemplated whether the Framework tries to lay out a new approach that will be developed
over the next ten years. Ken Bogen agreed that the Framework should be a launching pad for
stimulating further developments in the science, but felt that the document does not accomplish
3-7
-------
that goal. He thought that the document does describe ways to aggregate risks from multiple
chemicals, but acknowledged that many issues related to accumulating risks were still
undeveloped. Specifically, he pointed out that issues will arise related to distributions of
individual risks and inequities that call for estimating risks using different approaches (such as
using different potency factors) for different groups. He encouraged EPA to develop examples
within the document that illustrate different scenarios, such as the kind that might be encountered
with siting of facilities where one needs to consider different communities and issues of
vulnerability, behavior and other inequities.
Attention turned back to Ken Sexton's diagram and whether or not it adequately captured the
phases of a cumulative risk assessment. Beth Mileson revisited issues of definition and scope to
question whether cumulative risk assessment is "total" risk, comparative risk, or integrative risk.
She noted that it appears to be focused on "total" risk, but felt that the definition was not clear.
Ken Sexton replied that when the glossary is added to the document, it might help to clarify some
of the definitions and usage of terms throughout the document. Douglas Crawford-Brown added
that the Framework is introducing a new process that is more open ended, deliberative, and
inclusive. As a result, it is important that the Framework introduce the types of topics that can be
addressed but it should not constrain the process. Mike Callahan clarified that the traditional
methods of risk assessment are still useful for certain questions but that cumulative risk
assessment brings a new perspective and methodology to answering new types of questions. The
Framework document introduces some of these issues and the subsequent guideline development
efforts will provide more specifics over the next few years.
Steve Olin commented that it would be helpful to include some hypothetical or actual examples
to illustrate how the process might work. He also suggested that data deficiencies, variability,
interactions, and the potential utility of PBPK modeling were issues that needed further
discussion in the draft Framework document. Jim Butler again questioned the overall goal of the
document and how far the Framework needs to go, recognizing that subsequent efforts will
address some of these issues. Margo Schwab felt that cumulative risk assessment is a step
3-8
-------
forward and a logical evolution from the traditional risk assessment paradigm. She reiterated that
what distinguishes cumulative risk assessment is the problem formulation emphasis, which can
hopefully incorporate epidemiology into such assessments. Jim Butler returned to the idea that
the Framework document is intended to describe what cumulative risk assessment is, rather than
how to conduct such an assessment.
Bill Rhyne posed a question about terminology and usage of the word "stressors," desiring
clarification that it encompassed many factors beyond traditional chemical stressors, such as
noise, physical, socioeconomic, and other threats. Ken Bogen added that the complexity of a
cumulative risk assessment lies in the challenge to evaluate these different types of stressors,
particularly with different susceptibilities and vulnerabilities of communities. Mike Callahan, in
response to earlier comments about limiting the scope of the document to existing EPA
mandates, provided a few examples on why the document needs to be as broad as possible. He
noted that some air pollution issues of concern to communities may be outside of EPA's
purview, but that a cumulative risk assessment will help to identify and quantify the magnitude
of such problems so other types of actions can be taken. He added that stakeholders will use the
information to address issues (e.g., zoning, siting, etc.) that are outside of EPA's authorities.
Amy Kyle commented that it might help to think about the Framework as a model, where
constructing it requires tying the pieces together. Douglas Crawford-Brown agreed and thought a
diagram could be added to help to illustrate how the different pieces and steps relate to the
questions posed. Ken Bogen noted that what is still missing is some type of risk acceptability
criteria, looking at population risk and the number of cases, which he believes is the ultimate
goal of cumulative risk assessment. Douglas Crawford-Brown offered that additive risk is
addressed throughout the draft Framework but examples are needed on more complex types of
risks. Ken Bogen encouraged the authors to look at the NRC report on combining risks.
3-9
-------
4.0 DISCUSSION SESSION II - PROCESS ISSUES
Discussion on process issues focused on the following topics:
• Stakeholder Involvement Throughout the Process
• Using the Results of Cumulative Risk Assessment
• Cumulative Risk Assessment in the Broader Context of Environmental Analysis and
Decision Making
Presented below are summaries of the main comments and suggestions as well as detailed
dialogue from the process issues discussion session. Presented first are the main points from the
session (Section 4.1) followed by more extended discussion (Section 4.2) which led up to the
final suggestions.
4.1 Summary of Comments and Suggestions
Discussion on process issues focused on stakeholder issues, with some commentary on the use of
cumulative risk assessment in decision making. Comments related to stakeholder involvement
concentrated on (1) the level of stakeholder discussion in the document, (2) how to get
stakeholders involved, and (3) the role of stakeholders in cumulative risk assessment.
4.1.1 Stakeholder Involvement
Comments on the document's stakeholder involvement discussion were generally positive with
some suggestions on possible changes. Several participants thought that the narrative was too
general and that other EPA documents referenced in the draft Framework present much of the
same type of information on the process for (and benefits of) stakeholder involvement.
Specifically, two references from the draft Framework document (USEPA 1999b and 200Ib)
4-1
-------
provide much of the background on stakeholder involvement and repetition in this document is
not necessary. In the draft Framework document, the stakeholder discussion should be reduced
to focus on those issues that are specific to cumulative risk assessment. Extensive comments
were provided on how to promote stakeholder involvement in cumulative risk assessment
projects, particularly with respect to providing a level playing field for communities and other
groups that may not have technical resources available to them. Several individuals felt strongly
that the document should call attention to this issue and provide guidance on how to address this
situation. One specific comment on the document suggested that the list of stakeholders
presented on page 17 be expanded to include "affected" industry. Some felt that this list was
geared more toward community-based assessment and that it needed to acknowledge cumulative
risk assessment being applied for more national issues.
With respect to the role of stakeholders, several members of the group suggested that the
document provide a clearer picture of the potential role of stakeholders throughout all three
phases. Stakeholders can be vital in the problem formulation phase (e.g., identifying sources and
chemicals of interest), the analysis phase (in selecting the tools to be used in the assessment), and
in interpreting the results. As such, the document should clarify that stakeholders may be
involved in the risk management team throughout the life of an assessment, which may often
require considerable time commitments. A few experts noted that while it is important to have
participation from stakeholders, EPA should maintain final discretion on major decisions
affecting the outcome of a cumulative risk assessment. Several individuals raised the issue of
EPA developing trust with stakeholders, through up front planning and discussion of the
expectations and possible outcomes from an assessment. Specifically, it is important at the
beginning of a project to determine the possible end results, particularly with respect to the
uncertainties and the required level of confidence in the conclusions. One participant stated that
part of the front end planning should communicate if the assessment will produce a quantitative
answer or if it is more likely that a ranking or comparative risk-type outcome is more feasible.
With this comes the responsibility of all parties to acknowledge the difficulty in performing a
cumulative risk assessment and the limitations of the results.
4-2
-------
4.1.2 Use of Cumulative Risk Assessment in Decision Making
Discussion on the use of cumulative risk assessment in the larger context of decision making
focused on adding narrative to the document on when to use cumulative risk assessment and how
to maximize the utility of the assessment in the larger decision making process. Several
individuals suggested that EPA expand the discussion to describe when cumulative risk
assessment might be an appropriate tool to address a problem. Suggestions were made to present
examples of the types of questions that cumulative risk assessment can and cannot answer.
Furthermore, the Framework should also mention to the reader that a decision should be made up
front to determine if cumulative risk assessment is even appropriate for the issue under
consideration.
Several experts noted that this tool appears to be particularly applicable to permitting and siting
issues where it can be used in a prospective manner to determine potential risks from new
facilities or activities. Several individuals advocated that the Framework document provide
clearer guidance to readers that such assessments can be most effective when done up front,
possibly as a screening-level analysis, to improve decision making on proposed projects.
Participants acknowledged that this type of application may be useful in determining the scope
and magnitude of a problem, which can then trigger more detailed analysis that will assist
decision makers in determining the potential risks from new facilities or sources. Similarly,
several experts noted that because of the complexity of such assessments, it is important to keep
the project manageable in size and scope. One participant added that the results of an assessment
will likely be meaningless if the assessment is so complex such that no one can understand the
process or results. One other point was made was to emphasize that risk estimates cannot be
interpreted on an individual basis rather than, or in addition to, a population basis. This is
particularly the case with epidemiological data. It is a critical misperception that risk assessment
results apply to the individual. For instance, if the risk is 3/100,000 for children who have
exposure X, for instance, we can only say that if 100,000 children have exposure X, we expect 3
4-3
-------
of them to get sick. We cannot say that if a child's exposure is X then he has a 3% chance of
getting sick.
4.2 Detailed Discussion
Ken Sexton opened the discussion by stating the importance of stakeholders in scoping the study
and problem formulation phase. Bill Rhyne described two projects where stakeholder
participation worked particularly well. Margo Schwab noted that stakeholders can be crucial in
considering what stressors should be evaluated, which will help to ensure that the assessment
will provide information on issues of interest to the stakeholders. Jim Butler added that the
CCRI project had a very narrow focus (air toxics), which was decided early in the assessment
because those were the issues of concern to the stakeholders. Bill Rhyne commented that one of
the important roles of stakeholders is to obtain support for a project because they can reach out to
other groups and bring them into the project. Steve Olin emphasized the importance of carefully
identifying who the stakeholders are, noting that industry is an important stakeholder because
they can provide data for use in an assessment. Also, he added that it is important to keep the
stakeholders involved and engaged throughout the process, so they do not lose interest by the
time the study ends and the results are produced. He also commented on the importance of
agreeing up front on the issues to be addressed. Ken Bogen commented that cumulative risk
assessment will be used in big decisions and that the Framework should emphasize that stakes
are higher in cumulative risk assessment. He concluded by adding that stakeholders deserve
access to resources to help them actively participate in the technical decisions that occur
throughout an assessment.
Elizabeth Boa stated that several other EPA documents have addressed the issue of public
participation and stakeholder involvement so the Framework does not need to address some of
the basic issues related to this topic. In particular, EPA has a draft policy out for public comment
that should be referenced here. Douglas Crawford-Brown noted that it may be important to set
boundaries on the level of participation noting that cumulative risk assessments will often take
4-4
-------
considerable time commitments and people often will not stay interested throughout the entire
process. He noted that some groups will not participate at all because of a lack of trust in risk
assessment and EPA in general. Similarly, it is important on the front end to define the decisions
and structure for the interaction among stakeholders as part of setting ground rules. Norris
McDonald added that many groups do not believe the results of risk assessments because they are
often too complex. He voiced the opinion that communities will be more likely to embrace a
process that focuses on a relatively limited number of parameters, particularly those closely
linked to public health. Jim Butler stated that it is important for the public participation
discussion in the document to be broad and flexible so that it can be tailored to site specific
needs.
Douglas Crawford-Brown added that the issue of margin of safety needs to be addressed
particularly because cumulative risk assessments are likely to address situations involving many
types of stressors, including multiple facilities and multiple chemicals. This will naturally raise
questions as to whether margins of safety are eroded by the presence of so many facilities and
risk agents acting simultaneously on a community. Ken Bogen followed up on this comment by
adding that it is important ro recognize how cumulative risk assessments will be used in decision
making and the need for risk acceptability criteria. Ken Sexton stated that cumulative risk
assessment is likely to require a dramatic increase in the time and resources that would be needed
to do such an assessment, which should be discussed more in the Framework document. Mike
Callahan acknowledged that cumulative risk assessments can be conducted at many levels,
including as a screening tool, which can help to keep the scale of the study more manageable and
reduce the level of effort. Douglas Crawford-Brown added that the document should also
mention that the level of uncertainty may also be much greater than in a traditional risk
assessment. Mike Callahan agreed that uncertainty is likely to be greater, however, decision
makers need information and even uncertain information is an improvement over no information
at all. Margo Schwab noted that both quantitative and qualitative analysis can be done and that
up front planning will help to clarify which of the issues can be addressed quantitatively.
4-5
-------
Douglas Crawford-Brown continued the discussion by noting that as stakeholders get involved in
an assessment, basic decisions need to be made on what issues are to be addressed, whether risk
assessment is the best tool for addressing the concerns, and the scope of the assessment. Norris
McDonald added that this is a good opportunity to work with communities and some local
groups might take the opportunity to lead projects. Jim Butler noted that the Chicago
Cumulative Risk Initiative worked very well because one person was selected by the stakeholder
organizations to represent all of the stakeholder groups involved in the project. He suggested
that the document show examples of cases where stakeholder involvement has and has not
worked.
Amy Kyle revisited the issue of a level playing field for stakeholders and acknowledged that this
is a challenge. Ken Bogen stated that the idea of a level playing field is even more important in
cumulative risk assessment because of the complexity involved. Stakeholder groups should have
a "right to counsel" to help them participate throughout the process, otherwise their involvement
will be meaningless. Amy Kyle added that the document should be explicit on what it might take
to provide a level playing field for communities who may not have the same technical resources
that other groups, such as industry, might have. Ken Sexton agreed with this discussion and
added that the document should reflect that stakeholders have roles in all three steps of a
cumulative risk assessment (the document currently only clearly reflects stakeholder involvement
in the problem formulation phase). Douglas Crawford-Brown concurred, noting that
stakeholders can be critical players throughout the process, for instance in selecting the models to
be used in analysis. Beth Mileson agreed that stakeholder involvement can be influential in those
types of decisions. Steve Olin also acknowledged the need to have better stakeholder
involvement throughout all steps of the process. In the interpretation phase, all stakeholders can
play a role in drawing conclusions from the analysis. He acknowledged that some community
groups will know more about the local landscape which will be helpful in identifying stressors
and sources during the problem formulation phase as well as in interpreting the significance of
the results. Ken Sexton noted that it is often difficult to obtain community participation because
of a lack of interest. He hoped that the situation would be different for cumulative risk
4-6
-------
assessment because the stakes might be higher due to the nature of issues likely to be addressed.
He encouraged EPA to review a document from Environmental Science and Technology, 1999,
by Chess & Purcell, about what works and does not work with respect to community
involvement.
Norris McDonald asked whether there is a difference between the target audience for the
Framework document and stakeholders and wondered if states should be included as
stakeholders. Ken Sexton mentioned that stakeholders include many types of groups and that the
document discusses the importance of assembling a risk management team. He commented that
the document might be expanded to address this issue. Steve Olin thought that the list on page
17 was a good starting point but that it should be expanded to include the "affected' industry.
Jim Butler noted that the text states that EPA will decide if stakeholder involvement is necessary,
which he felt was too vague with respect to how these types of decisions might be made. Beth
Mileson added that there are so many different types of risk assessments and decisions to be
made other than community-based studies that it is important to recognize other types of
scenarios that call for national scale decisions. She added that she agreed with the sentiment
expressed in the document that said basically "if you are not going to take the advice of
stakeholders, don't invite them." Amy Kyle commented that she felt the document handled the
stakeholder involvement situation quite well, particularly considering that each situation will call
for specific needs.
Douglas Crawford-Brown, in response to Jim Butler's earlier comment, suggested that some text
be added regarding EPA's discretionary power on when and when not to include stakeholders.
Elizabeth Boa stated that it is hard to know in advance if an issue is significant enough to include
stakeholders. Ken Bogen acknowledged this situation and suggested the possibility of
performing a screening analysis first which might trigger the need for more study and prompting
notification of the public so they can get involved. Douglas Crawford-Brown disputed this
hypothesis because of the importance of the stakeholders' knowledge in the initial scoping and
screening process, where significant sources or exposure routes could be missed without their
4-7
-------
involvement. Norris McDonald stated that communities are often in front of the government in
identifying problems and provided examples where local groups initiated efforts themselves. He
acknowledged the difficulty in performing a cumulative risk assessment, but provided a warning
that if it is "impossible to do it" then the results will not be useful or believed by the
stakeholders. Margo Schwab followed up by emphasizing the importance of discussing the
expected outcomes up front so it is clearer what the results will represent, particularly relative to
the qualitative conclusions. This may also entail planning the type of analysis (e.g., ranking or
comparison) as well as prioritizing attention on select stressors. Douglas Crawford-Brown added
that this type of up-front planning goes along with EPA becoming more proactive in the process.
Beth Mileson commented that people have different ideas of what cumulative risk assessment is
and what the questions to be answered should be. She also recognized the importance of
focusing the assessment so it does not get overwhelming. Douglas Crawford-Brown emphasized
the importance of the initial stakeholder phase and getting people to agree on the scope of a
project so people do not back out at the end of the process because they do not like the results.
Ken Sexton added that it may be helpful in the problem formulation phase to establish a time line
on how long it may take to obtain results and to determine if that time line is consistent with the
decision makers' needs. Ken Bogen thinks that the Framework document should point out
responsibilities of parties such as state legislatures, which might play a role in setting boundaries
for the assessment as well as the degree of stakeholder involvement. Douglas Crawford-Brown
thought that this issue should not be addressed in the Framework, as it is a policy issue. Ken
Sexton noted this as an area of disagreement where some people think that certain issues should
be addressed in the Framework., while others think it is outside of the scope of this technical
document.
Ken Bogen introduced the issue of when a cumulative risk assessment should be conducted. He
recommended that the Framework advocate the use of cumulative risk assessment to identify
problems before they occur. He stated that the tool is particularly useful for siting decisions, in
permitting, and in other situations where risks should be evaluated before an action is taken that
4-8
-------
may result in unacceptable risks. He lamented that too often such an assessment will be
conducted after the fact when it is too late to remedy the situation. He encouraged EPA to revise
the document to state that cumulative risk assessment will be most useful to decision makers if it
is used as a planning tool at the beginning of the process. He emphasized that cumulative risk
assessment can be particularly powerful if used in a proactive manner. Douglas Crawford-Brown
acknowledged the importance of such use of cumulative risk assessment, but added that
decisions on when to use cumulative risk assessment can be made by the scientific community,
the policy community, and the stakeholders. Elizabeth Boa stated that it might be helpful to have
language in the document on what situations might be appropriate for cumulative risk assessment
to be used. Douglas Crawford-Brown suggested that alternate text could be added to the
document to achieve this goal. He suggested that the document provide examples of the types of
situations where cumulative risk assessment is particularly appropriate, such as addressing issues
of multiple facilities and chemicals, as well as in the siting of new facilities. Ken Bogen agreed
that the document should point out to the reader the types of uses and questions that can be
answered by a cumulative risk assessment. He felt strongly that EPA has an obligation to use
cumulative risk assessment in a manner that is most beneficial to society, which would be before
decisions need to be made and actions are taken. Douglas Crawford-Brown noted that there are
societal considerations that are beyond the scope of the Framework which are under the purview
of society in general.
4-9
-------
5.0 DISCUSSION SESSION III - TECHNICAL ISSUES
Ken Sexton identified the technical issues for discussion as follows:
Approaches to Cumulative Risk Assessment
Biomarkers of Exposure or Effect
Uncertainty/Variability in Cumulative Risk Assessment
Vulnerability
Combining Chemical and Non-chemical Stressors
Combining Different Types of Risk
Limits on Information Provided by Cumulative Risk Assessment
Distinguishing Between Cumulative Risks and Cumulative Impacts
Ken Sexton then asked the group if there were certain topics under the technical issue category
that they felt should not be discussed or, conversely, if additional issues should be added. Margo
Schwab recommended that "epidemiology" be added as a topic because she felt that
epidemiological studies provide data on health impacts at the population level. She also said that
it is important to make the distinction between individual health studies and population studies
and that the Framework document is a good place to address the broad scope of epidemiological
studies.
Several experts noted that the draft Framework document addresses quantitative risk but does not
discuss qualitative risk to the same extent. For that reason, they recommended that the issue of
"completeness" (including both qualitative and quantitative risk assessment) be added to the list
of technical issues. Douglas Crawford-Brown stated that there are established methods for
dealing with variability and uncertainty in risk assessment and wondered if these methods can be
used in the same way for cumulative risk assessment. To address Douglas Crawford-Brown's
comment, Ken Sexton recommended that "variability" be added to the "uncertainty" technical
issue.
5-1
-------
Following this brief discussion, the participants addressed the technical issues. The following
section is organized by technical issue in the order they were discussed and includes the overall
recommendations on how to address that particular issue in the Framework document.
5.1 Vulnerability
Ken Bogen stated that the Framework document should emphasize the importance of treating all
endpoints consistently when possible with regard to vulnerability. There is a current disparity
with carcinogenic and non-carcinogenic risks such that there is a means of addressing
vulnerability for non-carcinogenic risks (e.g., modifying the uncertainty factor as 1, 10, 100, etc.
to account for more vulnerable segments of the population) but there is no equivalent method to
address vulnerability for carcinogenic risks. Ken Bogen stressed that analyses should be done as
consistently as possible in cumulative risk assessment which would involve harmonization of
carcinogenic and non-carcinogenic risks, possibly through the use of consistent uncertainty
factors in cancer and non-cancer assessments.
Margo Schwab noted that another way to address this disparity would be to take out the
information on the cancer/non-cancer dichotomy and say that the methodology is currently
evolving toward having a similar, systematic approach for different endpoints. Beth Mileson
commented on the complexity of the vulnerability issue and recommended that the Framework
keep a placeholder for methodologies not yet developed so that information can easily be
incorporated at a later time. Amy Kyle said that the document should address, when you are
considering different types of endpoints, how you can summarize these effects. She agreed that
harmonizing results should be a goal but does not think the document should endorse certain
directions for harmonizing (i.e., should not try to combine chemicals' effects when they have
different mechanisms of action).
Margo Schwab stated that the issue of vulnerability should come up sooner in the document and
that it needs to be addressed as an overarching issue. Amy Kyle felt that the question of health
5-2
-------
disparities as related to the issue of vulnerability needs to be addressed in the document. Norris
McDonald wondered if there is enough information on environmental justice as related to
vulnerability in the document. Finally, Steve Olin said that the synergy of vulnerability is
missing in the current document and that it needs to be addressed when dealing with multiple
stressors and vulnerability. Subsequent discussion addressed the importance of incorporating
social sciences into cumulative risk assessments, particularly those that address complex non-
chemical stressors.
5.2 Combining Chemical and Non-Chemical Stressors
There was general agreement that combining chemical and non-chemical stressors is problematic
and the experts offered various suggestions as to how this issue could be addressed in the
Framework document. Douglas Crawford-Brown said one way to address the issue would be to
make the text in the document general enough so that it could apply to both chemical and non-
chemical stressors (e.g. by reference to fraction of individuals with an effect, severity of effect,
etc, all of which are concepts that apply to all effects regardless of cause). Ken Bogen noted that
the document should say early on that it is not possible to quantify everything with current
science and should not give the indication that you can. Jim Butler commented that because this
is a Framework document it should have a broad scope and include all stressors and that it is
alright that you don't know how to deal with all of them quantitatively. He added that even
though the document is not focused on methodology, it might be helpful for the reader to have
examples on how chemical and non-chemical stressors could be combined.
Margo Schwab indicated that one example of combining stressors would be to have a simple
matrix (e.g., with toxicity high and low and exposure high and low, and then consider the
high/low for other factors, like vulnerability (e.g., sensitivity of geographic areas - where one
area has 95% health insurance coverage and another area has 50% health coverage). So even
though you cannot discuss socioeconomic issues on the same level as chemical toxicity, you can
still address it along with chemical stressors and consider it in decision making. Beth Mileson
5-3
-------
noted that since all factors cannot be quantified to the same extent, another approach would be to
consider comparative risk analysis. This would allow for the evaluation of different factors that
cannot be easily combined. Amy Kyle added that it is important to try to identify all the different
stressors and that even though it is problematic to compare apples and oranges there is some
merit in doing it as long as you acknowledge the shortcomings. Ken Sexton noted that the draft
Framework makes explicit the type of value judgments that may often be made and that value
judgments are more important in these assessments than in traditional human health risk
assessments (since multiple types of effects are being combined, and some judgment must be
made as to their relative severity).
Steve Olin noted that the interactions that can occur from different types of stressors need to be
considered in the Framework document. He gave the example of chemical exposure combined
with radiation, and cited the McKone and Chen paper, as a good recent reference. Also, Ken
Bogen said that the major stressors to be included in the evaluation need to be identified early on,
otherwise anything could be included as a stressor at any point in the assessment and that would
be problematic.
5.3 Combining Different Types of Risk
Some reviewers felt that the current discussion regarding combining different types of risk in the
draft Framework document was adequate, while others thought that some changes and additional
information were needed. Ken Sexton said he thought the document was appropriately
ambivalent in this area, stating that combining different types of risk can be done many different
ways without really advocating one particular method. Douglas Crawford-Brown was also
generally comfortable with the section although he thought some text should be included which
states that there are disparities in combining different types of risks, that it cannot always be
done, and that failure to place all risks into a common metric does not make the assessment
irrational.
5-4
-------
Ken Bogen noted that the section left out one of the more obvious methods used to combine
different types of risk. This method, used in the 1994 NRC report, is a systematic and
probabilistic approach to endpoints that can be addressed in that way (for cancer, it is the
occurrence of cancer, for non-cancer it is the occurrence of a particular effect). According to Ken
Bogen, the risk becomes more straightforward when using the probabilistic approach, yet the
document barely addresses this methodology. Jim Butler said that the section mentions GIS as a
means to combine different types of risk in passing and thought that some additional text would
be helpful, informing the reader that GIS can provide a visual way of developing a matrix (by
allowing the viewer to overlay maps of different effects, and look for concordances).
Amy Kyle felt that the current section was not really fleshed out and that text should be added to
show a way of representing risk from different types of measures. Margo Schwab added that this
is an example of how epidemiology could be included in the analysis, particularly on background
risks and exposures. Norris McDonald noted that cumulative risk assessment in general should
focus on problem-solving when comparing different types of risk. Finally several reviewers
agreed that the discussion regarding background exposure should probably be expanded in the
document. They felt that the fact that cumulative risk assessment addresses background
exposure is important since traditional risk assessments generally do not. Overall, people felt
that the document can present the different approaches available, assess the strengths and
weaknesses of each, and not recommend use of any particular one.
5.4 Uncertainty/Variability Issue
The experts offered several general suggestions for modifying this section. Steve Olin noted that
the current discussion on uncertainty is somewhat generic and is not really specific to cumulative
risk assessment. He added that text on this topic could be combined with a discussion of the
importance of variability and how it should be considered in cumulative risk assessment. Ken
Bogen felt that the document needs to emphasize the fact that meaningful quantitative estimates
in cumulative risk assessment can only be generated if you adequately address uncertainty. The
5-5
-------
fact that all inputs in cumulative risk assessment are likely to be typically upper-bound values
underscores the need to track the level of uncertainty.
Douglas Crawford-Brown commented that people reading the document need to know why the
emphasis on uncertainty analysis is important (e.g. that it is related to issues such as margin of
safety). He also thought the document should say that some kinds of uncertainly analyses cannot
be done quantitatively, but that this should not be an excuse to ignore them. Building on that
point, Jim Butler said that the current section should be expanded to show that the limitations of
the analysis include more than just statistical uncertainties. This broadens the concept of
uncertainty from just quantitative uncertainty to actual confidence in the assessment. Margo
Schwab said the uncertainty issue illustrates the need to determine up front what the scope and
issues of the assessment are and let the reader know how certain you can be about the results and
how much uncertainty will result. She added that this kind of discussion may be used to help
frame the stressors to be evaluated.
In terms of specific comments, Amy Kyle said that she did not agree with the point on page 56
which states that information that is more certain should be weighted higher. She felt that it is
wrong to evaluate different factors that way. Douglas Crawford-Brown noted that for the second
set of bullets on page 57, the 3rd bullet on models should be combined with the second bullet.
Also, Bill Rhyne recommended that an example be included in the section to show how
uncertainty analysis can be helpful.
5.5 Approaches to Cumulative Risk Assessment
Several reviewers felt that changes were needed in this section. Douglas Crawford-Brown
thought it was confusing to have the four approaches presented separately in the section, since
you can use any combination (including all of them) in doing a cumulative risk assessment. Jim
Butler said that the section should be expanded to include more approaches than the four
currently listed, which seem to be limited to a chemical focus. He mentioned screening and
5-6
-------
hazards analysis as examples for inclusion. The screening and hazard analysis approach involves
not calculating risks but comparing ambient air concentrations (and other factors) and then using
a weight-of-evidence approach and using geographic data. He concluded by encouraging EPA to
consider options other than calculating risks that can give a sense of the hazards in a geographic
area.
Amy Kyle commented that the section title, General Approaches to Cumulative Risk
Assessment, is misleading because currently only a few random approaches are discussed. She
suggested that instead of starting from the methods, it might be better to focus on how to address
the scope of issues in the assessment (i.e., determine what can be evaluated quantitatively and
qualitatively) and then discuss appropriate methods for these types of assessment. Douglas
Crawford-Brown agreed with Amy Kyle's comment that the section appears to be a random
collection of methods and sensed that too many things are jammed into this section that do not
necessarily all go together.
5.6 Biomarkers
This part of the discussion focused more on the role of biomarkers in cumulative risk assessment
rather than the specific text in the draft Framework document. Amy Kyle noted that biomarkers
have particular value in cumulative risk assessment because they integrate exposures. Therefore,
some consideration should be given as to what biomarkers could be used for. Douglas Crawford-
Brown said that biomarkers can be used for ground-truthing the models' results (for models used
in cumulative risk assessment) and to see how far the models are off. Margo Schwab agreed that
biomarkers can give you an indication of ground-truth in cumulative risk assessment, in that it
measures the total assault. She said that biomarkers will be useful, similar to other specific
health indicators. You can also use other health indicators similar to the cumulative hazard
approach for the Chicago study as another means to evaluate cumulative risk.
5-7
-------
Ken Sexton noted that biomarkers have implications for biomonitoring and that the CDC is
attempting to do a nationwide study using biomarkers. Jennifer Sass noted that such nationwide
studies could benefit from collecting additional information. For example, she said that having
information included on a person's death certificate regarding their occupation(s) and the
duration of their occupation(s) could provide beneficial information. Margo Schwab made the
point that even though we are collecting data that we don't know how to interpret now, we may
be able to at some point, then we can go back and use the data for more analyses. She added that
sometimes surveillance of certain health characteristics can be initiated based on the need for
data, such as a health tracking study being conducted by the Pew Commission. Steve Olin
concluded this discussion by noting that there are certain limitations to the use of biomarkers in
cumulative risk assessment.
5.7 Distinguishing Between Cumulative Risks and Cumulative Impacts
Douglas Crawford-Brown said that 'risk' is the probability of the situation occurring, and
'impact' is the realization of the situation happening, but was not sure if that is what the authors
of the document meant in developing their distinction. It also is not clear why the distinction was
being drawn; i.e. what effect this would have on the conduct and interpretation of a cumulative
risk assessment. He does not want the document to be over-quantitative, because sometimes you
can deal with probability without the actual numbers. Ken Bogen felt that you should say what
you can quantify up front to reduce the residual issues left over to deal with. Although Jim
Butler thought that Ken Bogen's comment might apply in the long-term, he said that because the
Framework is so broad, there is no way that you can quantify everything. He interpreted the
section as you can have different approaches in cumulative risk assessment such as traditional
(quantitative) risk assessment and more qualitative assessment such as NEPA.
Norris McDonald said he thought that the purpose of the Framework document and its
limitations had been fairly well-defined on the last paragraph of page 58. He suggested that the
text be brought forward earlier in the document so that the limitations of cumulative risk
5-8
-------
assessment could be clearly understood. Ken Bogen agreed and thought the word 'chemicals'
should be replaced with 'toxic agents.'
Beth Mileson questioned whether the document made an adequate distinction between
cumulative risk and cumulative impacts. She felt that the document did sometimes, while at
other times the distinction was blurred. Mike Callahan of EPA responded that the document is
intentionally ambiguous so that it would increase applicability for the reader. Also, when the
document says "adverse impact" it should not be interpreted as health concerns only. Amy Kyle
said she thought it was good for the document to be intentionally ambiguous. Following Mike
Callahan's explanation the participants agreed that it was not necessary to change the current
terminology. At the close of the discussion, Ken Sexton reiterated the comments that the
distinction between impact and risk is somewhat blurred in the document. He also questioned
why cumulative impacts needed to be included in the discussion at all.
5.8 Epidemiology Studies
Margo Schwab said that there are two types of epidemiological studies; analytical (trying to
establish cause and effect, usually involving clinical trials and cohort studies) and descriptive
(surveillance of health statistics in an area, sometimes with individual information - the goal of
these studies is to describe an area). She said that it is important to make a distinction between
these two types of studies. Epidemiology studies serve as a surveillance indicator that can be
used to assess past as well as present effects at the population level. As a result, these health
indicators can determine the background level which can be pulled into the cumulative risk
assessment and allow you to study the effect from the project(s) and site(s), of interest. Beth
Mileson suggested that epidemiology studies can be used as a trigger to do a cumulative risk
assessment. At the close of this discussion Ken Sexton noted that some reorganization of the
epidemiology text is probably needed in the document.
5-9
-------
5.9 Issue of Completeness (Quantitative/Qualitative)
This discussion was somewhat limited. There seemed to be agreement among the participants
that qualitative assessment needs to have a similar level of discussion in the document as
quantitative assessment. Douglas Crawford-Brown commented that the document should state
that currently available methods will not necessarily drive the overall type of assessment to be
done. The document should say up front how complete the analysis will be. He then noted that
completeness of conception and rigor of execution are two separate concepts and they should be
kept separate and given equal weight.
5-10
-------
6.0 OBSERVER COMMENTS AND DISCUSSION
During the observer comment period on the first day, three people made short statements, Lowell
Smith, Alec McBride, and Elizabeth Margosches.
Lowell Smith, from EPA/NCEA (but representing himself), voiced the opinion that the document
should be as broad as possible. He explained that EPA's research roles extend beyond just those
areas addressed in the legislative mandates (e.g., global warming) and felt that the Framework
would be most useful to the scientific community if it provides a broad perspective on how one
might approach such cumulative risk assessment. He also stated that the Framework needs to
reflect the complexity of the types of issues that are likely to be encountered, ones that go beyond
just chemical stressors and incorporate ecological, economic, cultural, and other factors that often
must be considered in a decision. He provided examples of the complex issues that EPA's
research program addresses and supported efforts to revise section 2 of the document to reflect
the complexity of these issues.
Alec McBride, from EPA/OSWER, followed up on earlier discussion about the target audience
for the Framework and noted again that the document is focused on EPA as the target audience,
because EPA cannot require that people outside the Agency use it. He emphasized that the
Framework will have impacts beyond just EPA because the scientific community is likely to use
the document for technical information when planning and implementing cumulative risk
assessments.
Elizabeth Margosches, from EPA/OPPT, suggested that possibly a modular approach could be
used to reorganize the report, which might help with the unevenness in the presentation of
information. Based on previous experience with risk assessment documents and guidelines, she
suggested that some of the detailed discussion could be pulled out of the main body of the report
and placed in appendices. This would make the Framework clearer to the reader.
6-1
-------
On the second day of the meeting, brief observer comments were presented by Henry Schuver,
Reggie Cheatham, Lowell Smith, and Elizabeth Margosches. Henry Schuver, from EPA's
RCRA program, commented that for a risk model to be reasonable it must be calibrated with
existing health effects data. Reggie Cheatham, from EPA's Office of Environmental
Information, stated that it is important to recognize the role of stakeholder involvement in the
analysis and interpretation phases. He noted that one aspect of this issue relates to determining
how decisions to go forward with an assessment (or not to) can be influenced by stakeholder
involvement. Lowell Smith felt that the stakeholder involvement discussion was good but
suggested that the Framework include a recommendation that communities coordinate with
EPA's Office of Environmental Justice during the cumulative risk assessment process. Elizabeth
Margosches mentioned that areal data presentation is a strong analytical tool that can help to
related disparate types of data. She mentioned the work of Lance Waller and other researchers
who are using spatial data analysis tools including GIS for analysis of public health data and for
making inferences from such data sets.
6-2
-------
APPENDIX A - LIST OF PARTICIPANTS
-------
Peer Consultation Workshop - Draft Framework for Cumulative Risk Assessment
Panel Members
Name
Ken Bogen
Elizabeth Boa
Jim Butler
Douglas Crawford-Brown
Amy Kyle
Norris McDonald
Beth Mileson
Stephen Olin
Bill Rhyne
Jennifer Sass
Margo Schwab
Ken Sexton
Affiliation
Lawrence Livermore National Laboratory (LLNL)
American Chemistry Council
Argonne National Laboratory
University of North Carolina, Department of Environmental
Sciences and Engineering
University of California, Berkeley School of Public Health
Center for Environment, Commerce and Energy
Technology Sciences Group, Inc.
International Life Sciences Institute (TLSI)
H&R Technical Associates, Inc.
Natural Resources Defense Council (NRDC)
Johns Hopkins School of Public Health
University of Minnesota School of Public Health
A-l
-------
APPENDIX B - LIST OF OBSERVERS
-------
Peer Consultation Workshop - Draft Framework for Cumulative Risk Assessment
Observers
Name
Lucy Ament
Jeanne Bailey
Ed Bender
Michael Callahan
Angelina Duggan
Steve Gibb
Loren Hall
Marty Halper
Roger Hawks
Edward G. Jordan
Steve Knott
Elizabeth H. Margosches
Alexander McBride
Pat Phibbs
Alan Roberson
Henry Schuver
Scott Schwenk
Lowell Smith
Henry Topper
Cindy Gordon
Bill Wood
WALK-INS
Terry Quill
Martin F. Kovacs
Margaret MacDonell
Reggie Cheatham
Affiliation
Pesticide & Toxic Chemical News
American Waterworks Association
EPA -Office of Science Policy
EPA Region VI
American Crop Protection Association
Risk Policy Report
EPA-Office of Civil Rights
EPA -Office of Environmental Justice
BASF Corporation
BASF Corporation
EPA-Risk Assessment Forum
USEPA (7403M)
EPA, Office of Solid Waste
BNA-Daily Environment Report
American Waterworks Association
USEPA-OSW-RCRA Corrective
Action
EPA-Risk Assessment Forum
EPA - NCEA
EPA- Office of Pollution Prevention
and Toxics
American Petroleum Institute
EPA-Risk Assessment Forum
Duane Morris
Toxcel LLC
Argonne National Laboratory
USEPA
(202) 887-6320/ lament@crcpress.com.
(202) 628-8303/jbailey@awwa.org
(202) 564-6483
(214) 665-2787
(202) 872-3885
(703) 4 16-85787 Riskpolicy@yahoo.com
(202) 564-7289
(202) 564-2601
(919) 547-2870/ hawksr@basf.com
(919) 547-2889/jordane@basfcom
(202) 564-3359
(202) 564-7636
(703)-308-0466
(202) 452-4 106/pphibbs@bna.com
(202) 628-8303/aroberso@awwa.org
(703) 308-8656/schuver.henry@epa.gov
(202) 564-6667
(202) 5 64-3 3 89/smith.lowell@epa.gov
(202)-260-6750
(202) 682-8482/gordonc@api.org
(202)-564-3361
(202) 776-7894
(703) 335-5670
(630) 252-3243
(202) 260-3085
B-l
-------
APPENDIX C - DISCUSSION TOPICS
-------
DRAFT August 9, 2001
The Risk Assessment Forum
Framework for Cumulative Risk Assessment
Discussion Topics
August 22nd and 23rd, 2001
Arlington, VA
Discussion Questions
The following topics are provided to help guide the discussions during the peer
consultation workshop. When considering these topics, keep in mind that the purpose of a
framework is to identify key issues to inform a wide variety of interested and affected parties.
Therefore a balance must be struck between adequately characterizing the issues and providing
an excessive level of technical detail. The framework also discusses a process for gathering
information and defining the problem with stakeholders. When considering the process topics,
recognize that our goal is to focus the scope of the cumulative risk assessment on a parsimonious
set of stressors, sources, pathways and potential adverse effects that are of interest to a particular
place or community.
Overall, the RAF technical panel is seeking discussion and input on the questions and
topics that follow. We are not setting policies for how to deal with the issues, but rather
identifying principles and professional judgements which may lead to more specific guidance
later.
Comment on whether the Framework adequately captures, describes, and reasonably
organizes the key issues for cumulative risk assessment? Include in this discussion
whether key terms have been identified and clearly defined.
What additional issues, if any, should be covered?
Keeping in mind that the Framework is intended to address issues that are specific or
unique to cumulative risk assessment, comment on whether some of the topics presented
in the draft Framework are too generic for the intended purpose?
The Framework attempts to identify where methods and data are currently lacking or are
underdeveloped for application to cumulative risk assessment. Based on a suggestion
during the meeting with other Federal scientists, research oriented issues have been
highlighted in an appendix to the Framework. Comment on whether the Framework
adequately conveys the state of the science with respect to cumulative risk assessment
and the areas that are in need of further research and development.
C-l
-------
DRAFT August 9, 2001
To assist with this discussion, the following list includes a description of some of the key
technical issues encountered by the Risk Assessment Forum Technical Panel during
development of the draft Framework for Cumulative Risk Assessment. This is not a
comprehensive list of cumulative risk assessment issues but could serve as a starting
point for the review discussions. Where possible, the relevant sections of the Framework
document have been referenced for each issue.
Process Issues
Stakeholder Involvement Throughout the Cumulative Risk Assessment Process
[Framework Section 2; especially 2.1.3 and 2.3]
Stakeholder involvement is recognized as an important aspect of
cumulative risk assessment. The draft Framework highlights this
involvement in the Planning and Problem Formulation Phase of
the assessment. Examples of interested and affected parties are
provided and the importance of defining roles and responsibilities
is discussed. In addition, the importance of discussing the possible
outcomes of the assessment with stakeholders is emphasized.
What suggestions do you have for engaging stakeholders early?
What kind of background or ground rules would be helpful? What
roles and responsibilities would be most appropriate for
communities and interested parties?
Using the Results of Cumulative Risk Assessment [Framework Section
4.5]
The intended use of the cumulative risk assessment should be
considered at each step of the process, from Planning and Problem
Formulation through completion of the assessment. It is important
to recognize that the cumulative risk assessment provides
important information but is not the only contributor to the
decision making process. How the cumulative risk assessment fits
into the decision making process should be addressed when
discussing possible outcomes with stakeholders during the
planning phase. What kind of information could stakeholders
contribute to cumulative risk assessments? Under what
circumstances do you feel stakeholder participation is most
effective?
C-2
-------
DRAFT August 9, 2001
Cumulative Risk Assessment in the Broader Context of Environmental
Analysis and Decision Making [Framework Sections 1.4 and 1.5 and
Section 4; especially 4.4.2 and 4.5]
An issue related to using the results of cumulative risk assessment
is understanding how these assessments fit, both conceptually and
practicably, into broader contexts. For example, cumulative risk
assessment may be a component of assessing cumulative effects
under the National Environmental Policy Act (NEPA). Further,
these assessment may support Community-Based Environmental
Protection efforts. What are the most appropriate situations for
conducting cumulative risk assessments?
C-3
-------
DRAFT August 9, 2001
Technical Issues
Approaches to Cumulative Risk Assessment [Framework Section 3,
especially introduction]
There are a number of ways to approach cumulative risk
assessment, either starting with the NRC paradigm or using a
different approach. Each approach may present its own challenges
in methods, data, and analysis. Four example approaches are
discussed in the draft Frameworkfor Cumulative Risk Assessment:
combining toxicity before calculating risk, calculating risk
factors, using biomarkers and biomonitoring, and calculating
other types of probability statements Are important approaches
missing from this list? Comment on whether there is a better way
to organize this discussion on approaches to cumulative risk
assessment.
Using Biomarkers of Exposure or Effect [Framework Section 3.1.3]
Among the approaches to cumulative risk assessment discussed in
the draft Framework., use of biomarkers and biomonitoring
information shows promise for providing an integrated measure of
where an individual falls on the continuum between exposure and
effect. However, existing methods are inadequate for assessing
complex situations involving a large number of stressors. Further,
experience using biomarkers and related information in decision-
making is limited. Comment on the discussion of biomarkers of
exposure and effects? Given the current state of the science,
discuss the utility of this approach in the near term and in the
future. Does the Framework include the appropriate level of
discussion of this topic? How can the discussion be improved
(consider, for example, whether the Framework discussion
adequately characterizes how biomarkers can be used with more
traditional risk estimates)?
Uncertainty in Cumulative Risk Assessments [Framework Section 4.3]
Uncertainty analysis is an important aspect of risk assessment (and
policy analysis in general). However, historically, dealing with
uncertainty has been a short-coming of many assessments.
Cumulative risk assessments present new challenges for
uncertainty analysis. For example, assessing cumulative risks will
involve combining data of varying quality. Perhaps more
C-4
-------
DRAFT August 9, 2001
important, assessing cumulative risks will involve the use of "soft"
assumptions. These are assumptions which may have a high
degree of uncertainty that is difficult (or not possible) to quantify.
Comment on whether the Framework adequately characterizes the
importance of uncertainty analysis in cumulative risk assessment.
What additional discussions of uncertainty should be included in
the Framework (and in what sections of the document)?
Vulnerability [Framework Section 3.2.1.6]
As applied to cumulative risk assessment, it is useful to think of
four components to vulnerability: the susceptibility or sensitivity
of the human or ecological receptors; the differential exposures of
the receptors; the differential preparedness of the receptor to
withstand the insult from exposure; and the differential ability to
recover from the effects. The issue for cumulative risk assessment
is how to consider these aspects of vulnerability and their potential
impacts on risk. Comment on the discussion of vulnerability in the
draft Framework. Has the state of the science been captured in
this discussion? How can the discussion of this issue be
improved?
Combining Chemical and non-Chemical Stressors [Framework Sections
3.2.1.5 and 3.2.2.5]
Viewing cumulative risk assessment as an evaluation of the
accumulation of stressors presents many challenges. These may be
seen when attempting to combine, in some meaningful way, the
risks from multiple chemicals that may act as synergistic,
antagonistic, or additive doses leading to a single effect. The
situation is exacerbated when non-chemical stressors (e.g.,
radiation, biological agents, and psychological stress) are
considered. Comment on the Frameworks discussion concerning
the combining of disparate environmental stressors. In
commenting, consider the state of the science with respect to
understanding the effects of different stressors acting together
(e.g., chemical exposure and viral infection). What can be added
to the Framework to adequately convey the state of the science in
this area?
Combining Different Types of Risk [Framework Section 4.1.3]
C-5
-------
DRAFT August 9, 2001
Conveying the combined risks from multiple chemical and non-
chemical stressors, in a meaningful way, is the ultimate challenge
for cumulative risk assessment. Experience in this area is
extremely limited. Indices, common metrics (e.g., Disability
Adjusted Life Years - DALYs) and graphical (e.g., GIS)
approaches have been explored but much methods development
work remains to be completed. Cumulative risk assessment can be
a valuable part of the decision making process, but only if the
results are conveyed in a meaningful way. Comment on the
Framework's discussion concerning the combining of disparate
measures of risk. Do the example approaches discussed in this
section capture the state of the science in the area? In particular,
consider the role of valuation (i.e., the assignment of societal
values to disparate health outcomes) implicit in some of the
approaches. Suggest changes or additions that may improve this
discussion.
Limits on the Information Provided by Cumulative Risk Assessment
[Framework Section 2.3]
There are many factors that place limits on what can be learned
from a cumulative risk assessment. These include: how the
problem is defined during the Planning and Problem Formulation
phase, what data are available for the assessment, what models are
available for the assessment, and what resources are available for
the assessment (e.g., expertise, time, financial support, etc.). The
adequacy of existing data may be a key factor limiting the results
of a cumulative risk assessment. One outcome of the assessment
may be the identification and prioritization of data gaps and
research needs. This also applies to the availability of models for
the assessment. Comment on the Frameworks attempt to
characterize the limits that may be encountered when using
cumulative risk assessment in the decision making process. What
can be added to improve this discussion?
Distinguishing between Cumulative Risk and Cumulative Impacts
{Framework Sections 4.4 and 4.5]
The Framework attempts to clarify how cumulative risk
assessment relates to community assessments and decisionmaking.
In particular, the Framework emphasizes that cumulative risk
assessment is only a portion of the information needed for
decisions. Other factors such as employment and quality of life
C-6
-------
DRAFT August 9, 2001
are considered for many decisions. When these other factors are
brought into the assessment, the approach is often called
cumulative impact assessment. Comment on how well the
framework distinguishes between cumulative risks and cumulative
impacts. Are the two concepts well characterized? How can the
discussion be improved to clarify the role of cumulative risk
assessment in the assessment of cumulative impacts?
C-7
-------
APPENDIX D - AGENDA
-------
United States
Environmental Protection Agency
Risk Assessment Forum
Peer Consultation Workshop on Technical Issues
Associated with Developing a Framework for
Cumulative Risk Assessment
Marriott Crystal Gateway Hotel
Arlington, VA
August 22-23, 2001
Agenda
Workshop Chair: Ken Sexton,
University of Minnesota
WEDNESDAY, August 22, 2001
8:3 0AM Registration
9:OOAM Welcome & Introductions DavidBottimore, Versar, Inc.,
9:15AM Background Bill Wood, Risk Assessment Forum (RAF),
U.S. Environmental Protection Agency (U.S. EPA),
9:30AM Presentation of the Draft Framework for Cumulative
Risk Assessment Michael Callahan, Region VI, U.S. EPA,
10:OOAM Break
10:1 SAM Presentation of the Workshop Discussion Topics
Ken Sexton, Workshop Chair
11:00AM Initial Discussion of Discussion Topics
12:OOPM Lunch
D-l
-------
w
EDNESDAY, August 22, 2001 (continued)
1:15PM Discussion Session I - Over-arching Issues
Comment on whether the Framework adequately captures, describes, and
reasonably organizes the key issues for cumulative risk assessment.
What additional issues, if any, should be covered?
Comment on whether some of the topics presented in the draft Framework
are too generic for the intended purpose.
Comment on whether the Framework adequately conveys the state of the
science with respect to cumulative risk assessment and the areas that are in
need of further research and development.
4:00PM Observer Comments
5:OOPM Wrap-Up and Adjourn
D-2
-------
THURSDAY, August 23, 2001
8:30AM Discussion Session II - The Cumulative Risk Assessment Process
Stakeholder Involvement Throughout the Process
Using the Results of Cumulative Risk Assessment
Cumulative Risk Assessment in the Broader Context of Environmental
Analysis and Decision Making
10:OOAM Observer Comments
10:30AM Break
11:00AM Discussion Session III - Technical Issues
Approaches to Cumulative Risk Assessment
Using Biomarkers of Exposure or Effect
Uncertainty in Cumulative Risk Assessments
Vulnerability
Combining Chemical and Non-Chemical Stressors
Combining Different types of Risk
Limits on the Information Provided by Cumulative Risk Assessment
Distinguishing Between Cumulative Risks and Cumulative Impacts
12:OOPM Lunch
1 :OOPM Discussion Session III - Continued
3:OOPM Observer Comments
3:30PM Break
3:45PM Wrap-Up and Next Steps
5:OOPM Adjourn
D-3
-------
APPENDIX E - PRESENTER OVERHEADS
-------
Overheads From David Bottimore's Presentation
E-l
-------
Peer Consultation Workshop
on the
Draft Framework for Cumulative Risk Assessment
August 22-23, 2001
w
to
David Bottimore
Versar, Inc.
Crystal Gateway Marriott
1700 Jefferson Davis Highway
Arlington, VA 22202
-------
Overview of Peer Consultation Workshop
Review of Agenda
Introduction of Participants
EPA Presentations
Chair - Discussion Topics and Groundrules
Observer Comments
Post Meeting Activities - Workshop Report
-------
Overheads From Bill Wood's Presentation
w
-------
Risk Assessment Forum
Framework for Cumulative Risk Assessment
V
William P. Wood
Executive Director
Risk Assessment Forum Staff
Ci
jg FM Off ce of Rtsiddianc Dev
-------
Risk Assessment Forum
Risk Assessment Forum's Mission:
To promote consensus on risk assessment issues
and to ensure that this consensus is incorporated
into appropriate Agency risk assessment
guidance.
-------
Risk Assessment Forum
rn
oo
Characteristics of Forum Projects:
Selected because they present controversial or
precedent setting scientific or science policy
questions for the Agency
Intended to guide the Agency as a whole rather
than any specific program
Designed to be regulation neutral
Shaped and managed by the scientists in the
Forum
-------
Risk Assessment Forum
m
Forum Products Provide Guidance to EPA Risk Assessors
Agency Guidelines (e.g., Guidelines for Ecological Risk
Assessment)
Guidance Documents (e.g., Supplementary Guidance for
Conducting Health Risk Assessment of Chemical Mixtures)
Technical Papers (e.g., Special Report on Environmental
Endocrine Disruption: An Effects Assessment and Analysis)
Visit the Forum Web Page at:
www.epa.gov/ncea/raf
-------
Risk Assessment Forum
m
i
O
Cumulative Risk Assessment Guidance
Development - Phase I
Science Policy Council (SPC)
- Guidance on Cumulative Risk Assessment.
Part 1. Planning and Scoping (July 1997)
SPC Requests that the Forum Begin Developing
Cumulative Risk Assessment Guidance (FY'99)
Science Advisory Board Consultations (FY'OO and FY'Ol)
-------
Risk Assessment Forum
Cumulative Risk Assessment Guidance
Development - Phase I cont'd
m
Meetings with Federal and State Government Scientists
(May 2001)
Public Peer Consultation Workshop (August 2001)
Science Advisory Board Review (Fall 2001)
Goal: Complete the Framework for Cumulative Risk
Assessment in FY'02.
-------
Risk Assessment Forum
Cumulative Risk Assessment Guidance
Development - Phase II
rn
K)
Identify/Develop Illustrative Cumulative Risk
Assessment Case Studies
Develop Technical Issue Papers on Selected Cumulative
Risk Topics
Develop Proposed Guidelines for Cumulative Risk
Assessment (FY'04).
-------
Risk Assessment Forum
rn
UJ
Cumulative Risk Assessment Technical Panel
Michael Callahan, EPA Region VI, Chair
Office of Science Policy
Office of Pesticide Programs
Office of Pollution Prevention and Toxics
Office of Environmental Justice
Office of Emergency and Remedial Response
Chemical Emergency Preparedness and Prevention Office
Office of Solid Waste
Office of Air Quality Planning and Standards
National Center for Environmental Assessment
Regions V, VI, and IX
-------
Risk Assessment Forum
tn
The Framework
Building from ongoing cumulative risk assessment
experience, the Framework is intended to capture the
basic elements of the cumulative risk assessment
process. The Framework:
1) should provide a flexible structure for cumulative risk
assessment issues (capable of evolving with
experience);
2) should define key terms, concepts, and basic
principles to promote a common language on
cumulative risk assessment;
-------
Risk Assessment Forum
The Framework cont'd
3) will serve as a basis for the development of
cumulative risk assessment guidelines;
i
4) will not provide substantive technical guidance.
-------
Overheads From Mike Callahan's Presentation
w
—
o\
-------
Framework for
Cumulative Risk
Assessment
EPA Risk Assessment Forum
Technical Panel
on Cumulative Risk Assessment
-------
-------
Cumulative Risk Assessment
"Traditional" Risk Assessment:
- Where we've been
Cumulative Risk Assessment (CRA)
- Why change?
Framework: What is CRA?
Guidelines: How do we do CRA?
What is today's meeting about?
-------
-------
-------
Framework vs. Guidelines
Framework: General description of the topic.
An information document laying out scope
of the subject and how various parts fit
together. (This document)
Guidelines: Description of how it's done,
including boundaries (e.g., limits of "good
science") not to be exceeded. (Several years
away)
-------
Today's meeting
Report current progress/status on the
Framework for Cumulative Risk
Assessment
Report on issues we've encountered
Listen to your thoughts on the subject
-------
Types of Issues
Process issues: Extent of public
participation, organization of
Framework, etc.
Technical/scientific issues: Feasibility of
certain components, etc.
Policy issues: Requirements, etc. (we
will not discuss these today)
-------
-------
Process issues
Working Definition
Cumulative risk assessment: The
examination of the accumulation over time
(across sources, across routes, etc.) of
stressors or exposures that can cause
adverse effects, and then the integration of
the effects these stressors or exposures
cause into an estimate and characterization of
the risk caused to the individual or population
by the stressors acting together.
-------
Goal of Cumulative RA
Using the commonly accepted definition
of risk as "probability of harm", the goal
of a cumulative risk assessment is:
-To address and hopefully answer
questions related to the probability of
harm, to human health or the
environment, from multiple stressors
acting together.
-------
When do we do a CRA?
Cumulative risk assessment is a tool
It is not appropriate for every task
Cumulative risk assessments will be
most useful in situations where
questions need to be addressed
concerning the impacts of multiple
stressors acting together
Currently, there are methods limitations
-------
Process issues
Organization of Report
Introduction
Problem Formulation Phase
Analysis Phase
Interpretation Phase
Glossary
References
-------
Process issues
Planning/Scoping,
Problem Formulation
Public participation description in
Chapter 2 reflects recent Agency
trend.
Conceptual model and analysis plan
Discussion of possible outcomes
-------
Process issues
Use of CRA
How does CRA fit into the decision
making process?
How does CRA fit into broader context
of various legislation, including NEPA?
What can a CRA - and what can't a
CRA - tell us?
-------
-------
Technical issues
Stressors Acting Together
Combination toxicology/Combining risk
Risk factor approach
Biomarkers or biomonitoring
QALYs, DALYs, LLEs and other
-------
Technical issues
Combining different risks
Can different types of risk be
combined?
Additivity vs. independence
Interactions
Common metric approach
Index approach
-------
Technical issues
Vulnerability
Susceptibility/Sensitivity
Differential exposure
Differential preparedness
Differential ability to recover
Question: How do these factors change
risk?
-------
Technical issues
Uncertainty
Few good examples of uncertainty
analysis for Cumulative Risk
Assessments
New GIS-based technology poses new
challenges in uncertainty analysis
What type of analysis would be useful to
a decision-maker?
-------
-------
Current Activity
EPA is moving toward cumulative risk
approaches in certain situations
Methods are not completely developed,
but some parts of cumulative risk
assessment methodology exist now
Guidelines are perhaps 4-6 years away
-------
Schedule
Currently in technical discussion mode
Finish SAB draft by October 1, 2001
Peer review (incl. SAB) Fall, 2001
Framework final early 2002
Science Policy Council to address
policy issues
Case studies developed 2001-2002?
Guidelines development starts 2002?
-------
Overheads From Ken Sexton's Presentation
E-40
-------
W
Peer Consultation Workshop
on the
Draft Framework for Cumulative Risk
Assessment
Ken Sexton - Chair
-------
Chair's Opening Remarks
w
-k
to
Peer "Consultation" - Obtaining input on technical issues earlier in the process
through a more informal dialogue with experts with diverse perspectives
Goals for Meeting - Provide input to EPA on technical issues related to the
Framework
Discussion Topics - 3 sets (1) overarching issues, (2) process issues, and (3)
technical issues
Ground Rules - process issues, do's and don'ts
Post Meeting Activities - Workshop report that summarizes discussion and
comments on Framework
-------
Discussion Session I - Over-arching Issues
• Comment on whether the Framework adequately captures, describes, and reasonably
organizes the key issues for cumulative risk assessment.
• What additional issues, if any, should be covered?
• Comment on whether some of the topics presented in the draft Framework are too
generic for the intended purpose.
• Comment on whether the Framework adequately conveys the state of the science with
respect to cumulative risk assessment and the areas that are in need of further research
and development.
Discussion Session II - The Cumulative Risk Assessment Process
w • Stakeholder Involvement Throughout the Process
£ • Using the Results of Cumulative Risk Assessment
• Cumulative Risk Assessment in the Broader Context of Environmental Analysis and
Decision Making
Discussion Session III - Technical Issues
Approaches to Cumulative Risk Assessment
Using Biomarkers of Exposure or Effect
Uncertainty in Cumulative Risk Assessments
Vulnerability
Combining Chemical and Non-Chemical Stressors
Combining Different types of Risk
Limits on the Information Provided by Cumulative Risk Assessment
Distinguishing Between Cumulative Risks and Cumulative Impacts
-------
Discussion Session I - Over-arching Issues
Comment on whether the Framework adequately captures, describes, and reasonably organizes the
key issues for cumulative risk assessment? Include in this discussion whether key terms have been
identified and clearly defined.
What additional issues, if any, should be covered?
Keeping in mind that the Framework is intended to address issues that are specific or unique to
cumulative risk assessment, comment on whether some of the topics presented in the draft
w Framework are too generic for the intended purpose?
rv
j\
The Framework attempts to identify where methods and data are currently lacking or are
underdeveloped for application to cumulative risk assessment. Based on a suggestion during the
meeting with other Federal scientists, research oriented issues have been highlighted in an
appendix to the Framework. Comment on whether the Framework adequately conveys the state of
the science with respect to cumulative risk assessment and the areas that are in need of further
research and development.
-------
Discussion Session II - The Cumulative Risk Assessment Process
Stakeholder Involvement Throughout the Process
Using the Results of Cumulative Risk Assessment
Cumulative Risk Assessment in the Broader Context of Environmental
Analysis and Decision Making
w
-------
Discussion Session II - The Cumulative Risk Assessment Process
Stakeholder Involvement Throughout the Cumulative Risk Assessment
Process [Framework Section 2; especially 2.1.3 and 2.3]
Stakeholder involvement is recognized as an important aspect of
cumulative risk assessment. The draft Framework highlights this
involvement in the Planning and Problem Formulation Phase of the
assessment. Examples of interested and affected parties are provided
w and the importance of defining roles and responsibilities is discussed.
& In addition, the importance of discussing the possible outcomes of the
assessment with stakeholders is emphasized. What suggestions do
you have for engaging stakeholders early? What kind of
background or ground rules would be helpful? What roles and
responsibilities would be most appropriate for communities and
interested parties?
-------
Discussion Session II - The Cumulative Risk Assessment
Process
Using the Results of Cumulative Risk Assessment [Framework Section 4.5]
The intended use of the cumulative risk assessment should be
considered at each step of the process, from Planning and Problem
Formulation through completion of the assessment. It is important to
recognize that the cumulative risk assessment provides important
information but is not the only contributor to the decision making
F process. How the cumulative risk assessment fits into the
' decision making process should be addressed when discussing
possible outcomes with stakeholders during the planning phase.
What kind of information could stakeholders contribute to
cumulative risk assessments? Under what circumstances do you
feel stakeholder participation is most effective?
-------
Discussion Session II - The Cumulative Risk Assessment Process
Cumulative Risk Assessment in the Broader Context of Environmental Analysis and
Decision Making [Framework Sections 1.4 and 1.5 and Section 4; especially 4.4.2
and 4.5]
An issue related to using the results of cumulative risk assessment is
understanding how these assessments fit, both conceptually and practicably,
into broader contexts. For example, cumulative risk assessment may be a
component of assessing cumulative effects under the National Environmental
Policy Act (NEPA). Further, these assessment may support Community-Based
£ Environmental Protection efforts. What are the most appropriate situations
for conducting cumulative risk assessments?
oo
-------
VO
Discussion Session III - Technical Issues
• Approaches to Cumulative Risk Assessment
• Using Biomarkers of Exposure or Effect
• Uncertainty in Cumulative Risk Assessments
• Vulnerability
• Combining Chemical and Non-Chemical Stressors
• Combining Different types of Risk
• Limits on the Information Provided by Cumulative Risk Assessment
• Distinguishing Between Cumulative Risks and Cumulative Impacts
-------
Discussion Session III - Technical Issues
Approaches to Cumulative Risk Assessment [Framework Section 3,
especially introduction]
There are a number of ways to approach cumulative risk assessment,
either starting with the NRC paradigm or using a different approach.
Each approach may present its own challenges in methods, data, and
analysis. Four example approaches are discussed in the draft
Framework for Cumulative Risk Assessment combining toxicity before
g calculating risk, calculating risk factors, using biomarkers and
biomonitoring, and calculating other types of probability statements.
Are important approaches missing from this list? Comment on
whether there is a better way to organize this discussion on
approaches to cumulative risk assessment.
-------
Discussion Session III - Technical Issues
Using Biomarkers of Exposure or Effect [Framework Section 3.1.3]
Among the approaches to cumulative risk assessment discussed in
the draft Framework, use of biomarkers and biomonitoring information
shows promise for providing an integrated measure of where an
individual falls on the continuum between exposure and effect.
However, existing methods are inadequate for assessing complex
situations involving a large number of stressors. Further, experience
using biomarkers and related information in decision-making is limited.
- Comment on the discussion of biomarkers of exposure and
- effects? Given the current state of the science, discuss the utility
of this approach in the near term and in the future. Does the
Framework include the appropriate level of discussion of this
topic? How can the discussion be improved (consider, for
example, whether the Framework discussion adequately
characterizes how biomarkers can be used with more traditional
risk estimates)?
-------
to
Discussion Session III - Technical Issues
Uncertainty in Cumulative Risk Assessments [Framework Section 4.3]
Uncertainty analysis is an important aspect of risk assessment (and
policy analysis in general). However, historically, dealing with
uncertainty has been a short-coming of many assessments.
Cumulative risk assessments present new challenges for uncertainty
analysis. For example, assessing cumulative risks will involve
combining data of varying quality. Perhaps more important, assessing
cumulative risks will involve the use of "soft" assumptions. These are
assumptions which may have a high degree of uncertainty that is
difficult (or not possible) to quantify. Comment on whether the
Framework adequately characterizes the importance of
uncertainty analysis in cumulative risk assessment. What
additional discussions of uncertainty should be included in the
Framework (and in what sections of the document)?
-------
Discussion Session III - Technical Issues
Vulnerability [Framework Section 3.2.1.6]
As applied to cumulative risk assessment, it is useful to think of four
components to vulnerability: the susceptibility or sensitivity of the
human or ecological receptors; the differential exposures of the
receptors; the differential preparedness of the receptor to withstand
the insult from exposure; and the differential ability to recover from the
effects. The issue for cumulative risk assessment is how to consider
these aspects of vulnerability and their potential impacts on risk.
- Comment on the discussion of vulnerability in the draft
" Framework. Has the state of the science been captured in this
discussion? How can the discussion of this issue be improved?
-------
Discussion Session III - Technical Issues
Combining Chemical and non-Chemical Stressors [Framework Sections
3.2.1.5 and 3.2.2.5]
Viewing cumulative risk assessment as an evaluation of the
accumulation of stressors presents many challenges. These may be
seen when attempting to combine, in some meaningful way, the risks
from multiple chemicals that may act as synergistic, antagonistic, or
additive doses leading to a single effect. The situation is exacerbated
when non-chemical stressors (e.g., radiation, biological agents, and
psychological stress) are considered. Comment on the
Framework's discussion concerning the combining of disparate
environmental stressors. In commenting, consider the state of
the science with respect to understanding the effects of different
stressors acting together (e.g., chemical exposure and viral
infection). What can be added to the Framework to adequately
convey the state of the science in this area?
-------
Discussion Session III - Technical Issues
Combining Different Types of Risk [Framework Section 4.1.3]
Conveying the combined risks from multiple chemical and non-
chemical stressors, in a meaningful way, is the ultimate challenge for
cumulative risk assessment. Experience in this area is extremely
limited. Indices, common metrics (e.g., Disability Adjusted Life Years -
DALYs) and graphical (e.g., CIS) approaches have been explored but
much methods development work remains to be completed.
Cumulative risk assessment can be a valuable part of the decision
making process, but only if the results are conveyed in a meaningful
way. Comment on the Framework's discussion concerning the
combining of disparate measures of risk. Do the example
approaches discussed in this section capture the state of the
science in the area? In particular, consider the role of valuation
(i.e., the assignment of societal values to disparate health
outcomes) implicit in some of the approaches. Suggest changes
or additions that may improve this discussion.
-------
Discussion Session III - Technical Issues
Limits on the Information Provided by Cumulative Risk Assessment
[Framework Section 2.3]
There are many factors that place limits on what can be learned from a
cumulative risk assessment. These include: how the problem is
defined during the Planning and Problem Formulation phase, what
data are available for the assessment, what models are available for
the assessment, and what resources are available for the assessment
(e.g., expertise, time, financial support, etc.). The adequacy of
existing data may be a key factor limiting the results of a cumulative
risk assessment. One outcome of the assessment may be the
identification and prioritization of data gaps and research needs. This
also applies to the availability of models for the assessment.
Comment on the Framework's attempt to characterize the limits
that may be encountered when using cumulative risk assessment
in the decision making process. What can be added to improve
this discussion?
-------
Discussion Session III - Technical Issues
Distinguishing between Cumulative Risk and Cumulative Impacts
[Framework Sections 4.4 and 4.5]
The Framework attempts to clarify how cumulative risk assessment relates to
community assessments and decisionmaking. In particular, the Framework
emphasizes that cumulative risk assessment is only a portion of the information
needed for decisions. Other factors such as employment and quality of life are
considered for many decisions. When these other factors are brought into the
assessment, the approach is often called cumulative impact assessment.
- Comment on how well the framework distinguishes between cumulative
^ risks and cumulative impacts. Are the two concepts well characterized?
How can the discussion be improved to clarify the role of cumulative risk
assessment in the assessment of cumulative impacts?
-------
APPENDIX F - FRAMEWORK CONCEPTUAL DIAGRAMS
(FROM DOUGLAS CRAWFORD BROWN)
-------
1. Historical Introduction: How did we get to this point of
thinking about Cumulative Risk Assessment? What are
the issues that drove the need for such a tool?
2. What is Cumulative Risk Assessment? What are some
examples of its application? How does it differ from
previous assessments in these examples? How does it
address the limitations in traditional regulatory risk
assessments that drove the issue?
3. What are the steps of a Cumulative Risk Assessment?
What is the process of discourse conducted in these
steps? What is the content of discourse conducted in
these steps? What are the principles that ensure the
process and content are of the necessary quality?
Problem
Formulation
Data&
Analysis
Synthesis &
Interpretation
F-l
-------
Problem
Formulation
1. Purpose: The goal of Problem Formulation is to state
concisely and clearly the environmental situation under
consideration, the values at risk from this situation, the
parties to be included in the discourse, the categories of
answers that must be generated, the risk characteristics to
be included in these answers, and the role of these
answers in decisions.
2. The explicit steps of Problem Formulation are (the first
step has a hyperlink to a representative second level of
the framework):
Define the
geographic
region
Identify the
values
at risk
i
Define the
characteristics
of risk
^
>„
Produce
conceptual
model
1
Define the
decision
process
Identify
the
stakeholders
Produce
rnnrpntual I
w w i i \j\^ ky i u t-i i •
model
v
Engage
the
stakeholders
i
Define the
assessment
problem
F-2
-------
Define the
geographic
region
Purpose: To be clear as to the geographic boundaries of the
environmental situation and affected community, including all
sources that might impact that community even if these are
not contained in the boundaries where risks and impacts are
being assessed.
Questions to be addressed:
• What are the criteria used to identify the geographic
boundaries of a problem?
• Using these criteria, what are the boundaries for this
community?
Example answers:
• Within 50 km of the industrial complex.
• Orange County, NC and parts of Wake County, NC.
F-3
-------
Data&
Analysis
1. Purpose: The goal of Data & Analysis is to assemble
necessary data, assess the quality of those data, estimate
exposures and risks to the population in the community,
and summarize uncertainty and variability for these
estimates.
2. The explicit steps of Data & Analysis are (the first step has
a hyperlink to a representative second level of the
framework):
Identify
community
stressors
Identify
exposure-
response
1
Characterize
risk
Identify
receptor
populations
Calculate
total
exposure
>„
^
Ider
expo
pathn
itify
sure
ways
f
Identify
background
levels
Characterize
uncertainty
and
>„
Characterize
variability and
equity
F-4
-------
Identify
community
stressors
Purpose: Identify the kinds and sources of stressors in this
community.
Questions to be addressed:
• What is meant by a stressor?
• Which stressors are relevant for this kind of assessment?
• Using these criteria, what are the relevant stressors for this
community?
Example answers:
• Stressors are physical, biological, chemical and social
factors that increase the incidence of disease.
• All factors that increase the incidence of childhood
leukemia.
• Viruses, benzene, radiation.
F-5
-------
Synthesis &
Interpretation
1. Purpose: The goal of Synthesis & Interpretation is to
summarize the results in ways that inform the original
decision framework, are understandable to the
stakeholders, and allow identification of both risks and their
causes.
2. The explicit steps of Synthesis & Interpretation are (the first
step has a hyperlink to a representative second level of the
framework):
Assess
acceptability of
risk variability
Identify
vulnerable
subpopulation
Identify
mitigation
resource
Identify
mitigation
strategies
Summarize
confidence in
risk estimates
Identify
major causes
of risk
So
FIX
it!
F-6
-------
Assess
acceptability
Purpose: To determine whether the variation in risk is
acceptable in terms of the criteria of equity established in
Problem Formulation.
Questions to be addressed:
• Is there an acceptable level of variability?
• Is the upper bound of risk acceptable ?
Example answers:
• No, there is wide variation across the community.
• Yes, so while there is not equity, no individual rights are
being violated.
F-7
-------
APPENDIX G - POST-MEETING WRITTEN COMMENTS
-------
Elizabeth Boa
Post-Meeting Written Comments From Elizabeth Boa
G-l
-------
Elizabeth Boa
The Risk Assessment Forum
Draft Framework for Cumulative Risk Assessment
Peer Consultation Comments of Elizabeth Boa
Summary:
I support the development of a cumulative risk assessment framework, as there is a need
in to coordinate and direct the multiple, highly variable efforts within EPA. However this
document is so general that I don't believe that inconsistency in conducting cumulative
risk assessments can be avoided among and within program offices. It is unclear how the
framework could be applied in any program. It appears to be a summary of several other
EPA documents/sources and hits on concepts only superficially. There is very little
science to underpin the broad cumulative risk assessment approach of multiple unrelated
stressors and thus the framework is held together only by theory.
At a minimum, before the framework goes any further, e.g. to SAB review, it needs to
add case studies that illustrate application of the suggested processes. It is through
development and working through such case studies that the applicability of the
procedures in the framework can be assessed.
Currently, different EPA program offices take different approaches to cumulative risk
assessment. The variety of approaches that have proliferated, and the diverse terminology
accompanying these approaches, leads to confusion among the regulated community and risk
assessors about cumulative risk assessment.
The Framework needs to differentiate among these various approaches, especially as applied to
risk assessments of chemicals, to prevent the misapplications of the terminology and the
approach both by regulators and non-governmental organizations and other risk experts.
Cumulative risk assessments need to follow a clear, scientifically sound procedure to evaluate
cumulative risks to be useful in policy making or regulating. I suggest that the Framework
improve its organization by more clearly focusing on the two overall categories: approaches to
cumulative risks with a chemical focus which may or may not be restricted to those with
common mechanisms of action; and those including stressors beyond chemicals. My preliminary
review found that there are four general approaches to cumulative risk assessments for chemical
stressors and two overall approaches to cumulative risks of multiple stressors. Some approaches
are very clear about the adverse endpoint or effect, naming a specific common endpoint, while
other approaches are not. The cumulative risk assessment may consider a single common
adverse effect or it may consider broadly defined effects such as human health or ecological
effects or regional cancer risk. Each may be useful for different purposes such as screening to
compare regional variations in stressors (as opposed to risks) or as a basis for regulating
chemicals which cause a common adverse effect by a common mechanism.
As an illustration of the confusion around "cumulative risk," I have listed the
many terms used to express the concept used within and outside the EPA:
G-2
-------
Elizabeth Boa
Aggregation
Body burden
Comprehensive chemical exposure
Comprehensive risk
Cumulative effects
Cumulative environmental hazards
Cumulative hazard
Cumulative pollution
Holistic risk
Incremental impact
Integrated risk
Mixture
Multipathway exposure
Multiple chemical
Multiple exposures
Multiple risk
Overall regional risk
Risk "in context"
Synergistic effects
Total load
Total risk
As variations of cumulative risk assessment proliferate, it will become increasingly important to
distinguish among the variations in meaning so analytical approaches can be distinguished. This
is a role that the Framework can embrace. For example, the term "aggregate risk" has been used
synonymously with the term "cumulative." However, as defined by the EPA Office of Pesticide
Programs (OPP), aggregate is the term for single chemical, multiple pathway exposures causing
a common effect. In the context of pesticide risk assessment at least, aggregate should not be
used to describe risks from multiple chemicals.
The use of specific, precise terminology is needed to identify various approaches to risk
assessment that are labeled "cumulative," so that a common lexicon for the various forms of
"cumulative" risk assessments can be developed. "Cumulative risk assessment" also needs to be
precisely defined especially as it applies to risk assessments of chemicals to help prevent the
misapplications of the terminology and the approach both by regulators and non-governmental
organizations.
Comment on whether the Framework adequately captures, describes, and
reasonably organizes the key issues for cumulative risk assessment?
I. The document can be improved. There is a need for a better differentiation between
what has been considered cumulative risk as a regulatory term (FQPA) and this much
broader presentation of cumulative risk. Chemical focused cumulative risk can be better
described in this document and kept separate from the other broader "integrated" risk
assessments this document also describes which include more risks than only chemicals.
G-3
-------
Elizabeth Boa
II. This document is also written at different levels: some is very basic and some
is highly technical. Needs to be written with a clear idea of the audience and
target that audience. The target audience should be risk assessors and managers,
so the document should be written at that level.
III. Considering the state of the science, the goal of this document would be better
to identify the areas of research needed. Currently, attempting a cumulative risk
assessment would involve a great deal of uncertainly.
IV. The document needs to focus on what we are trying to accomplish with a
cumulative risk assessment: usually, this is lowering risks through regulations.
This needs to be considered in the problem formulation stage when stressors are
identified.
V. I agree with the working backward, population based approach generally,
however, this has not been how cumulative risk assessments have been conducted
under the FQPA. FQPA works "forward" from identified pesticides. How will
this approach be integrated into the document? A preferable approach is to work
backward from specific, identified health effects rather than from all stressors in a
community to identify and evaluate the stressors that are actually a problem.
Although identifying the cause(s) of a particular effect is not always possible,
focusing on particular health effects and working backward is more economical
and efficient than focusing on all community stressors indiscriminately.
Include in this discussion whether key terms have been identified and
clearly defined.
I. Need new terminology to differentiate between the previously defined
"regulatory cumulative risk" term (that is, the FQPA definition, which is highly
specific and very technical) and the very much broader cumulative risk approach
that is described here. To avoid confusion between the regulatory definition and
the broader one, perhaps assigning the term "integrated risk" would be more
accurate and less confusing. The use of specific, precise terminology is needed to
identify various approaches to risk assessment that are labeled "cumulative," so
that a common lexicon for the various forms of "cumulative" risk assessments
can be developed.
What additional issues, if any, should be covered?
I. The document should focus on the assessments conducted or underway in EPA.
There are assessments by Office of Pesticide Programs, the Office of Water,
(disinfection byproducts), the National Air Toxics Assessment (NATA), which
are chemical only and which may provide lessons learned and help define general
principles for chemical cumulative risk assessment. Those that have been
conducted on broader stressors have included ecological risk assessments which
could be described here. Also, the framework should include examples of
G-4
-------
Elizabeth Boa
community risk assessments: what have they shown in terms of the usefulness of
their results, the techniques used, can they stand up to technical review in terms
of methodology and results. Many of the community risk assessments used for
environmental justice are aggregate exposure assessments as opposed to risk
assessments. Examples of this type of assessment need to be described in the
document, and compared with the approaches suggested in this document.
Likewise, the framework should describe how multi-stressor ecological risk
assessments are conducted and how this population based paradigm can be
applied to human populations. This document does not include much on
ecological cumulative risk assessment, and it is therefore unclear whether this
framework is intended for human health risk only.
II. The Framework should be built from these experiences to develop principles
for cumulative risk assessments, and describe lessons learned and best practices.
III. The document could be more useful to those doing cumulative risk
assessment and facing issues such as model selection. The OPP is conducting a
cumulative risk assessment on organophosphates (OPs); one of the biggest
difficulties being encountered in this attempt is the identification of a dose-
response model that is consistent with the data and the mode of action. The
Framework could provide input to OPP and other offices facing the challenge of
selecting models or data for cumulative risk assessments.
IV. The framework should also include information on how cost-benefit analysis
can be included to determine alternative risk reduction strategies after risk
characterization is completed.
V. The Framework needs to give more details on selecting dose versus response
additivity when combining chemicals in a cumulative assessment. The
framework advises selecting dose additivity as a default when for the most part
response additivity is the more appropriate method. This issue is covered in the
Mixtures guidance and comes up in the FQPA application of cumulative risk
assessment..
Keeping in mind that the Framework is intended to address issues that are
specific or unique to cumulative risk assessment, comment on whether
some of the topics presented in the draft Framework are too generic for
the intended purpose?
I. The document could spend less time up front on the assessment process used
for single chemicals and describe how the process applies to cumulative risk
assessment specifically.
Comment on whether the Framework adequately conveys the state of the
science with respect to cumulative risk assessment and the areas that are
in need of further research and development.
G-5
-------
Elizabeth Boa
I. In places the document mentions that research and data are needed. However
overall it does not capture that we don't have any experience with such
assessments, that most data and methods are not available to conduct these. Any
attempts to do so have been experimental and need thorough evaluation to
determine how technically accurate and useful they are.
II. This document is somewhat inconsistent in describing the combination of
stressors as opposed to risks, particularly when non-chemical stressors are part of
the assessment, without considering the interaction of the stressors which is the
key part of cumulative risk assessments. Part of the process of cumulative risk
assessment is characterizing the risk posed by the stressors ACTING
TOGETHER. The chemical and non-chemical stressors may not act together, they
may be independent, or be additive, synergistic, or antagonistic. The framework
needs to discuss how the interaction of the non-chemical stressors will be
accounted for in the assessment.
To assist with this discussion, the following list includes a description of
some of the key technical issues encountered by the Risk Assessment Forum
Technical Panel during development of the draft Framework for Cumulative Risk
Assessment. This is not a comprehensive list of cumulative risk assessment issues
but could serve as a starting point for the review discussions. Where possible, the
relevant sections of the Framework document have been referenced for each
issue.
Process Issues
Stakeholder Involvement Throughout the
Cumulative Risk Assessment Process [Framework Section
2; especially 2.1.3 and 2.3]
What suggestions do you have for engaging
stakeholders early? What kind of background or
ground rules would be helpful? What roles and
responsibilities would be most appropriate for
communities and interested parties?
I. Communities, interested parties, and stakeholders all have a role in providing
input into cumulative risk assessments. As the Agency increasingly relies on
science as the foundation of its decisions, the question of how best to make sure
science is sufficiently represented in stakeholder processes has emerged as a vital
issue. Recognizing this fact, EPA's Science Advisory Board (SAB) last year
initiated an evaluation of how best to promote "a full and careful consideration of
all available science" in stakeholder-based environmental decisions.
G-6
-------
Elizabeth Boa
I strongly urge EPA to address in the Framework the need for assuring the use of
scientific information, underscoring the fundamental place of science in EPA
decision-making. As noted by the SAB in its Oct. 7, 1999, letter to then-
Administrator Carol Browner, the concept of the general public interest "lies at
the heart of many of our most cherished democratic institutions," and the Agency
would best serve this interest when it carefully considers all available science as
well as addressing stakeholder needs and interests. I second this SAB position.
The Framework should clearly describe the importance of credible scientific
information in Agency deliberations and provide the best guidance possible on
how to ensure that such science plays an appropriate role in stakeholder
involvement processes.
II. Addressing scientific uncertainty, and diligently working to reduce it through
research and the use of all available data, must be a central component of Agency
decision-making. Informing stakeholders as accurately as possible about what the
Agency knows, does not know, and assumes in the face of limited data must be
central to stakeholder involvement.
Using the Results of Cumulative Risk
Assessment [Framework Section 4.5]
What kind of information could stakeholders
contribute to cumulative risk assessments? Under
what circumstances do you feel stakeholder
participation is most effective?
The communities and interested parties would be valuable in highlighting
concerns, values, behaviors as well as constructing a complete picture of risks to
the community. The community could help identify holistically the major health
or ecological effects they are facing.
Cumulative Risk Assessment in the Broader
Context of Environmental Analysis and Decision
Making [Framework Sections 1.4 and 1.5 and
Section 4; especially 4.4.2 and 4.5]
What are the most appropriate situations for
conducting cumulative risk assessments?
I. There are several situations: concern about exposure from a source (or
sources) of multiple chemicals, a decision about where to site a facility, or
concern about a specific health or ecological effect that may be caused by
multiple stressors.
II. The various applications of cumulative risk assessment, which use various
quality information should be used for different purposes: regulatory;
screening for more research; risk ranking; or comparison among regional
G-7
-------
Elizabeth Boa
risks. I agree with document (page 11) That it is very important to apply
CRA in the context of the decision to be made or the problem to be solved.
III. The framework does not give guidance in the problem formulation stage for
putting an analysis in context when it is required by law. For example, the OPP
cumulative risk assessment is justified by the requirement of FQPA. It is not
justified by identification of a particular problem being solved such as a health
problem. The framework needs to give guidance to the Agency on providing a
context or a reason for a cumulative risk assessment being conducted beyond a
law or a regulatory requirement.
Technical Issues
Approaches to Cumulative Risk Assessment [Framework
Section 3, especially introduction]
Are important approaches missing? Comment on
whether there is a better way to organize this
discussion on approaches to cumulative risk
assessment.
I. Clearly delineate in the Framework the types of cumulative risk assessment that are
used now in the agency for chemical risk assessments v. the community based risk
assessments that are discussed in this document. These are different approaches with
different outcomes and for different purposes. These approaches are very different in
how they should be conducted and their ultimate outcome and purpose.
II. For chemical cumulative risk assessment, there are two categories that I see. A. For
cumulative risk assessments of non-carcinogenic chemicals used in regulating or
permitting, such as those conducted by EPA's pesticide office, it is appropriate only to
link risks of chemicals that act by a common mechanism, with parallel dose-response
curves, to cause a well-defined common adverse effect at a common site. This is to
ensure that the regulation of chemicals acting by common mechanism will be aimed at
reducing a particular risk and will not artificially reduce the risk cup by including
inappropriate chemicals. The requirement of common mechanism is consistent with the
EPA mixtures guidance recommendations for dose additivity. Combining multiple
stressors by dose additivity needs an understanding of mechanism of action of each
stressor and therefore the interaction among the stressors. For very few chemicals do
we have this understanding. The document at 32 says that the prediction of how
specific mixtures of toxicants will interact must be based on an understanding of the
mechanisms of such interactions. This is true. The framework doesn't state that we
usually don't have this knowledge. It is very complex. Under FQPA, the risk
assessment is only looking at common mechanism causing common effect to assume
dose additivity and yet it is a very complex determination of interaction.
G-8
-------
Elizabeth Boa
B. For a "screening" cumulative risk assessment used for ranking purposes,
chemicals can be grouped by common target organ, as a surrogate for common
mechanism in the absence of data (that is, with mode/mechanism of action unknown).
But such an assessment must assume the chemicals have parallel dose-response curves
and cause a common adverse effect. The common mechanism requirement is waived in
this case because that information is likely not known in a screen. A "screening level
cumulative risk assessment" attempting to combine multiple risks should be only for
the purpose of, for example, developing an understanding of the various regions and
populations most at risk. Screening level assessments should be used only for
prioritizing further data collection and research. For example, the EPA's National Air
Toxics Assessment (NATA) draft report combines diverse risks rather than common
adverse effects with a common mechanism of action.
III. The framework does not address chemical mixtures issues such as assumption
of response additivity. It does not adequately address the fact that most chemicals
do not interact but act independently, and cause different effects and/or operate
through a different mechanism of action. That is where response additivity should
be used not dose additivity. Examples of how independence would be handled
should be included. The Framework needs to indicate that response additivity
should also be used when assessing the cumulative risk of chemical exposures
below individual chemical's NOAEL unless there is evidence that identical
metabolic pathways detoxify the chemicals.
IV. At page 34, the document says that toxicants may interact by causing different
effects at different receptor sites. I don't understand: What is the basis for that
statement? Based on that information, there is no basis to assume an interaction.
That implies independence of action. That provides no basis for "adding" these
risks together. See III. above.
V. The Framework does not give guidance on choosing an index chemical and
supporting that choice or for selecting among available dose response data sets.
This guidance could help the OPP with its OP cumulative risk assessment; in that
case an index chemical was chosen based on a good data base for that chemical
but without supporting documentation as to why other chemicals were not chosen
as the index chemical. The Framework could also help OPP with selecting the
dose-response data set. Their OP cumulative risk assessment uses surrogate dose-
response data for the index chemical and the other chemicals, rather than the best
available data.
VI. P. 45:1 don't understand the statement that using the Relative Potency Factor
can be applied to all toxic endpoints for the chemical group being assessed when
the similarity is justified on mechanistic grounds. Is similarity based on the
mechanism of action? Or the SAR? If there is an understanding of a mechanism
of action for one endpoint it doesn't mean that the same group of chemicals will
G-9
-------
Elizabeth Boa
all work with the same mechanism of action for all endpoints. This needs to be
determined for each endpoint.
VII. Multi-stressor assessments for ecological risk consider that an array of factors of
chemical, physical, or biological origin may be active, either concurrently or
sequentially, in the deterioration of natural ecosystems (Foran and Ferenc 1999).
Ecological risk is not addressed to any extent in this document.
Using Biomarkers of Exposure or Effect [Framework
Section 3.1.3]
Among the approaches to cumulative risk
assessment discussed in the draft Framework, use
of biomarkers and biomonitoring information
shows promise for providing an integrated measure
of where an individual falls on the continuum
between exposure and effect.
This can only be done if there is a lot more information on mechanism of action
of the chemical and how other stressors are affecting that mechanism.
However, existing methods are inadequate for
assessing complex situations involving a large
number of stressors
OR even a single chemical.
Comment on the discussion of biomarkers of
exposure and effects?
This document needs to discuss the limitations of what realistic use
biomarkers are at this point, and distinguish between biomarkers of
"exposure" and biomarkers of "effects." Exposure will not necessarily lead to
effects. Additionally, biomarkers of effects do not indicate the cause of the
effects.
Given the current state of the science, discuss the
utility of this approach in the near term and in the
future. Does the Framework include the
appropriate level of discussion of this topic? How
can the discussion be improved (consider, for
example, whether the Framework discussion
adequately characterizes how biomarkers can be
used with more traditional risk estimates)?
G-10
-------
Elizabeth Boa
The Framework should give more guidance on when available biomonitoring data
should be used rather than exposure modeling or as a reality check for
monitoring. The OPP cumulative risk assessment of OPs relies on complex
exposure modeling although biomonitoring data are available. The Framework
should address when real data should be used over model outputs.
Uncertainty in Cumulative Risk Assessments [Framework
Section 4.3]
Cumulative risk assessments present new
challenges for uncertainty analysis.
In cumulative risk assessment, there is a great deal more uncertainly introduced, more
than has been dealt with in single chemical assessments. It strikes me that there is
uncertainty to a level never seen before. It is imperative to quantify these uncertainties
so we know how close the result is to actual risk. Uncertainty AND variability need to
be dealt with in the beginning of the risk assessment process and throughout.
Perhaps more important assessing cumulative risks
will involve the use of "soft" assumptions. These
are assumptions which may have a high degree of
uncertainty that is difficult (or not possible) to
quantify.
The Framework should clarify that these "soft" assumptions will introduce a great
deal of uncertainty the more they do not rely on science.
Comment on whether the Framework adequately
characterizes the importance of uncertainty analysis
in cumulative risk assessment. What additional
discussions of uncertainty should be included in the
Framework (and in what sections of the document)?
I. The Framework should discuss in the risk characterization section that a
range of risk (best case, worst case, average case numbers) should be included
to give people an idea of uncertainty.
II. The risk characterization needs to acknowledge the data gaps and the severe
limitations these result in by using defaults. Where there is no information, this
fact has be carried through on the risk characterization for the decision-maker.
III. The document should also include that uncertainty models used should
include "average" and "least conservative" assumptions.
Vulnerability [Framework Section 3.2.1.6]
G-ll
-------
Elizabeth Boa
As applied to cumulative risk assessment it
is useful to think of four components to
vulnerability: the susceptibility or sensitivity of the
human or ecological receptors: the differential
exposures of the receptors: the differential
preparedness of the receptor to withstand the insult
from exposure: and the differential ability to
recover from the effects. The issue for cumulative
risk assessment is how to consider these aspects of
vulnerability and their potential impacts on risk.
Comment on the discussion of vulnerability in the
draft Framework. Has the state of the science been
captured in this discussion? How can the
discussion of this issue be improved?
I. Definitions issue: susceptibility and sensitivity traditionally encompass all the other
aspects that are now labeled "vulnerability." Change "sensitivity and susceptibility" to
"biological differences."
Combining Chemical and non-Chemical Stressors
[Framework Sections 3.2.1.5 and 3.2.2.5]
Combining Different Types of Risk [Framework Section
4.1.3]
Comment on the Framework's discussion
concerning the combining of disparate measures of
risk. Do the example approaches discussed in this
section capture the state of the science in the area?
In particular, consider the role of valuation (i.e.. the
assignment of societal values to disparate health
outcomes) implicit in some of the approaches.
Suggest changes or additions that may improve this
discussion.
1.1 am not sure whether different stressors, which act independently, can even be
added up or what it would mean even if they were. This is very different from
cumulative risk of chemicals. This type of assessment is not truly assessing
"cumulative" risks from a toxicological standpoint and is more appropriately
labeled "number of hazards from multiple stressors." I think a comparative risk
approach would be more useful than a combination approach. Isn't it more
important to figure out the risk drivers than the total risk in a meaningless
numerical score?
G-12
-------
Elizabeth Boa
II. The addition of diverse hazards provides limited information beyond comparison,
for example, of variations among number of regional hazards. The Framework should
include that the assessor needs to consider what such a combination of diverse risks
would mean scientifically — and articulate a justification for assuming additivity of
diverse risks — and from a policy standpoint.
III. For assessments of cumulative risks to human health, combining multiple, unrelated
chemical and non-chemical hazards, the hazards need to be assigned weights in terms of
probability of risk and in terms of the relative adverse effect.
IV. The Framework needs to address how it will deal with combining in a ranking system
data that are high quality with data that are highly uncertain.
V. It is not clear how the Framework would suggest the Agency integrate what may be
incommensurables e.g. cancer and non-cancer effects, health and eco effects,
extrapolated risks based on animal studies and actual risks based on epidemiology.
VI. Caveats need to be made with regard to the estimated hazard numbers, indicating
that adding diverse hazards provides a number that has limited descriptive
capabilities or meaning and gives a possibly misleading picture of the magnitude of
risks to which populations are exposed. Without a clear understanding of the purpose
for acquiring the information from the cumulative assessment, this information can
easily be taken out of context.
Limits on the Information Provided by Cumulative Risk
Assessment [Framework Section 2.3]
Comment on the Framework's attempt to
characterize the limits that may be encountered
when using cumulative risk assessment in the
decision making process. What can be added to
improve this discussion?
I. We should try to identify the data gaps and quantify how big they are. The
framework should include information on what models are out there and their
availability.
II. One thing that could be addressed is that cost benefit analysis should be
instituted as part of the cumulative assessment.
G-13
-------
Post-Meeting Written Comments From Ken Bogen
G-14
-------
Ken Bogen
Comments of K.T. Bogen on the
USEPA Review Draft
Framework for Cumulative Risk Assessment
Comments below are organized with reference to section number of the USEPA Review Draft
Framework for Cumulative Risk Assessment (NCEA-F-1098; 2001).
General: Specific aims of the Framework are not clear
CRA is now fairly established as an acronym for Comparative Risk Assessment. Also, technical
aspects of how to aggregate risks have been addressed in detail in several EPA guidelines as well as
in the NRC 1994 report, Science and Judgment in Risk Assessment, that already are cited in the
Framework, which apparently is neither intended to add to nor comment substantively upon these
existing risk-cumulation methods. Consequently, it may be argued that the proposed "cumulative"
risk assessment Framework is nothing but a generic description of risk assessment that is properly
executed to consider all relevant risks associated with one or more risk sources of concern. That is,
the only new problem being addressed by the word "cumulative" in the Framework appears (but is
nowhere clearly stated) to be simply how best to define a set of risk sources that are (or could?, or
should?) be considered relevant to a specific public-policy decision. SIN this context, some careful
thought might be given to an alternative name for the Framework. The use of the word
"Assessment" (rather than, say, "Analysis and Characterization") suggests a framework intended to
address risk management issues that appear to be beyond the intended scope of the present
Framework, because it currently presents no background or information concerning the array of
different methods often used to make decisions involving uncertain risks, costs and benefits. Thus,
it appears that the intended, and more restricted, focus is a framework for Cumulative Risk Analysis
and Characterization (CRAC) or Risk Assessment for Multiple Endpoints and Sources (RAMES).
A Framework for RAMES might be denoted "FRAMES". Nevertheless, "CRA" is used below to
denote Cumulative Risk Assessment.
Section 1. Introduction
As noted above, this section needs to define the nature and purpose of the Framework more clearly,
particularly in relation to previous/parallel EPA documents such as the SAB report, Toward
Integrated Environmental Decision-Making (EPA-S AB-EC-00-011; 2000). The latter report gives
examples of regulatory mandates that require integrated environmental decisionmaking. The
Framework for Cumulative Risk Assessment likewise should list and discuss how at least an
illustrative set of regulatory and nonregulatory EPA mandates currently and/or potentially could
require and/or benefit from CRA.
This section should spell out specific issues that are newly being addressed in the proposed
Framework, and explain—and illustrate using a few simple, diagrammed examples—how these new
issues differ from problems addressed by previous documents. Using these illustrative examples,
G-15
-------
Ken Bogen
this section should state clearly (in generic terms, if necessary) some examples of why and/or under
what circumstances current applications of existing approaches might be considered deficient, and
thus why a new framework is needed. To the extent that existing approaches might be considered
deficient, why is this so? Do such deficiencies arise from overly narrow statutory mandates, overly
narrow regulatory implementation of existing statutory mandates, inappropriate/inconsistent
implementation of mandated regulatory goals, or lack of appropriate procedures and a
suitable/consistent methodology for undertaking CRA? For example, if the application of current
RfD methods to set MCLs fails to consider background doses in exposed subpopulations, then this
indicates that those methods were simply defective methods to implement appropriate regulatory
toxicology, not that a whole new framework is needed. Is the new framework intended to facilitate
new regulatory and/or scientific goals that involve aggregate risk, or rather to improving the
achievement of previous goals that were not met because aggregate risks were not considered
appropriately?
The text in current Section 4.5, para. 3 line 1 (p. 60, 2nd line from bottom) should also appear in
Section 1, as a specific example of a CRA application and CRA goals.
A key potential application of CRA is for long-term/anticipatory planning, e.g., with regard to siting
new sources of pollution in specific areas. Implications of using this method should be discussed.
E.g., use of CRA implies that, where not already in place (give examples), new procedures need to
be developed for planning for and implementing source-specific caps on the siting of new sources
in regions identified through CRA to be approaching or at a maximum level of acceptable
environmental health stress.
This section should end with a brief (1-paragraph) introduction to, and a corresponding schematic
diagram summarizing the elements/steps that comprise, the proposed framework. Consideration
might be given to renaming section 1 to refer to a substantive topic, e.g., "1. The Need for a
Cumulative Risk Framework".
Section 2. Planning and Problem Formulation
On page 16 (sec. 2.1.2, line 6), the draft now states that "The issue of background exposures should
be discussed and agreements reached (see Appendix E)." This statement is very problematic, insofar
as the "issue" of background risks might reasonably be considered to be a technical issue driving the
"correct" prediction of source-induced risk, particularly for noncancer (e.g., threshold) endpoints,
and as such ought not be subject to political interpretation/agreement. If I am currently
environmentally exposed to X mg/kg-d of a toxic agent (e.g., mercury), if the accepted ADI for the
agent is A mg/kg-d, and if the focus of a CRA is whether or not to accept a new source that imposes
a new exposure to me of Y mg/kg-d, a relevant CRA issue must be whether or not X + Y < A (or to
what extent X + Y < A, considering uncertainties involved), and not whether Y < A, regardless of
how participants in any CRA scoping discussion may feel about this. Accordingly, this section, and
Appendix E, need to be developed with more rigor. The framework needs to discuss in generic terms
how background levels of various factors like those listed in Figure 3 (conceptual model shown on
page 21) might influence the different endpoints identified in the conceptual model, and how these
expectations may therefore logically relate to technical requirements of CRA methods. The diagram
G-16
-------
Ken Bogen
of this conceptual model should reflect feedback loops not currently diagrammed (e.g., loss of
property value, aesthetics, etc., tend to negatively affect the socioeconomic system, which in turn
tends to increase rates of crime, traffic accidents, and communicable-pathogen transmission).
In discussing the scope of conceptual models that can be considered in CRA, some expression should
be given to the concept that the more overly broad the CRA scope is, the more difficult it will be to
implement systematic and consistent CRA methods, and so the more difficult it will be to interpret
and apply CRA output to help resolve a policy question. Conversely, CRAs with more narrowly
defined scope of problems may more easily contribute substantively to public policy decisionmaking.
There is text on p. 23 (section 2.3, 1st para., last 2 lines) that refers to the extent to which CRA-team
"members accept" CRA conclusions. This text unreal!stically implies that those members generally
will and/or should themselves be free to define the scope of CRA. The Framework should point out
that statutory mandates, corresponding regulations, and legal interests (e.g., property rights or other
due process rights) may constrain and/or define most if not all risk acceptability criteria applicable
to CRA.
On page 17 (lines 3-4 from bottom) it is stated that "it is important that community involvement be
sought and encouraged," and on p. 18 provision of "the opportunity to obtain technical assistance"
is listed among highlighted Guidelines for Stake holder Involvement. However, these briefly stated
recommendations for supporting stakeholder involvement are totally inadequate to address the
absolute requirement for substantive access to trusted technical expertise in order for any stakeholder
to participate meaningfully in a CRA process in which their vital environmental health interests are
at stake. Here, the Framework fails to note the unique feature of CRA that makes it more of a high-
stakes process than more typical (non-CRA) risk assessments. Namely, because CRA generally will
focus on integrated risk of multiple endpoints arising from multiple sources through multiple
pathways, it generally will involve integrated risks substantially larger than those addressed in a
conventional (non-CRA) assessment. Insofar as CRA may typically involve such high stakes, the
Framework ought to propose consideration of a legal right to "environmental health defense counsel"
for stakeholders participating in a CRA process, analogous to the right to legal defense counsel
during criminal procedures.
Finally, it should be emphasized in this section that CRA is a potentially resource-intensive type of
analysis. For this reason, the Framework should refer to the utility of identifying/developing and
adopting a set of conservative-default bounding assumptions and methods (CRA-inference bridges)
that would be applied to CRA-related factors in the absence of better information (analogous to
defaults used for routine risk assessments). In any particular case, CRA defaults could be used (in
combination with other routine defaults) to define a worst-case cumulative-risk scenario to be applied
routinely prior to CRA/scoping to determine whether a more formal/rigorous CRA process is
required. Reference to the need to develop such defaults, and illustrations of specific aspects of CRA
for which such defaults are needed, should be emphasized throughout the Framework (Sections 1 -4).
Section 3. Cumulative Risk Analysis
The logical distinction between Sections 3 and 4 needs to be clarified, insofar as certain topics
appears to be addressed de novo in each of these sections . For example, Section 3.1.1 (Combining
Individual Stressor Risks) appears to duplicate concepts covered in Section 4.1.3 (Combining
G-17
-------
Ken Bogen
Different Types of Risk). Section 3 should begin with a clear schematic diagram
summarizing/illustrating the generic type of technical/logical problem posed by CRA. This diagram
should be more generic than current Figure 3 (conceptual model), and should in general terms
represent a simplified version of the general conceptual problem being addressed in any CRA. Some
heuristic notation might be helpful in this regard. For example, it would be useful to adopt some
notation for and to illustrate m multiple stressors or sources (S,) that may individually or jointly affect
any among a total of n different health and/or ecotoxicity endpoints (Ey) considered (via regulatory
requirement and/or by the CRA team) to be relevant in a given CRA, where / = 1,2, ..., w, andy =
1, 2, ..., n. The diagram should how n different (but not necessarily mutually exclusive) subsets of
S, may each affect each of the Ey, For example, each member of the source set Zt = {Sl3 S3, S4}
might affect endpoint Eb sources Z2 = { S2, S3, S4, S5} might affect endpoint E2, and so on, where (Eb
E2, E3, ..., EJ might denote (cancer, asthma, reproductive/developmental toxicity, ..., serious
ecological disruption}. Cumulative or aggregate risk, R,^, should be defined as some (yet to be
defined) function of the entire set of Ey conditional on Zy, Such an illustration would be analogous
to (or could simply reprint) the corresponding figure that appears in the first section of Appendix I
of Science and Judgement in Risk Assessment (NRC, 1994), which appendix is referred to in
Chapter 11 (titled "Aggregation") of that report, which chapter and appendix specifically address the
topic of calculating and characterizing aggregate measures of environmentally induced health risk.
Sections 3.1-3.2 of the Framework (General Approaches to Cumulative Risk Assessment) need to
be researched better and to be rewritten accordingly, citing and referring to concepts that appear in
more of the relevant literature (e.g., NRC, 1994, Chapter 11 and Appendix I). This revised
discussion would benefit by explaining what specific questions or issues are being addressed in terms
of the figure proposed above relating S;> to Ey, In particular, Sections 3.1 and 3.2 do not refer to a
coherent description or any example of a purely probabilistic approach to CRA (e.g., like the
approach described in NRC, 1994, Chapter 11 and Appendix I), and thus fail to give any indication
that such an approach exists and may offer key advantages. For example, the analytical mechanics
of performing CRA are simplified greatly by adopting such a purely probabilistic framework, by
focusing the analysis on quantifying the cumulative/aggregate risk R^ defined as the
risk/probability/likelihood of inducing one or more of the (insofar as feasible, independent) endpoints
EJ defined via regulatory requirement and/or by the CRA team to be relevant in a given CRA. As
noted by NRC (1994; Chapter 11 and Appendix I), when defined in this way, R^ is readily
calculated using de Morgan's rule as
in the case of independent Ep where R7(E/) = Prob(E/) = the risk or likelihood of inducing endpoint
EJ. Recently, Bogen (2001) used this approach to quantify combined risk of cancer and noncancer
endpoints induced by the chemical trichloroethylene (TCE), including quantitative characterization
of associated interindividual variability and associated uncertainty (including uncertainty regarding
mechanism of carcinogenic action). This general approach is readily extended to cases involving
multiple sources and/or endpoints, as explained in NRC (1994, Chapter 11 and Appendix I).
Technical hurdles involved in implementing this approach become those of defining the set of
G-18
-------
Ken Bogen
relevant (preferably independent) endpoints Ep and of quantifying Prob(E/) (the likelihood of
inducing each/h adverse health or ecotoxic response considered unacceptable) as a function of Zy,
As noted inNRC (1994), if it is possible to adopt endpoint-specific riskmodels all ofthe form, R7(E/)
= 1 - expf-H^Z;)] in which Hy denotes a hazard-type function of a corresponding set Zy. of sources
affecting the/h endpoint, then Eq. (1) reduces to
Rcum=l-exp -
V
Such a probabilistic approach would of course require explicit models ofthe risk of occurrence of
each endpoint (including noncancer endpoints) considered. The Framework currently only refers
to existing, nonprobabilistic, approaches to noncancer endpoints that are currently used or are being
considered by EPA. The recent report by Bogen (2001) illustrates an alternative probabilistic
approach to noncancer endpoints, in which methods used for integrated quantitative treatment of
uncertainty and variability are made consistent with those used for probabilistic assessment of cancer
risk. This report addresses many issues concerning the implementation of probabilistic methods for
noncancer endpoints, and cites a number of related references (e.g., Lewis, 1993; Dourson et al.,
1994; Slob and Pieters, 1998)—none of which issues or references are discussed in the current
Framework. Reference in the Framework (in Section 4.1.3.1, p. 50-51) to an ad hoc procedure of
assuming that all RfDs correspond to a given (e.g., 2 x 10"4) risk of endpoint occurrence should be
dropped and replaced by references to well-documented probabilistic approaches for noncancer
endpoints (Lewis, 1993; Dourson et al., 1994; Slob and Pieters, 1998; Bogen, 2001), particularly in
view ofthe fact that not all ofthe extrapolation factors now routinely used for these endpoints pertain
to uncertainty (one pertains to variability).
Any approach to CRA must carefully define the set of relevant endpoints Ey, The Framework should
emphasize that precisely how this is done has important logical and practical implications for how
Rcum may be calculated and interpreted. For example, the risk R7(E/) on inducing endpoint Ey may
differ among different people in a population at risk for some Ey (e.g., cancer conditional on all
carcinogen exposures), but may be unaffected by interindividual variability (e.g., in exposure or
susceptibility) for other Ey (such as ecological or aesthetic effects). Defining the latter risks R7(E/)
in terms of individual risk per se will thus complicate calculating R^ if a probabilistic approach to
CRA is used, and perhaps if other approaches are used as well. In contrast, the probabilistic
approach to CRA could be facilitated by defining R/E^) in terms of population risk, i.e., in terms
ofthe predicted number N(Ey,) of cases of Ey, For example, if the definition R7(E/) = ProbfN^,) >
1] were adopted, then the probabilistic approach corresponding to Eqs. (1) or (2) would be simplified
because R7(E/) (thus defined for each interindividually variable or "heterogeneous" endpoint) would
be calculated by integrating over corresponding risk values pertaining to all individuals at risk (see
Bogen and Spear, 1987; NRC, 1994; Bogen, 1995), thereby yielding a variability-independent
measure of cumulative risk. Risk-equity issues could still be addressed by using additional, equity-
related risk-acceptability criteria—see NRC, 1994; Bogen, 1995. Alternatively (or additionally),
similar simplification can be achieved for all heterogeneous endpoints Ey. by defining R7(E/) only with
respect to those persons in the population at risk who are reasonably maximally exposed (e.g.,
individuals adjacent to a proposed source), or to those persons who will incur the greatest increased
G-19
-------
Ken Bogen
risk (e.g., children or other members of a sensitive subpopulation who might be located adjacent to
a proposed source).
Although population risk is a penultimate form of cumulative risk, and would thus appear to be a
natural focus of CRA, it is barely mentioned in the Framework. The quantitative relationship
between jointly uncertain and variable individual risk and uncertainty in population risk, and the
requirement to characterize uncertainty in population risk in order to calculate quantities such
potentially CRA-related quantities as Prob[N(E/.) > 1], are not mentioned at all in the Framework.
These relationships were first identified by Bogen and Spear (1987), and are further discussed and
illustrated in additional references (see Bogen, 1990; NRC, 1994; Bogen, 1995).
The first line from the bottom on p. 35 that reads "At the current state of the science, these factors
[addressing the issues related to vulnerability, meaning differential susceptibility, sensitivity,
exposure, or resilience] have not been extensively developed beyond correlations between mortality
rates and several socioeconomic factors such as income" is not at all accurate. Variability with
regard to susceptibility was discussed in detail in NRC (1994), and the current state of knowledge
concerning epidemiologically based (e.g., oncogene-specific) risk factors provides empirical data
upon which at least crude estimates of the magnitude of heterogeneity in susceptibility to toxic
response can be based. Current approaches implemented (e.g., by EPA) to address risk of noncancer
endpoints routinely employ a 10-fold factor to address heterogeneity in sensitivity. More
quantitative probabilistic approaches that address variability in susceptibility for noncancer endpoints
were mentioned above (Lewis, 1993; Dourson et al., 1994; Slob and Pieters, 1998; Bogen, 2001).
Section 4. Interpretation (Risk Characterization)
As its title indicates, the focus of this section is not at all clear. "Interpretation" would appear to
involve some aspects of risk management, whereas risk characterization refers to summarization of
risk-analysis output in a form that is useful to risk managers, i.e., that is relevant to a specified or
potentially relevant set of risk-acceptability criteria. Some redundancy with issues covered in
Section 3 was discussed above. I would suggest that the material in Section 4.1.3. (which might be
renamed, Characterizing Cumulative Risk Using a Common Metric) seems to be the central problem
in risk characterization for CRA. In this Section, again, a purely probabilistic approach is not
presented clearly (or really at all), as was discussed above in detail. Section 4.3 (Uncertainty
Analysis) does not (but should) refer also to variability analysis, and should accordingly discuss not
just Morgan and Henrion (1990), but also (or rather) relevant references that explicitly address the
analysis of joint uncertainty and interindividual variability in predicted environmental risks, and the
related problem of compounded uncertainty in risk analysis (e.g., Bogen and Spear, 1987; Bogen,
1990,1994,1995; NRC, 1994).
References
Bogen, K.T., and R.C. Spear. 1987. Integrating uncertainty and interindividual variability in
environmental risk assessment. Risk Anal. 7, 427-36.
G-20
-------
Ken Bogen
Bogen, KT. 1990. Uncertainty in Environmental Risk Assessment. Garland Publishing, Inc., New
York, NY, 195 p.
Bogen, K.T. 1994. A note on compounded conservatism. Risk Anal. 14, 379-81.
Bogen, K.T. 1995. Methods to approximate joint uncertainty and variability in risk. Risk Anal. 15,
411-19.
Bogen, K.T. 2001. Methods for Addressing Uncertainty and Variability to Characterize Potential
Health Risk from Trichloroethylene Contaminated Ground Water at Beale Air Force Base in
California: Integration of Uncertainly and Variability in Pharmacokinetics and Dose-Response.
UCRL-ID-135978 Rev. 1. Lawrence Livermore National Laboratory, Livermore, CA..
Dourson, M.L., S.P. Felter, andD. Robinson. 1996. Evolution of science-based uncertainty factors
in noncancer risk assessment. Regulatory Toxicol. Pharmacol. 24, 108-120.
Lewis, S.C. 1993. Reducing uncertainty with adjustment factors: Improvements in quantitative
noncancer risk assessment. Fund. Appl. Toxicol. 20, 2-4.
National Research Council (NRC). 1994. Science and Judgment in Risk Assessment. NRC
Committee on Risk Assessment of Hazardous Air Pollutants, National Academy Press,
Washington, DC.
Slob, W., and M.N. Pieters. 1998. A probabilistic approach for deriving acceptable human intake
limits and human health risks from toxicological studies: General framework. Risk Anal. 18,787-
798.
G-21
-------
Post-Meeting Written Comments From Jim Butler
G-23
-------
Jim Butler
COMMENTS ON THE
FRAMEWORK FOR CUMULATIVE RISK ASSESSMENT
by
James P. Butler, Ph.D.
Argonne National Laboratory
August 2001
The following comments are based on my review of EPA's Framework for Cumulative Risk
Assessment (August 2, 2001 Review Draft). My comments address issues raised in the questions
that were provided to help guide discussions during the peer consultation workshop on August 22-
23,2001.
OVERARCHING ISSUES
Comment on whether the Framework adequately captures, describes, and reasonably
organizes the key issues for cumulative risk assessment?
Many of the key issues have been adequately captured and described, often with real insight into the
practical implications of implementing this type of framework. The authors are to be commended
for tackling such an important and technically challenging project. Because semantics are often a
problem with cumulative risk assessments, the Framework will be particularly helpful in providing
consistent definitions for risks that are referred to as cumulative, total, aggregate, incremental,
holistic, integrated, etc.
With various interpretations of cumulative risks today, it is important to clearly state the goal of the
Framework in the introduction. To put the significant changes being proposed here in context, the
approach that EPA risk assessments have historically followed should also be explained (i.e., a
methodology which focuses on the incremental risks of contaminants emitted from a single source).
In addition, the possible uses of the results of a cumulative risk assessment (e.g., during the
permitting process) should be included early in the document.
What additional issues, if any, should be covered?
It should be emphasized that a cumulative risk assessment can be done on a variety of scales, from
looking at aggregate risks of a single pesticide (under the FQPA) to a broad screening analysis of
community hazards. On a related note, the Framework places most of its emphasis on chemical risk
assessments, especially in Section 3 on the analysis phase. Approaches for placed-based cumulative
risk assessments need to be included for evaluating exposure and risk from multiple sources within
a given geographic area. For example, often this type of analysis would provide stakeholders with
G-24
-------
Jim Butler
a more complete picture of a particular facility's emissions in relation to environmental loadings
from currently operating facilities in the same community.
The National Environmental Policy Act (NEPA) approach to evaluating cumulative impacts is
mentioned in the Framework as having parallels with cumulative risk assessment, but then is not
presented as a possible model to use (especially for community-based assessments).
Finally, how does the guidance in the Framework "fit" with other guidelines that have recently been
issued, e.g., from the Science Policy Council, Office of Pesticide Programs, OAQPS's National-
Scale Air Toxics Assessment, and the Cumulative Exposure Project?
Comment on whether some of the topics presented in the draft Framework are too generic for
the intended purpose?
At this point early in the development of the framework, I think the organization of the report is
more important than if some topics are too generic (and in some cases, too specific). There are so
many variations on the theme of cumulative risk assessment, and this report addresses only some
of them. It is probably better to keep it all the different approaches and levels of analysis under the
umbrella of cumulative risk assessments in this framework because of some common concepts, and
also because methods and data may be transferable to situations other than what they were originally
intended for.
I suggest revising the report so that the conceptual, overarching components of cumulative risk
assessment are clearly explained and organized in a concise framework, including a schematic
representation of the process. The more detailed technical issues and methodologies could then be
presented as a series of issues papers that would be appended to the Framework report. The level
of detail would, of course, vary depending on the topic, and may include the generic material from
other guidance documents. This kind of modular approach would give EPA more flexibility in
further developing the framework, while having the advantage of keeping the overall framework
consistent and relatively simple. The issues papers (or white papers) should also include details on
additional approaches to cumulative risk assessment that are currently being explored (including
screening-level analyses, place-based assessments, comparative risk assessments, NEPA cumulative
effects analyses, and hazard assessments) that may fit more easily within a flexible, streamlined
framework than they would in the present framework. In addition, the issues papers could include
summaries of case studies of cumulative risk projects that would extend the Framework from
theoretical to practical approaches and applications. This approach would also address the
disconnect between the rigid analysis approaches of Section 3 and the flexible interpretation
methods of Section 4.
Comment on whether the Framework adequately conveys the state of the science with respect
to cumulative risk assessment and the areas that are in need of further research and
development.
Long-term research areas that would improve the process for evaluating cumulative risks have been
identified in Appendix F. Important data gaps and information needs are also identified in the text,
albeit often at a superficial level. However, future inserts may provide supplementary information
G-25
-------
Jim Butler
on key topics (e.g., biomarkers). The development of a series of issues papers (as described above)
would also strengthen the discussion of the state-of-the-science with respect to cumulative risk
assessment-related topics.
PROCESS ISSUES
Stakeholder Involvement Throughout the Cumulative Risk Assessment Process
The level of stakeholder involvement at different stages of the process is not always clear. A
suggestion that I have for engaging stakeholders early is to identify a "point person." This is not to
imply that stakeholders must speak with a single voice (not likely in any case), but that they have
at least one person to help facilitate interactions and identify available technical resources and other
sources of information. Another suggestion for engaging stakeholders early is to set up a public web
site for the project. A variety of resources can be posted, including cumulative risk tools and
databases, project-related news, list of experts, glossary, reports, related links, etc. An online
discussion forum could also be included on the web site as a more interactive way of exchanging
information with stakeholders.
In Section 2.1.3, it is stated that EPA is the decision maker and will determine if stakeholder
involvement will be useful. This sounds too open-ended, with the potential to avoid the more
difficult cases. What are the checks on this process issue? Also, EPA and the stakeholders should
agree early in the process about timing for the public release of results during the course of the
study.
Using the Results of Cumulative Risk Assessment
In discussions of possible outcomes with stakeholders, it is important to discuss not only outcomes,
but also the limitations on the information that is generated. By this, I mean not just statistical
uncertainties, but also what the overall study limitations are. For example, there could be limitations
related to the inherent quality of input data, mapping (boundaries of emissions), spatial distribution
resolution, inconsistent treatment of chemical classes/mixtures in various databases, etc. Know what
the appropriate applications of the information are ahead of time.
There are also situations where stakeholders are focused on a specific issue, e.g., they may be
interested in only permitted or regulated sources. This focus may not only rule out complete
categories of stressors and pathways, it is intended to yield cumulative risk assessment results that
will have specific uses, e.g., to influence the permitting process or EPA's risk management efforts.
Cumulative Risk Assessment in the Broader Context of Environmental Analysis and Decision
Making
One of the most obvious situations for conducting cumulative risk assessments is in an
environmental justice context. But I think a more immediate question is what are existing
environmental analysis and decision-making approaches and tools that may be adapted for use in
cumulative risk assessments? The situations that would benefit from cumulative risk assessments
are (or will become) readily apparent. The problem has been how to go about evaluating cumulative
G-26
-------
Jim Butler
risks today, often with limited resources. For years, cumulative effects have been assessed under
NEPA. Similarly, EPA's community-cased environmental protection and comparative risk
assessment efforts both have some principles in common with cumulative risk assessment. These
types of approaches and outcomes have not been addressed sufficiently in the Framework.
Another approach that may be more practical in the short-term is the use of Geographical
Information Systems (GIS) as a way to present cumulative hazards and/or risks and thus inform the
decision-making process. The use of GIS and related mapping techniques was mentioned as a
possible approach for avoiding the common metric dilemma, but should be described in greater
detail. This approach has been used for screening cumulative hazards of air toxics in the Chicago
metropolitan area.
TECHNICAL ISSUES
Approaches to Cumulative Risk Assessment
The four approaches discussed in the Framework report are a good starting point, especially for
single chemical/multiple sources scenarios (they seem less suited to community-based scenarios
with exposures to mixtures of hundreds of chemicals). But as mentioned in some of my other
comments, there are a variety of related approaches that may provide useful interim tools until a
more comprehensive cumulative risk methodology can be developed. There are times when
estimating risks, as a screen, may be more cost-effective or simply the level of analysis that is
needed (or available) for the decisions being made. For example, hazard and risk-based screening
approaches have been used in Chicago and Baltimore, respectively. Other examples include the
NEPA approach for a somewhat qualitative, big picture view of cumulative impacts or possibly a
weight-of-evidence approach using available EPA databases. These kinds of practical approaches
should be included with case study examples or lessons learned. It would be better to not limit the
framework by the available methods, but to structure the document so that additional modules can
support the general framework as additional methods are developed. Another issue paper that could
be appended to the Framework report is a discussion (and status) of EPA's efforts to harmonize
cancer and noncancer assessment techniques, which is an important methodological hurdle to enable
cumulative risk assessment to move forward.
Using Biomarkers of Exposure or Effect
The Framework discussion does not adequately characterize how biomarkers can be used with more
traditional risk estimates. However, that may be remedied in the insert planned for Section 3.4.
While the use of biomarkers in cumulative risk assessments is currently limited, the potential for
verifying the modeled results is promising. Biomarkers will also be important for defining
background exposures and in health surveillance studies. This section of the Framework can be
expanded as additional biomarker and biomonitoring research is conducted.
Combining Chemical and non-Chemical Stressors
G-27
-------
Jim Butler
One problem with combining chemical and non-chemical stressors will be the introduction of value
judgments. For example, the same psychological stressors affect people to varying degrees. The
need to explicitly acknowledge these values judgments should be stressed. It may be helpful to
illustrate the difficulty of combining different types of stressors by including a hypothetical
example. A visual or graphical representation of different stressors could be presented, along with
a discussion of the significant limitations (e.g., caveat about not implying a causal relationship). The
chemical exposure and viral infection example (mentioned in the discussion question) could be
further explored along these lines.
Combining Different Types of Risk
As discussed under Process Issues, the use of GIS is a promising approach for presenting different
types of risks to stakeholders. It is one way to present integrated risk results without mathematically
combining disparate measures. The approach used for air toxics screening in the Chicago
Cumulative Risk Initiative involved the use of GIS and indices of toxicity, emissions, and ambient
concentrations.
The Framework sometimes makes cumulative risk assessment sound like an all or nothing
evaluation, i.e., all accumulated risks must be accounted for and then integrated. That's an ideal
endpoint, but useful information can still be provided before having all the methods available to do
this. Instead of a total of all risks, sometimes we just need to identify the stressors or sources that
contribute the most to a population's risk. A relative assessment of cumulative risks (or hazards or
exposures) may satisfy the project's original objectives, yet technically not be an integrated
assessment of all accumulated risks.
Limits on the Information Provided by Cumulative Risk Assessment
If it is known that the results of a particular cumulative risk assessment will be severely limited
because of a lack of data or available methods, it may be advisable to start with a screening analysis
to set priorities for a subsequent more detailed, focused study.
Appendix C mentions data quality issues, including the problem of combining data of different
quality in a single assessment. Other related problems include combining data from different
sources and/or different originally intended uses (like many of EPA's databases).
Distinguishing between Cumulative Risk and Cumulative Impacts
The Framework mentions cumulative impacts in the context of NEPA, but does not provide enough
information for the reader to understand the process and potential applicability to cumulative risk
assessments. The advantages of highlighting this methodology as a possible starting point for
certain cumulative assessments was discussed above. A thorough analysis of the pros and cons of
this type of cumulative impacts assessment should be appended to the report.
Completeness (Quantitative vs. Qualitative)
G-28
-------
Jim Butler
The Framework report is geared toward the quantitative assessment of cumulative risks. While
ultimately preferable, this often is not possible with existing methods. Plus, there is the issue of
available resources; even a screening-level analysis can be labor and cost-intensive. Cumulative risk
assessment should be considered a tool that can be used at different levels, as long as there is also
a clear understanding of the different levels of uncertainties. Screening analyses often employ semi-
quantitative or qualitative approaches, e.g., an impacts matrix, that would enable at least a broad
assessment of risks to go forward. I think the approach section should be inclusive and not rule out
other types of more qualitative cumulative assessments, including screening approaches, hazard
analyses, impacts assessments, comparative assessments, etc.
G-29
-------
Post-Meeting Written Comments From Douglas Crawford-Brown
G-30
-------
Douglas Crawford-Brown
Review of Framework for Cumulative Risk Assessment
Douglas Crawford-Brown, University of North Carolina at Chapel Hill
General Comments
The document is a good start, and the Agency is to be commended for beginning a
review process before a draft is fully ready. The history, methods and use of cumulative
risk assessment is a new and rapidly expanding field, and a difficult one to summarize
succinctly. Given this complexity, it would be surprising if the first drafts were able to
capture the issues and distill a coherent framework.
At the same time, it is especially important that any Framework document establish at
least a partial structure to what can be an overwhelming array of issues. The present
document begins in that direction, and then falters as the details begin to pour in. While it
represents a good start, and offers some insights into what a framework might look like,
it does not in the end produce an explicit framework. The resulting discussion does not
present a methodological framework. It might be argued that such a framework would be
too technical and should be the domain of a later guidance document. That seems a fair
enough argument. But the document should provide at least a conceptual framework, a
systematic exploration of the issues that will arise, the order in which they are likely to
arise, the kinds of answers that might be produced, and how these answers fit into the
overall framework of analysis. I don't think it does that at present, but could with some
improvements to both content and organization.
As an example, and only an example, I append a sample framework of questions that
might be asked (see the files cral.doc through cra7.doc). These are organized into a stage
on Problem Formulation, a stage on Data and Analysis, and a stage on Synthesis and
Interpretation. For each of these stages, I have provided a representative flow of issues
that might be raised. For each issue, I then have provided some representative questions
that might be asked, and some representative answers.
My sense of a framework document is that it does not provide detailed answers to
specific questions. It does not lay down hard and fast rules for how the cumulative risk
assessment must proceed. It is based on recognition that any assessment, and even a
guidance document, will result from a dialogue between several parties. The goal of the
framework then is to bring structure to that dialogue, to provide some guidelines as to the
issues that will be discussed, why these issues are important in the flow of assessment,
the range of possible approaches to addressing these issues, etc. It provides a map, and in
a sense a checklist, through which the parties may proceed somewhat systematically to
be sure they have raised all of the important issues.
Again, I don't think the current draft does that. It has a focus and order at times, but this
evaporates periodically and the reader is left with a series of discussions that probably are
important in some sense, but are not articulated clearly. It is not clear how each issue
G-31
-------
Douglas Crawford-Brown
relates to larger issues, and how all of these issues fit into a systematic framework of
assessment stages. I believe this is a natural state of affairs in first documents,
particularly when the issue is as conceptually complex as the one surrounding cumulative
risk assessment. But there is still a need for great improvement, with some framework
superimposed on the discussion. It should be possible to summarize this framework both
verbally and graphically on a single page, and to use that organization throughout the
document. I won't pretend that the example I provide in the attachments is the ideal
example, but it does give the flavor of my thoughts.
Answers to specific questions
The following responses apply to specific questions addressed in the three sessions of
Peer Consultation.
Over-arching Issues:
1. Does the framework adequately capture, describe and organize the key issues for
cumulative risk assessment? I believe my answer is contained in the general section
above. The document does raise many important issues, but these are not organized
into a coherent framework, or at least one I could follow. And there are issues
barely mentioned but significant (discussed below).
2. What additional issues, if any, should be considered? There must be a better
discussion of how the need to include all important factors will be balanced against
a desire to be as rigorous (mathematically) as possible. There must be a better
discussion of the spectrum of results from fully qualitative to fully quantitative and
probabilistic (and everything in between), allowing the reader to understand why
one or the other part of the spectrum might be desirable or feasible at any moment.
There should be a discussion of the kinds of questions that can be legitimately
answered by cumulative risk assessment and the kinds that cannot. I think the
authors can give insight into that issue without wandering into the domain of policy;
i.e. specify legitimate uses of the tool without specifying the kinds of policy options
which must be considered. There should be some discussion of the aspects of a
particular community circumstance that are on the table, and which aspects are not.
For example, can the discussion include whether the EPA is to the appropriate body
to address community concerns? Can it include whether risk assessment is the
appropriate tool? Etc.
3. Are some topics discussed too generically for the intended purpose? I believe
this may the case at many places in the document. I will comment on this more fully
in the last section. Perhaps the problem is not that the discussions are too generic,
but rather that they don't take place in a flow of reasoning that ultimately will give
them meaning. As a result, the reader encounters the issues without having had the
stage set clearly so they know why each issue is important in the overall flow of the
assessment.
G-32
-------
Douglas Crawford-Brown
4. Does the framework adequately convey the state of the science with respect to
cumulative risk assessment, and the areas needing further research and
development? To a degree, although more on the former than the latter. With respect
to the former, I do think the document lacks a good discussion of probabilistic
approaches to risk assessment, which allows one to move away from simply
summing already conservative quantities such as HQs. There are new methods to
allow this, and the document only presents the routine regulatory risk assessment
approaches. I think this will cause a real problem if conservatism is compounded
improperly.
Process Issues:
1. Is stakeholder involvement throughout the process described? Yes, the authors
have done a pretty good job at this. There might be slightly more detail provided as
to why stakeholders are being brought in. What role do they play in defining values,
in defining the assessment problem, in selecting methodologies, in interpreting
results, etc?
2. Is the use of results from the assessment discussed? That part needs to be
improved. The reader should have some guidance as to the kinds of questions that
can be answered by a cumulative risk assessment, and the kinds of questions that
cannot. Iwouldn 't want the document to prescribe the kinds of policy choices that
can be made (and the document avoids this), but it does need to provide insight into
legitimate and illegitimate uses of the tool as these pertain to the scientific basis of
the tool.
3. Is the broader context of cumulative risk assessment within environmental
analysis and decision making discussed? On the first issue, the answer is a qualified
yes. It is discussed at probably the correct level of detail, perhaps with the need for
an explanation of how cumulative risk assessment relates to other assessment tools
(such as air monitoring). I think there is a need to specify in a bit more detail the
decision structure. The reader does not receive any insight into the kinds of
decisions to which the tool will be applied. I understand the need to stay away from
policy issues here, but a decision framework focused only on risk characterization
would contain issues of decision principles as to the kinds of risk summaries needed
as input into any decision, and would still be useful even if it does not extend into
risk management.
Technical Issues:
1. Comments on the discussion of approaches to cumulative risk assessment. I think
this section is deficient. The document should drop back to the issue of protection of
a reasonable fraction of the population against unreasonable (unacceptable, etc)
risk, and with some kind of margin of safety, and present a coherent framework in
which different approaches might be compared against these decision goals. I
G-33
-------
Douglas Crawford-Brown
disagree that the approaches mentioned are competitors. I think they feed
information into a more robust assessment, and should be treated more as weight of
evidence considerations than as competitors.
2. Comments on the discussion of use ofbiomarkers of exposure or effect. I think it
is good to include this issue, but more detail could be given as to the uses of
biomarkers. They might be used to assign exposures, but they also might be used to
test models, to provide surveillance, etc. I also think the framework should draw
attention to a careful consideration of the strengths and weaknesses of using
biomarkers, and why these strengths might be more (or less, although 1 personally
think more) relevant in cumulative risk assessment than in past risk assessments.
3. Comments on the discussion of uncertainty analysis. I think this section is
somewhat deficient, but a good start. It does a good job of specifying types of
uncertainty, although it at times confuses variability and uncertainty, which should
always be kept distinct. The bulleted items on page 57 in the middle of the page are
not always so distinct. In fact, bullets 2 and 3 are really the same-just two causes of
uncertainty about model form. There should be some discussion as to why
uncertainty is considered; i.e. how it relates to concepts such as reasonable
confidence in protection, ample margin of safety, etc.
4. Comments on the discussion of vulnerability. This seems a good discussion, and
a necessary one. It is not clear how some of the causes of vulnerability ultimately
relate to the EPA mission, but I have no problem with that in a broad framework.
5. Comments on the discussion of combining chemical and non-chemical stressors.
This is not so well developed. There are many existing models for combining such
exposures, and none of these are mentioned. Perhaps it is OK to avoid discussion of
specific models in this document, but there should be some mention of classes of
models such as biologically-basedpharmacokinetic and dose-response models that
have the potential to include mixtures of exposures in moving organisms between
distinct states or stages of health (e.g. the Moolgavkar cancer model, or some
developed by Ken Bogen, or some developed by myself). The document should alert
the reader better to scientific advances in this area over the past 10 years. It should
describe clearly how default assumptions might be introduced, and why.
6. Comments on the discussion of combining different types of risk. This is
probably all that can be said on this issue at present. As with many of my comments
above, the need in the document is simply to better motivate the issue in the mind of
the reader. The document does a good job of pointing out difficulties, suggesting
some methods, and even calling up the possibility that effects should be left in a
matrix rather than combined into a single measure.
G-34
-------
Douglas Crawford-Brown
7. Comments on the discussion of limits of information provided by cumulative risk
assessment. I think this should be supplemented by a formal discussion of the
specific questions that can be answered by such an assessment, and the questions
that cannot, and why. The term "limits of information" generally refers to limits that
affect the ability to reach specific decisions. Some decisions can still be made under
limited information (e.g. ad campaigns rather than command-and-control
decisions). Providing a more formal framework of discussion may help in making
the decisions clearer in the document.
8. Comments on the discussion of distinguishing cumulative risk and cumulative
impact. I never really understood this distinction in the document. I can see why it
might be important, but the document didn 't explain the difference clearly, or why
the difference is significant for decisions. There needs to be an improvement in that
part of the document- not with respect to what it says but with respect to framing the
issue so the reader knows why the difference is being considered.
9. Comments on the issue of completeness of conception versus rigor of execution.
The document does not approach this issue at all. Cumulative risk assessment
involves a wide range ofstressors, receptor populations, kinds of information that
might contribute to the assessment, etc. The framework should help the reader
understand the full range of these issues, which can be treated quantitatively and
which must be treated qualitatively, how this treatment affects the ability to perform
specific tasks such as developing a single summary measure of risk (if that is
desired), and the conditions under which it is desirable or undesirable to leave out a
factor because it cannot be assessed or phrased in a way that is commensurate with
other factors. I would not want to see a factor left out simply because it cannot be
treated with the quantitative rigor of another factor, but we also don't want the
assessment compromised too much by poorly characterized factors. At times, it will
be necessary to give a little on the criterion of trying to be as quantitative as
possible, so as to avoid the loss of some key factors and, hence, a complete
conception of the problem in a community.
10. Comments on the issue of qualitative versus quantitative assessments. The
document mentions this several times, which is a strength. There should be an
explicit point in the framework where the reader confronts this issue and determines
whether quantification always is needed, when qualitative results might be OK, and
how these two kinds of results should be combined. Some examples could be given.
For example, the quantitative results might eventually be reduced to a more
qualitative scale (High, Medium or Low). Or the qualitative results could provide
"comments " tacked to the quantitative results. Or the assessment might simply raise
"redflags" associated with specific issues (e.g. density of emitters in a community;
presence of minority populations; special exposure pathways; etc); a high number of
such flags would indicate unacceptable cumulative risk, even if this isn 't quantified.
G-35
-------
Douglas Crawford-Brown
And so on. But this seems an important issue, especially since stakeholders are likely
to prefer more qualitative approaches. The experience of the EU in using qualitative
methods might be discussed, to show that "qualitative " is not "irrational. "
Further Specific Comments:
These comments apply either to points not discussed above, or to specific statements in
the document. In the latter case, page citations are given.
1. New science policy default procedures may be needed for cumulative risk
assessment (I would argue that they ARE needed). This should be mentioned.
2. There should be harmonization of cancer and non-cancer approaches. Given
probabilistic methods, this is now possible. It will not be feasible to combine these
two classes of effects under the existing approaches.
3. The document should emphasize the need for communities to have access to the
necessary scientific expertise, and not simply travel to meetings. This may include
access to models, data sets, etc. The OTAGprocess mgt be an example.
4. Page 17 should mention the role of discretionary power of the Agency. While
stakeholder participation is good and necessary, the Agency must eventually decide
what to do.
5. There should be some mention of the fact that such assessments may be done
retrospectively (to determine existing risks), prospectively (to assess the risks of
proposed facilities), or even creatively (to design a development plan for a
community).
6. Page 22, last paragraph. It is not clear why "hypotheses " are referred to here. I
don't see any sense in which scientific hypotheses appear, unless the author means
something like "possible contributors". The writing should be clearer on this point.
7. Page 25. At some point in this discussion, the issue of differences in severity of
effects should be raised. This might be related under the discussion ofQALYs, but it
seems reasonable to also raise the issue elsewhere (is this what is intended by the
discussion of categorical regression?).
8. Page 25. Risk factors are not so different from slope factors, UREs, etc, at least
conceptually. This issue should be raised in the context of weight of evidence, where
a framework is mentioned to combine information from different approaches.
9. Page 26. One problem with use ofbiomarkers is that the decision problem often
requires separation of contributions from exposure pathways so effective policies
G-36
-------
Douglas Crawford-Brown
can be located.
10. Page 28. My understanding of point of departure is that it does not refer to the
entire range of observed doses, but rather to the lower end of that range (from which
extrapolation may proceed if needed). But I could be wrong.
11. Page 29 and 30. RfDs andRfCs are not characterized by two-tailed confidence
intervals. There is significantly more confidence that the RfD is protective than that
it is non-protective. The document seems to suggest the threshold might lay
anywhere within a factor of 10 in either direction of the RfD. This is why it is
important to not simply addHQs in a cumulative risk assessment.
12. Page 32. Dose-rate effects are not mentioned, but they are important. It is
mentioned back on page 39, however. There probably should be a distinct section on
dose-rate and timing issues, and how they might complicate cumulative risk
assessment.
13. On a more general note, there might need to be a section talking about special
problems that arise in cumulative risk assessment, or perhaps problems that are
made worse under cumulative assessments. I am thinking of issues related to the
overall confidence in an assessment; resource needs; understanding of the margin of
safety that applies; etc. While cumulative risk assessment is not so different in
character from common methods used today, it does differ in terms of the level of
comfort with the results and their interpretation.
14. Section 4.1.1 think the methods discussed in this section all contain, at least
implicitly, an assumption of additivity. This should be mentioned (if I am correct)
and some mention given of methods that do not require additivity. That would
balance the discussion, which seems to me too heavily focused on existing
regulatory default procedures rather than possible advances.
15. At several points (e.g. on Page 51), there is too large of a distinction drawn
between analytic and deliberative steps. All steps involve both aspects, even those
classically assumed to be scientific. Science involves both analysis and deliberation
(that is why we have scientific bodes and not individual scientists speaking for the
field). And policy/management involves both analysis (the meaning of concepts such
as acceptable risk) and deliberation.
16. Page 57. A point-by-point sensitivity analysis can be misleading. A better
approach is one where uncertainty is factored into the analysis, rather than simply
adjusting each parameter one at a time across the range. The approach is based on
contribution to variance.
G-37
-------
Post-Meeting Written Comments From Amy Kyle
G-38
-------
Amy Kyle
Peer Consultation Panel
Cumulative Risk Framework Document
August, 2001 - Comments of Amy D. Kyle
Overall comments
Developing methods and approaches that allow for the consideration of cumulative exposures and risks is
important. Such approaches can better reflect human experience with pollutants and contaminants, as
well as other stressors, than can approaches that focus on single pollutants and incremental risks. As has
been widely documented, and is noted in the foreword at page vi, standard risk assessments typically
consider single pollutants and new releases or exposures, but not the accumulation of exposures over time
and the likelihood of an accumulation of effects.
The limitation of current risk assessment methods is very apparent to people from communities that are,
in a sense, at the receiving end of assessments. Estimates of risk levels that are quantified through
rigorous methods but exclude obviously relevant accumulations of exposures and effects have limited
credibility. Changes, either large or small, to reflect more of the actual burden of exposures and effects
that people are experiencing will improve the scientific validity and credibility of risk assessments.
The document, Review Draft, Framework for Cumulative Risk Assessment, August 2, 2001, represents an
impressive attempt to pull together, from varied sources, information that would contribute to a
framework for cumulative risk assessment. The document presents well-developed and nuanced
descriptions of many of these issues. The US EPA is to be commended for this important step.
The purpose of a framework is to identify key issues to inform a wide variety of interested and affected
parties. Therefore a balance must be struck between adequately characterizing the issues and
providing an excessive level of technical detail. The framework also discusses a process for gathering
information and defining the problem with stakeholder input.
When considering the process topics, recognize that our goal is to focus the scope of the cumulative
risk assessment on a parsimonious use of stressors, sources, pathways, and potential adverse effects
that are of interest to a particular place or community.
Overall, the RAF technical panel is seeking discussion and input on the questions and topics that
follow. We are not setting policies for how to deal with the issues, but rather identifying principles and
professional judgements which may lead to more specific guidance later.
G-39
-------
Amy Kyle
One way to look at this is to think first about what is most fundamental. What would one always
want to do when considering cumulative risks? Trying to identify the essence of the framework would be
helpful.
The process for involvement of stakeholders and integration of technical and policy concerns,
described in detail in the National Research Council's 1996 book entitled Understanding Risk, is not
unique to cumulative risk assessment but would be equally relevant for other kinds of risk assessment
exercises.
I do not agree that focusing on cumulative risk assessments for a particular place or community is
always the appropriate focus. Principles of cumulative risk assessment may be just as integral to
assessments that do not focus on a particular place or community.
Comment on whether the Framework adequately captures, describe, and reasonably organizes
the key issues for cumulative risk assessment. Include in this discussion whether key terms
have been identified and clearly defined.
The document describes many of the key issues for cumulative risk assessment. The narrative
sections of the document are generally clear and informative and have about the right level of detail for a
document such as this.
The document provides useful narrative but lacks a true framework. The document would benefit
by the addition of a section that describes how the pieces fit together, perhaps in the form of a schematic
or flow chart. A framework would provide an organized way of looking at the topic and thinking through
the steps. I would not agree that it would be better to develop the document solely as an issue paper
without working to more clearly define the essence of the framework.
Such a framework could begin by selecting a scale for an assessment and proposing what, among
possible exposures, stressors, and effects, would be considered and over what term. The steps in the
analytic-deliberative process laid out in section 2, beginning with a full scoping step, followed by
analysis, and concluding with characterization, are appropriate. The explanation would be improved by
acknowledging that the steps may be iterative and that results from one step may influence another.
It might be useful to start with principles that should always be considered, and then work
through more complex points that might apply in some cases.
More linking of how the various kinds of analyses described in the document might apply to
these steps could be helpful.
It is important to recognize that the scope of cumulative risk assessments will vary considerably,
as some may involve one contaminant, simply considered in terms of its cumulative exposures and
effects, while, at the other extreme, "place-based" assessments may attempt to capture a multitude of
contaminants and stressors and consider the cumulative result. The document provides, at page 19, three
cases that could be used to explain the differences in approach and issues to be considered at different
scales. For each case it would be valuable to develop a model or schematic for the steps that would be
involved, who would be in charge of each step, what would be decided, and what types of methods or
G-40
-------
Amy Kyle
considerations would be relevant. Such as schematic or model could then refer to other portions of the
report for narrative explanations.
The document may give too much emphasis to the most difficult case for cumulative risk
assessment, illustrated at page 7, as a complete analysis of contaminants and other stressors as they
interact and affect a population. This case may be informative in terms of explaining the ultimate scale
and reach of a cumulative risk assessment. However, it may be unwise to set such an intensive analysis
as the ideal case for the process. It may be a more productive to incorporate more of the accumulation of
exposures and effects in more modest assessments. More attention is needed to the implications of
cumulative risk assessment on the basic methods now used for risk assessment.
The document needs to be cognizant of the need for EPA to retain the ability to carry out its
responsibility to protect public health and the environment while also including stakeholders in a process
like this. It would likely be necessary and appropriate for the agency to develop an initial concept for an
application of cumulative risk assessment methods, perhaps in a case where existing methods are
identified as likely not resulting in sufficient protection for the public. The EPA could identify the scope
and level of detail for such an assessment, identify appropriate stakeholders, and initiate a discussion of
how to frame the problem, altering its initial conception as appropriate through discussions with
stakeholders. However, as noted in the document, the EPA would need to be able to set some reasonable
boundaries around such a process to ensure that progress is made in a timely fashion.
The term "risk management" team, as used at page 13 may not be appropriate, as some solutions
may involve actions other than risk management and because, in some cases, who would need to be
involved in solutions may not be apparent until the analysis is complete.
The discussion of combining risks, in some places, implies that it is essential to the framework
that risks be combined. While it is valuable to combine risks, this may not be possible or desirable in all
cases. First, all stressors and effects may not best be expressed in the form of "risks." Second, it may be
better to consider different categories separately. Methods required to convert everything into a common
metric (such as DALYs) may have more disadvantages than advantages. Moreover, the document should
not assume that all "risks" would necessarily be quantified. The presentation at the meeting said that the
goal was to define the probability of harm from multiple stressors. This may or may not be the most
appropriate goal. As was discussed at the consultation meeting, use of weightings such as high, medium,
and low or percentiles may allow the comparison and summing across a variety of disparate outcome
measures without trying to convert everything into one common form. Alternatively, an index approach
to scoring may be useful in some cases. Methods that show both the original scores and a combined score
are most informative and allow participants to see clearly the implications of methods used. The
discussion of this topic should be reviewed and clarified throughout the document.
What additional issues, if any, should be covered?
a. Completeness
The importance of considering the completeness of a cumulative risk analysis, given its purposes
and scope, should be added to the discussion. The analyses used should be selected in order to address
the problem that is identified. In other words, the problem should drive the method. (The method should
G-41
-------
Amy Kyle
not drive the problem). We need to be able to consider, what are we leaving out? And, when we notice
that we are leaving things out, we need to be able to use the best available approaches to integrate them.
Bounding estimates that incorporate different assumptions can be useful here, but the principle should be
to integrate everything that is important.
To look at this clearly, during the analytic-deliberative process, the team should consider how
much of the problem they are trying to address has been captured by the methods that they have decided
to use. This then allows for course correction if need be to address topics that are important but that have
not been well captured by the methods used. There is often a tension between doing an analysis that is
more complete but that uses methods that have more uncertainty or are less familiar and doing an analysis
that is narrower but can rely upon well-understood and familiar methods. The document should
acknowledge that it is not scientifically sound to leave out consideration of factors or issues that can be
reasonably anticipated to be important to the scope of an analysis. It is more scientifically sound to use
an imperfect model that addresses all of the factors of concern than a perfect model that clearly ignores
many of them. This is the crux of much of the interest in cumulative risk assessment - we can no longer
justify acting as if people are exposed to a single chemical de novo in every assessment.
b. Intermediate level of analysis
The concept of the difference in approach in looking at the population level and the release level
is useful though it leaves out the intermediate level of looking at cumulative risks from ambient
exposures, which can be quite useful and informative. It would obviously be most desirable to model
exposures and risks at the individual level. It is not, however, parsimonious. Useful ranges can be placed
on exposures and risks at a higher level of abstraction that can be very useful for policy decisions and
much less intensive of knowledge and processing.
c. Support to community stakeholders: level playing field
Providing sufficient support for successful participation by stakeholders, particularly those
representing non-governmental and community organizations, is a challenge for a cumulative risk
assessment process as it is for other agency processes. The document says, at page 18, that it may be
necessary to provide certain limited forms of tangible support or to provide incentives for participation.
These may well be appropriate. However, to truly create a level playing field will, in some cases, require
greater efforts, as community stakeholder groups may need access to adequate technical analysis and
resources to be able to fully participate in the process.
d. Vulnerability
With regard to vulnerability, it may be appropriate and useful to consider whether there is any
synergy of vulnerability. If both vulnerabilities and exposures accrue to particular segments of the
population, then cumulative risks may be significantly higher for some, and this should be pointed out. It
may also be useful to link the discussion of vulnerability to the national discussion of eliminating health
disparities. Cumulative risk assessment will be important to eliminating disparities in the field of
environmental health, and this should be recognized in the document.
e. Clarify that cumulative risk assessment means considering all exposures
G-42
-------
Amy Kyle
The document should clarify that cumulative risk assessment, even for a single chemical,
considers all exposures. This means both those earlier in time and those from all sources. It also
considers the implications for both background and multiple sources when assessing dose and response.
New exposures should be considered in terms of where they fall on a dose-response curve after
adjustment for previous exposures, accumulated exposures, and other sources. The discussion of
background exposures now in Appendix E should be integral to the document.
/ Incorporate consideration of actual exposure when discussing thresholds
The document should note that it will be important to review assumptions about thresholds for effects in
light of cumulative exposures, cumulative risks, and the variability of human responses. Exposures
should not be represented as being above a threshold unless it remains above a threshold when these
factors are taken into account. Thresholds for risk should not be assumed when the data do not support
this. (See RfD document for methylmercury as an example.)
g. Children
It would be appropriate and helpful to include a section that discusses issues for assessment of cumulative
risk with regard to children. How to think about time patterns of exposure and of susceptibility (the
former is mentioned at page 7 but not described in any detail) could be covered.
h. Environmental justice
It would be appropriate to add a greater discussion of how cumulative risk assessment (and management)
is related to environmental justice. This is discussed from an historic perspective, but not with an
operational perspective.
Keeping in mind that the Framework is intended to address issues that are specific or unique to
cumulative risk assessment, comment on whether some of the topics presented in the draft
Framework are too generic for the intended purpose.
The discussion of current risk assessment methods seems to be excessive. Perhaps other sources
could be referenced.
The section on biomarkers is far more generic than it needs to be and does not seem to be related
to the essential purpose of the document. Biomarkers could be particularly relevant for cumulative risk
assessment. Biomarkers of exposure would ideally integrate all sources of exposure at the level of the
individual. Biomarkers of effect could analogously integrate all sources of effect at the level of the
individual. Obviously, biomarkers would need to be carefully considered and chosen to actually have
such capacities. It might be worthwhile in this section to at least identify leading candidates for
biomarkers that could contribute to this effort. For biomarkers of exposure, hair level of mercury and
blood levels of dioxin come to mind as potentially useful, in part because the body burden metrics have
been related to adverse health outcomes in recent assessment documents. In addition, there may be,
perhaps in the future, some application of the field of proteomics to this area.
G-43
-------
Amy Kyle
There has been some discussion of providing a better statement of how cumulative risk
assessment could be used, in specific cases. There may be some benefit in doing this. I would not think
it would be helpful, however, to tie these to specific legislative mandates, as the overall approach is not
linked to any specific statute and could probably be relevant to all.
The framework attempts to identify where methods and data are currently lacking or are
underdeveloped for application to cumulative risk assessment. Based on a suggestion during
the meeting with other Federal scientists, research oriented issues have been highlighted in an
appendix to the Framework. Comment on whether the Framework adequately conveys the
state of the science with respect to cumulative risk assessment and the areas that are in need of
further research and development.
To fully describe the "state of the science" with regard to cumulative risk assessment would be a
considerable undertaking that would need to incorporate the state of the science for each of the methods
and approaches that could be used. This seems to be rather beyond the scope of a framework document
of this type. Perhaps the most useful approach would be to summarize the state of the science with
regard to the reasons to conduct cumulative risk assessment.
One critical management need is to expand the number of chemicals that have toxicity values in
IRIS. This database has become increasingly outdated.
How to address other stressors is obviously a critical area of research that would apply to
cumulative risk assessment and to any other form of risk assessment as well.
Other comments
The introduction to the document seems to include a number of questionable statements.
Considering cumulative risk is not in conflict with approaches that seek to reduce or eliminate releases of
contaminants. It is just as valid to limit releases of contaminants because the cumulative risk is too high
as because the risk of a single chemical, considered apart from everything else, is too high. The effect
may be measured in the population, but the control may still need to be at the source and could come
about as a result of either "command and control" or incentive-based approaches. (It appears to me that
much more has been claimed for incentive based approaches than has been achieved by them.)
The discussion of regulatory history at page 1 is somewhat odd. The "command and control"
approach is usually contrasted to incentive-based approaches, not to those that consider population-based
risks. It is not particularly clear to me that this has reached the point of diminishing returns. Consider
the regulation of air pollution from power plants or SUVS, for example. Some things are well controlled
and others are not. There is no great uniformity of approach.
The leading reason that I believe that people are concerned about current risk assessment methods
is because they clearly leave out things that are important. When the risks of single chemicals are
assessed, exposures that occur before today (background exposures) are usually ignored. This means that
the actual risk is almost always under-predicted, as people very often have other, pre-existing exposures
that affect where they are on a dose-response curve. Moreover, exposures to other things are also usually
ignored. This means that people are almost always facing a higher, and often a very significantly higher,
G-44
-------
Amy Kyle
risk than what is shown in any single chemical, de novo assessment. Consequently, regardless of the
strategy chosen to manage the risk, there is widespread support for an approach that better reflects what
people actually experience.
Use of the notion of "margin of exposure" should be excluded from this document. This
approach has been used most extensively in conjunction with an attempt to convert the starting point for
non-cancer risk assessment from a no-effect level to a 10% effect level. Consequently, the term is
somewhat tainted for other purposes.
Page vii - In the third paragraph, the assertion that a key role of science at EPA is to reduce uncertainties
in decision-making seems to be a rather odd way to sum up the role of science. I would think that, more
fundamentally, the role of science would be to provide the best available information for decisions to
protect public health and the environment, the mission of the agency.
The contrast between methods that focus on a chemical and those that focus on a population is
drawn too sharply. As noted elsewhere, it would be possible and advisable to consider cumulative risk
for a single chemical and would require a change to existing methods.
Page 1 ~ In the first paragraph, use of the phrase "so-called sources" implies that you don't think that the
elements listed are actual sources of pollution.
Page 3 ~ In the fist full paragraph, I would advise acknowledging that there was much more to the
evolution of the environmental justice movement than what is described here. The discussion of NEPA is
very interesting.
Page 7, point 2 ~ The notion of presenting an integrated picture of risk, as presented here, is a good one.
However, as noted elsewhere, trying to combine disparate types of risk into a single measure raises issues
in itself. I don't think it is wise to exclude the possibility of presenting different kinds of risks in a table
as a positive outcome of an assessment.
Page 7, point 3 ~ This area seems to be one for which considerable additional research will be needed.
Page 8 ~ In the definition, the areas in which exposures can be accumulated should all be spelled out (and
the phrase etc. should then be deleted.)
Section 1.4 — 1 would add the diagram of the analytic-deliberative process given in the NRC report
referenced.
Page 10, Section 1.5 — It is a good idea to introduce some of the key conceptual differences between a
decision analysis approach such as that in Clemen's book and the largely biomedical model reflected in
current risk assessment methods. I am not sure that this belongs in this section, however. It also probably
needs a more thorough explanation.
Page 13 - I would recommend using the term "analysts" rather than "risk assessors" to reflect the point
made elsewhere that a variety of kinds of analysis may be needed.
G-45
-------
Amy Kyle
Page 16 ~ Background exposures should be addressed in cumulative risk assessments, not just discussed
as stated here.
Page 21 - Conceptual models should be included in this document for assessments of varying scope,
including examples that are simpler than this one. It is not a trivial task to distinguish between what is
"known or determined" and what is "assumed." I am doubtful that it will always be helpful or appropriate
to insist on agreement on a single conceptual model. The first full paragraph on page 22 seems quite
naive. It also seems to be in conflict with other portions of the document that better reflect the
"deliberative" elements of the section of stressors, sources, receptors, and so on.
Page 22, Section 2.2.3. What does the following sentence mean? "Those hypotheses considered more
likely to contribute to risk are given priority." As far as I know, hypotheses don't contribute to risk. The
earlier discussion about the need for methods, including elements of decision analysis, that address the
problem identified should be reflected here.
Section 3.1- This section is rather incoherent and needs to be re-thought. It now includes a somewhat
random collection of methods that could contribute to a cumulative risk assessment. They are not
mutually exclusive.
Page 27 - Using "mode of action" as a way to combine risk for chemicals is problematic since the true
mechanism of action of most chemicals is unknown and many chemicals have more than one.
Section 3.2.1. - This section should acknowledge that there is not always a threshold for non-cancer
effects of chemicals (see recent EPA assessments for dioxin and methyl mercury for example.)
Page 31 — The notion of a "mode of action" as proposed in the 1996 draft cancer guidelines is based on a
rather selective application of science and seems to be technically dubious. The fragmentary information
that has been accepted as a basis for a hypothesized mode of action to date seems to reflect a policy goal
of the agency rather than a truly scientific approach.
Page 39 ~ Even though a linear model has been used for carcinogens, there may be cases when the
average lifetime dose may not be the appropriate exposure metric. It could be that the averaging period is
too long or that earlier exposures, particularly in children, may have a greater risk.
It seems highly unlikely that it will be possible in most cases to characterize chemical exposures "in terms
of which other chemicals are present, and when."
Page 49 ~ The "margin of exposure" terminology does not reflect risk. This terminology is currently
being applied to represent a ratio between an exposure level and a very high effect level (ED 10), not a no-
effect level. The explanation of how this term is being used in the document is not correct. The phrase
should be deleted to avoid this confusion.
Page 56 ~ The second bullet point is not correct in this context.
G-46
-------
Center for Environment, Commerce & Energy
Post-Meeting Written Comments From Norris McDonald
G-47
http://communities.msn.com/CenterforEnvironmentCommerceEnergy
-------
Center for Environment, Commerce & Energy
PEER CONSULTATION
Provided By
Norris McDonald
President
Center for Environment, Commerce & Energy
For The
Review Draft
Framework for Cumulative Risk Assessment
Submitted To The
United States Environmental Protection Agency
August 2001
G-48
http://communities.msn.com/CenterforEnvironmentCommerceEnergy
-------
Center for Environment, Commerce & Energy
Introduction
Risk Assessment is a complicated undertaking. Cumulative Risk Assessment can
become unmanageable if limitations are not placed on the input parameters. It becomes a
case of "too much information in, nothing useful can come out." Paradoxically,
cumulative risk assessment becomes useless if enough inputs are not included.
Legislation, such as the Food Quality Protection Act of 1996, is forcing the
Environmental Protection Agency (EPA) to consider the cumulative effects of multiple
chemical exposures occurring simultaneously and over time. Various entities at EPA are
examining the feasibility of formulating cumulative risk assessment models. The EPA
Science Policy Council [via the Risk Assessment Forum (RAF)], EPA Office of Pesticide
Programs, and the Office of Air Quality Planning and Standards are examining various
models. The framework is being reviewed by Federal and State scientists, peer
reviewers, EPA's Science Advisory Board and will require approval from EPA's Science
Policy Council.
My comments will address the RAF's request for input in the following areas:
organization of the framework, additional issues, issue of generic versus specific content,
effectiveness of the methods and data, adequacy of state of the science, and the need for
more research and development. My comments will also address process and technical
issues.
Finally, I commend the participants in the Peer Consultation Workshop held in
August 2001 for their insightful comments and contributions to my thinking about the
framework: Ken Bogen, Lawrence Livermore National Laboratory, Elizabeth Boa,
American Chemistry Council, Jim Butler, Argonne National Laboratory, Douglas
Crawford-Brown, Dept of Environmental Sciences & Engineering, Univ of North
Carolina, Amy Kyle, School of Public Health, Univ of California-Berkeley, Beth
Mileson, Technology Sciences Group, Inc., Stephen Olin, International Life Sciences
Institute, Bill Rhyne, H&R Technical Associates, Jeniffer Sass, Natural Resources
Defense Council, Margo, Schwab, Johns Hopkins School of Public Health, and David
Bottimore, Versar. The workshop provided an informative platform for analyzing the
G-49
http://communities.msn.com/CenterforEnvironmentCommerceEnergy
-------
Center for Environment, Commerce & Energy
technical issues associated with developing a framework for cumulative risk assessment.
It also provided a good forum for comparing our views with the views of a wide variety
of independent scientists.
Organization of the Framework
The framework is very concise for an EPA document—under 100 pages. I
understand that there was a directive to keep it under this number. Unfortunately, brevity
was accomplished at the expense of including numerous references in the framework to
other documents. I found this to be distracting and intimidating. It gives one the feeling
that unless you have reviewed a large volume of background, historical documents,
legislation and regulations, you could not reliably render an evaluation on the topic at
hand. These references take up 13 single-spaced pages in the back of the document.
Maybe the references could be included in a separate reference and appendix document.
The elimination of all of these references and the massive listing in the back of the
document would provide a more readable and user-friendly framework. The framework
is breaking new ground and the technical concepts need to be clearly identifiable. The
additional space would also provide more room for informative charts and figures.
The Conceptual Model on page 21 is a good generic description. It should be enlarged in
order to provide a better view of the technical linkages and elements.
The final 38 pages of the document contain only 5 figures and no charts. It appears as
though this valuable tool is utilized in the front of the document and only given token use
in the remainder of the document. These 5 charts do not contain informative concepts or
descriptive scientific diagrams. They are basically definitions. The policy analysis list
on page 56 appears to be out of place in this scientific framework.
Conceptual Model. The framework provides a good, short description for developing the
conceptual model. The framework acknowledges the importance of the conceptual
model as a risk communication tool both at EPA and in interactions with the public. The
chart on page 22 (Desired Outputs for Problem Formulation) is very informative but is
G-50
http://communities.msn.com/CenterforEnvironmentCommerceEnergy
-------
Center for Environment, Commerce & Energy
presented in a dull, single-spaced format. It should be more fully developed and
presented in a design similar to the one on page 21 (Conceptual Model).
The chart on page 40 should be included as an appendix.
Additional Issues/Issue of Generic Versus Specific Content
Analysis Phase. Although the framework is considered to be a technical document, very
little guidance is given to addressing data gaps in toxicity mixtures related to health
effects. This is really the heart and soul of cumulative risk assessment. Can we put all
of the chemicals in a particular situation along with potential health effects in a computer
and get predictable results? Our current institutions are territorially separated in
addressing the science. EPA is treading in Center for Disease Control (CDC) and
Agency for Toxic Substances and Disease Registry (ATSDR) territory. From a strictly
scientific approach, these agencies need to work together as a seamless unit, or maybe
they should be combined in an Executive Branch, Cabinet-level Department - -
Department of Environmental Disease (DED). Cumulative risk assessment modeling
could provide the science to drive the policy. Then the policy could effectively drive the
scientific implementation. The framework should at least consider a new scientific
paradigm.
Figures. The figures provide easy-to-understand explanations of the risk assessment
concepts. Again, these figures should be enlarged to make the information more
accessible. The simple listing of EPA's Risk Assessment Guidelines and Selected Policy
and Guidance Documents was particularly helpful. The short listing of documents was
not overwhelming and the guidelines were comprehensive and diverse, but not
exhaustive. More information could be provided in figure and chart form. Many
National Environmental Policy Act (NEPA) documents utilize this approach very
effectively.
"Corporations" should be added to the chart on page 17 and it should be moved to page 6
under intended audience. Of course, this relates to my biggest criticism of the document:
making EPA the only intended audience. I recommended in the workshop and I
G-51
http://communities.msn.com/CenterforEnvironmentCommerceEnergy
-------
Center for Environment, Commerce & Energy
recommend now, that stakeholders should also be included as the intended audience.
The interested and affected parties included on page 17 have as great a stake in the
success of cumulative risk assessment issues as EPA risk assessors and managers. I will
address this issue in more detail later in my comments.
Environmental Justice. The mention of environmental justice in the introduction is
woefully inadequate. It lists a rudimentary three-line history of the movement. Note:
Benjamin Goldman should be added to the 1987 Lee item. Goldman was the principal
author along with Lee (page 3). The environmental justice paragraph should at least
mention that the Executive Order addresses disproportionate impacts on minorities due to
racism in addition to the quote about "Environmental human health analyses, whenever
practicable and appropriate, shall identify multiple and cumulative exposures." In fact,
an additional paragraph should be included which describes the complexity of race
relations in the context of cumulative risk assessment. It provides an excellent example
for illustrating the scientific complexities involved in integrating an abstract confounding
variable into a cumulative risk model. Some mention should also be made about
researcher bias. EPA is struggling with internal discrimination cases. How can they be
trusted with environmental justice cumulative risk assessments when EPA managers are
unwilling to quantify the weaknesses of many of their scientists, researchers and other
staff? These weaknesses translate into bad science, bad policy implementation and bad
service delivery. The Marsha Coleman-Adebayo Case ($600,000 discrimination
judgment against EPA) is a good example of this weakness.
Intended Audience. I disagree with the framework's targeting EPA as the intended
audience and final arbiter of cumulative risk assessment for America. Although it is
commendable that EPA is taking the lead in framing a scientific foundation for
cumulative risk assessment, excluding stakeholders as principle decision-makers and
targeted audience are major flaws in the methodology of the framework. Section 2.1.3
Agreeing on Participants, Roles and Responsibilities, should be incorporated into Section
1.2. Intended Audience. Because the development of a cumulative risk assessment is in
its infancy, stakeholders will be secure in their positions as formulators and respected
decision makers in achieving a broad scientific consensus. Stakeholders should be
G-52
http://communities.msn.com/CenterforEnvironmentCommerceEnergy
-------
Center for Environment, Commerce & Energy
incorporated into the risk assessment planning team as decision makers with equal
decision-making authority. In considering placed-based versus chemical-based
assessments, stakeholders outside of the EPA oftentimes have different, and more
insightful scientific perspectives. Moreover, EPA should not have the sole authority to
decide whether stakeholder involvement is warranted, as proposed in the framework.
Sometimes EPA is culpable in scientific oversights related to informing and protecting
the public from pollutants. Some science is very simple but is made complex by EPA
scientists and policymakers. For instance, if you can see the air during a hot, summer
day, it is not healthy to breathe. That is an indisputable scientific observation. This
conceptual model would work just as well as, if not better than, parts per million
quantification. Scientific observation as it relates to policy decision-making should
ultimately be tied to improving human health. We can still see the air in our major cities
after 31 years of having a Clean Air Act. And the metric for enforcing compliance is
basically ignored. This model continues as asthma deaths are rising exponentially. How
will cumulative risk assessment address this scientific issue? It should mix the complex
scientific assessment with simple indices.
Adequacy of State-of-the-Science/Need for More Research and Development
The framework seems to surrender to the potential complexity of the cumulative risk
assessment model. The framework shouldn't surrender. Most real-world environmental
threats have finite, though complex, inputs. There should be more analyses of the
interaction of a variety of chemicals, in a real world model, with epidemiological inputs.
The stroke example on page 25 is good, but a few more pages of this type of inquiry
would be very helpful. This epidemiological and toxicological data could be illustrated
in a form similar to the conceptual model on page 21.
The loss of life expectancy section is very helpful and informative.
The Vulnerability section would be a good place to expand on the environmental justice
components. The specific items could provide additional insights into differential
exposure, differential preparedness and differential ability to recover. The EPA Office of
G-53
http://communities.msn.com/CenterforEnvironmentCommerceEnergy
-------
Center for Environment, Commerce & Energy
Environmental Justice should provide additional information in this section. The Center
for Environment, Commerce and Energy will, in addition to addressing chemical and
biological stressors, examine hazards other than chemical and biological stressors in its
alternative Framework for Cumulative Risk Assessment. For instance, racism and black-
on-black-murder and assaults are huge stressors in the black community. Our multi-
hazard risk assessment will include these non-chemical, non-biological stressors in a
framework.
Section 3.2.1.5 (Hazards Other than Chemical Hazards) on page 34 could be combined
with Section 3.2.2.5 (Exposures to Non-chemical Stressors). Section 3.2.2.4
(Subpopulations with Special Exposures) could be combined with a new Environmental
Justice/Vulnerability section. I would consider dropping the subpopulation designation.
It diminishes the importance of these groups. They aren't subs. They are populations.
Although examples of runners, children, the elderly and others are used, these
populations would be better served in the framework under a different heading. In the
context of cumulative risk assessment, these populations are just another variable.
The examples for addressing a common metric or index are informative.
The Risk Description 4.2 should be moved closer to the front of the document. The
paragraph at the bottom of page 58 should be move to the front of the document. It
succinctly describes the purpose of the framework. The italics should be removed
(Margo Schwab recommendation to remove italics).
Effectiveness of the Methods and Data/Process and Technical
The generic framework provides an acceptable starting document for developing a
cumulative risk assessment model. This framework provides an excellent roadmap for
future formulations. I did not find the Appendices to be very helpful in the current form.
The information in the appendices on methods, data, and R&D should be incorporated
into the body of the document. The resources and contacts could remain in an appendix.
G-54
http://communities.msn.com/CenterforEnvironmentCommerceEnergy
-------
Center for Environment, Commerce & Energy
Conclusion
The peer consultation process is a very important part of developing the
Framework for Cumulative Risk Assessment. It would be great if the final document
would be titled EPA & Stakeholder Framework for Cumulative Risk Assessment instead
of EPA Framework for Cumulative Risk Assessment. Although the process to date has
been effective, scientific consensus will probably suffer when preliminary and final
decisions are made. This could hinder the acceptance of the any final methodology at the
policy level. The framework states "Assessment of cumulative risk through complex
exposures is one of the high priorities of the Agency." It also states that, "The issue of
cumulative risk is also an important issue with the general public." Although scientists
will ultimately conduct epidemiological and toxicological research, the framework for
establishing the methodology for cumulative risk assessment should be an equal
partnership and collaborative effort with stakeholders.
G-55
http://communities.msn.com/CenterforEnvironmentCommerceEnergy
-------
Comments on the EPA Draft Framework for Cumulative risk assessment
Post-Meeting Written Comments From Beth Mileson
Mileson G-56
-------
Comments on the EPA Draft Framework for Cumulative risk assessment
Peer Consultation Comments on the EPA Framework for Cumulative Risk
Assessment
Beth E. Mileson, Ph.D., DABT
Technology Sciences Group, Inc.
1101 17th Street, NW
Washington, DC 20036
T:(202)828-8956
F:(202)872-0745
bmileson@tsgusa.org
Mileson G-57
-------
Comments on the EPA Draft Framework for Cumulative risk assessment
General Comments
The goal of the Framework and the charge to the RAF working group that developed the
document are not entirely clear in the draft document. One of the goals mentioned is to
present the scientific and technical aspects of cumulative risk assessment in the context
of EPA activities without ignoring the real world of total cumulative risks that people and
the environment are subject to. Thus the Framework defines cumulative risk assessment
in the broadest sense possible, generally equating cumulative risk with "total risk" to an
individual or population due to chemical, physical, biologic, and socioeconomic
stressors. Included in the consideration of cumulative risk assessment are risk
assessment of mixtures, integrated risk assessment as described by the SAB, combined
effects of chemical and physical agents, comparative risk assessment, and population-
based risk assessment. This all-inclusive concept of cumulative risk makes developing
focus for a cumulative risk assessment a daunting task that is not necessarily clarified in
the Framework draft.
A second goal of the framework is to help foster a consistent EPA approach for
conducting and evaluating cumulative risk assessments, providing operational definitions
for terms used in cumulative risk assessments and setting out basic principles.
Unfortunately, this document was not developed before a number of EPA Offices were
required to consider cumulative risk and developed their own definitions, principles and
methods for cumulative risk assessment. Thus the authors are in the awkward position of
developing a framework to fit around existing methods, rather than identifying unifying
principles for each Office to use as they develop methods for cumulative risk assessment
to suit their needs.
The framework might be more useful if the definition of cumulative risk assessment was
narrowed to focus on issues related to the EPA activities. Alternatively the broad picture
of cumulative risk assessment could be presented in the introduction and the risks
unrelated to EPA activities (e.g., traffic accidents, earthquakes) could be set aside for
consideration when appropriate, or by an agency other than EPA. In this way, the broad
concept of cumulative risk assessment could be presented and structured using a modular
approach, as discussed at the peer consultation. In either case, the document should
focus on practical guidance for cumulative risk assessment of factors relevant to the EPA
purview. The document does emphasize risk assessment of exposures to chemicals, but
periodically brings up unrelated hazards (e.g., pgs. 34-35).
Key Issues, Terms, and Framework Organization
The discussion of the definition of cumulative risk assessment lends more murkiness than
clarity to the concept, and could be simplified. A short simple definition of the word
"cumulative" would serve as a good start (i.e., increasing or enlarging by successive
addition), followed by adoption of the definition of cumulative risk assessment used by
the Council Cumulative Risk Subcommittee (presented on page 9 of the draft).
Mileson G-58
-------
Comments on the EPA Draft Framework for Cumulative risk assessment
Sections of the document are too generic to be useful in conducting a cumulative risk
assessment and might be modified to account for the unique circumstances of a
cumulative risk assessment. For example the section on stakeholder involvement is
largely composed of excerpts of other EPA documents and could be modified to capture
some of the unique types of input stakeholders may provide for cumulative risk
assessment. Other sections that merely review existing methods used by EPA are
sections on development of Reference Doses, Reference Concentrations and single
chemical carcinogen risk assessment (pages 28-31).
A number of EPA Offices have developed methods for cumulative risk assessment that
meet the regulatory or programmatic needs of the Office. Thus in some areas, the state
of the science in cumulative risk assessment is well developed, while in other areas, there
is little to go on. This is reflected in the draft document in an uneven level of detail, and
in Appendix F, on research and development needs. The treatment of R&D needs is
simplistic. To be useful the level of detail should be increased or references should be
included to provide details.
Stakeholder Involvement
Population- or community- based cumulative risk assessment will benefit from
stakeholder involvement throughout the process, as discussed in the document and at the
peer consultation. Some precautions might be included to make certain that this process
does not delay the assessment significantly. Section 2.3 seems to indicate that all
participants should agree on the assessment plan before proceeding, but in reality it is not
likely that everyone will agree.
Approaches to Cumulative Risk Assessment
The Framework includes sections described as general approaches to cumulative risk
assessment, though in reality the topics covered are tools that can be used as needed for a
given cumulative risk assessment. This was discussed at the peer consultation.
Uncertainty and Variability
Uncertainty and variability should be characterized and considered separately throughout
the process of a cumulative risk assessment.
Combining Chemical and Non-chemical Stressors
Sections 3.2.1.5 and 3.2.2.5 deal in part with the combined effects of stressors that may
interact to produce an adverse effect (such as the cumulative effects of inhalation of
radon gas and tobacco smoke) and in part with disparate stressors. It might be useful to
separate the concepts more clearly in the text.
Mileson G-59
-------
Comments on the EPA Draft Framework for Cumulative risk assessment
Combining Different Types of Risk
It is appropriate to include a discussion of DALYs and QALYs, and these may be useful
in comparative risk assessment, but the utility of these methods in cumulative risk
assessment is questionable. Conversion of adverse effects to a common metric is
important in evaluating cumulative risks due to multiple chemical exposures, but
combining disparate risks is not likely to provide a risk manager with sufficient
information to make decisions. The population's health and well-being are clearly at risk
due to such diverse adverse events as floods, traffic accidents, and exposure to toxic
chemicals, but a single estimate of risk to a population from all these adverse effects
combined does not provide a basis for decision-making unless a sensitivity analysis is
used to attribute risk to each stressor or groups of stressors (as in a comparative risk
assessment).
Mileson G-60
-------
Post-Meeting Written Comments From Stephen Olin
G-61
-------
Stephen S. Olin, Ph.D.
August 31,2001
COMMENTS ON THE DRAFT FRAMEWORK FOR
CUMULATIVE RISK ASSESSMENT
EPA Risk Assessment Forum
Review Draft NCEA-F-1098
Overall, the Cumulative Risk Assessment Technical Panel of the Risk Assessment
Forum is to be congratulated on the major step forward that this document represents.
When the Food Quality Protection Act (FQPA) was enacted in late summer of 1996, some
of us in the risk science field were more than skeptical about the near-term feasibility of
doing credible, data-driven, science-based cumulative risk assessments as called for in the
Act, even given its fairly narrow definition (common mechanism of toxicity) and focus
(mainly pesticides). As expected, it has indeed proven to be a daunting task - but
progress has been made, data are being generated, methodologies are being developed and
debated, and cumulative risk assessments for selected classes of pesticides are moving
forward.
The cumulative risk assessments envisioned in the Risk Assessment Forum's draft
Framework could be a quantum leap beyond those being done under FQPA. The scope
and complexity of the issues that fall within this broader concept of cumulative risk
assessment are enormous and, in many instances, unexplored. Yet it is time to begin to
think about these issues, to ask these complex questions, and to lay out a course that will
encourage and facilitate the dialogue, the research, and the testing of approaches
necessary to move this field forward. The need to understand the cumulative impact of
multiple stressors on public health and the environment demands it. This draft Framework
is a good start.
General Issues
The definition of cumulative risk assessment (p.8) in the draft Framework probably
needs some further discussion and fine-tuning. Actually, the definitions proposed by the
Science Policy Council for cumulative risk and cumulative risk assessment (footnote, p. 9)
capture my view of the appropriate scope and intent of these terms very succinctly. Of
course, there are terms imbedded in the Science Policy Council definitions that also need
to be defined in the Framework glossary (e.g., aggregate exposures), but there is a distinct
advantage in keeping the definitions simple to avoid limiting the scope of cumulative risk
assessment unnecessarily and perhaps inappropriately at this early stage its development.
However, having said that, I have one "limiting" proposal to offer that I think
would help. I suggest that multi-media risk assessments that involve only a single
G-62
-------
Stephen Olin
chemical or stressor be excluded from the term "cumulative risk assessment". Cumulative
risk assessment should imply the consideration of the combined exposures to and effects
of multiple stressors/agents/chemicals. Multi-media risk assessments on individual
chemicals/stressors can be referred to as such, or another term could be adopted. FQPA-
driven risk assessments are referring to single chemical, multi-source/pathway/route
exposures as aggregate exposures (see SPC definitions above) and to the associated risks
as aggregate risks. The advantage to excluding single chemical/stressor assessments from
the term "cumulative" is that it will focus the discussion of cumulative risks on the
important, larger issues that need to be addressed, such as interactions of
chemical/stressor effects, characterization of multi-stressor effects and risks, and the
public health context and implications. Methodologies for conducting single chemical,
multi-media risk assessments are much further along than for multi-stressor assessments,
and it would be beneficial not to give the mistaken impression that the latter is merely a
simple extension of the former by calling them by the same name.
Another smaller point on definitions: the draft document suggests in several places
that cumulative risk assessment must be "population-based" or "place-based". While
these are undoubtedly the broadest categories for application of cumulative risk
assessment approaches, cumulative risk assessments can also be built around a group or
class of chemicals. The current work on pesticides under FQPA is the obvious and
important example.
The draft Framework at least touches on many of the key issues for cumulative
risk assessment. The level of detail in the presentation and discussion of issues is
somewhat uneven; there were a number of specific comments on this during the peer
consultation, and it is recognized that this is still an early draft. It dwells mainly on
chemicals as stressors, which is understandable, but it would be well to expand the scope
to recognize the work that is being done in ecological risk assessment and radiation risk
assessment and microbial risk assessment and the different issues that arise in considering
these risks. These approaches have proceeded along parallel and somewhat independent
paths from the work on chemical cumulative risk assessment and have much to offer to the
discussion. They may not need to be treated in detail in this Framework but, as examples
of other significant classes of stressors and receptors in cumulative risk assessments, they
do warrant separate discussion.
On a related point, EPA has had a number of initiatives in recent years related to
cumulative risk assessment (although often using different terms to describe it). Some of
these are mentioned in the document in passing, but it would be helpful to the reader to
have more information on these initiatives. I would suggest that several of these could be
described in 2-3 page appendices to the Framework. Examples might include the
EPA/OPP work on methods for cumulative risk assessment of pesticides, the OAQPS
work on TRIM, and the Risk Assessment Forum's recent work on risk assessment of
G-63
-------
Stephen Olin
mixtures.
There are some other issues that, I think, need to highlighted or added to the
Framework:
Data Demands - Cumulative risk assessments will identify the need for many
different kinds of data and, often, large quantities of such data. It needs to be clearly
highlighted early in the document that this is likely to be a problem. In fact, it may be that,
for the near term, identification of critical data and research needs will be the primary
function of many cumulative risk assessments.
Variability - There is a section in the draft Framework on uncertainty, but
variability is not discussed in any detail. With multiple stressors, the inherent variability in
the factors affecting risk increases factorially. Temporal variation in combinations of
exposures, spatial exposure variation within a community, and within-population
variations in all of the factors affecting vulnerability are complex problems to address,
even if extensive data sets were available for modeling. Section 3.2.2.2 talks about
"variation of mixtures," but this is such a central issue that it deserves a full discussion and
a clear recognition early in the document.
Interactions - There are some paragraphs in the draft talking about the critical
issue of interactions. What would be helpful, I think, would be to add a few paragraphs
giving the reader a flavor (a taste) of what is actually known about interactions -
chemical/chemical, chemical/microbial, chemical/radiation, etc. There is a developing
literature on this topic that should be acknowledged in the Framework. On the other
hand, it should be made clear that this will continue to be a major research need for
cumulative risk assessment (as suggested in Appendix F). This is not the same as
developing methods for combining different kinds of risk.
PBPK/BBDR Models - In cumulative risk assessments of groups of chemicals, a
key question is, What is the cumulative effective dose to the target tissue, and how can it
be expressed for an individual or a population over time? Since this is the type of
cumulative risk assessment for which we are closest to obtaining quantitative answers, it
would be well to mention the potential value and encourage the development of
physiologically-based pharmacokinetic (or biologically-based dose-response) models that
can handle multiple chemicals simultaneously.
One of the questions asked of the peer consultants was whether some of the topics
presented in the draft Framework are too generic. There certainly are a number of
sections that are written in such a way that they could be lifted from this document and
placed in any other discussion of risk assessment. Some of these sections probably could
be deleted and replaced with an appropriate citation or two, or moved to an appendix;
examples are Sections 3.2.1.1 and 3.2.1.2. However, other sections probably should
remain in the document because, while the processes or issues they describe are applicable
beyond cumulative risk assessment, they are extremely important in population- or
community-based cumulative risk assessments; examples are several of the subsections of
G-64
-------
Stephen Olin
Section 2.1. These latter sections do need to be linked to cumulative risk assessment, and
one approach might be to add a sentence or two at appropriate points in each of these
sections illustrating how the process might play out in such an assessment. It is
recognized that this Framework provides an opportunity to pull together material that
already exists in other EPA guidance documents but has never been laid out systematically
in a single framework.
In fact, the draft document overall suffers from a lack of "for instance" and "for
example." There are many broad statements that are indisputable and may bring to mind
specifics for the cognoscenti but leave the average reader without a point of reference.
More 1-2 sentence follow-ons to these broad statements that would give an example of
what is meant or link the statement to cumulative risk assessment would add significantly
to the readability of the document.
Even more important, it would greatly enhance the impact of the document if the
authors could include in an appendix two or three samples (real or imaginary) of
cumulative risk assessments. Details would not be necessary, and each sample could
probably be limited to a couple of pages. The idea would be to walk through the process
described in the Framework with a specific case, so that the reader could begin to get
some sense of how the Framework might work and its practical utility. My impression
was that there was a good deal of enthusiasm for this suggestion among the peer
consultants.
I do want to acknowledge the evident efforts of the authors to "dejargonize" this
document. I appreciate that one does not need 10 years experience at EPA to
comprehend the ideas being presented and the approaches being discussed.
Another idea on which there seemed to be general agreement among peer
consultants was the value of capturing the Framework in a figure, a process flow diagram
that identifies the key components and generic steps involved in doing a cumulative risk
assessment. Several models for such a diagram exist, and the outline of a diagram for
cumulative risk assessment was drafted by one of the peer consultants during the
consultation. Having a process diagram can help the reader to see how the parts fit
together. It also provides a frame for building more detail into the process, as necessary,
by expanding individual boxes within the diagram or illustrating how cumulative risk
assessments with different objectives might play out differently in the Analysis and
Interpretation stages.
Another question for the peer consultation was whether the Framework
"adequately conveys the state of the science.. .and the areas that are in need of further
research and development." Some suggestions have already been made to improve the
coverage of the state of the science and more will be mentioned below, but overall the
G-65
-------
Stephen Olin
draft document at least raises many of the key issues. It doesn't, and shouldn't attempt to,
provide a comprehensive resource of scientific data or methodologies for cumulative risk
assessment. It might be more complete in its citations of key references. In fact, it cites
EPA documents extensively (which is fine) and is weak on non-EPA documents. For
example, the discussion of interactions and of available methods would be greatly
strengthened by a better coverage of the literature, identifying key papers and good
reference books and reviews. [I provided a few examples during the peer consultation,
but a focused literature search is needed.]
Appendix F is an excellent start on a list of critical areas of research for cumulative
risk assessment. The development of biologically-based dose-response models might be
added to the list, including both the "simple" case of multiple chemicals operating by a
common mechanism and more complex cases.
Process Issues
The questions posed to the peer consultants on process focused primarily on
stakeholder involvement. The draft Framework has a fairly thorough and thoughtful
discussion of process for the Planning and Problem Formulation Phase, but the
descriptions of the Analysis and Interpretation (Risk Characterization) Phases are focused
almost entirely on the technical issues. More attention should be given to process in these
latter phases. For example, the document acknowledges the importance of stakeholder
involvement throughout a cumulative risk assessment but barely mentions stakeholders in
the Analysis and Interpretation Phases. There is a good point made (in Section 2.3) about
discussing possible outcomes of the risk assessment during the Problem Formulation
Phase and revisiting those possible outcomes as a group (including all stakeholders) during
the Interpretation Phase. However, the implication in Section 2.3 that unanimity must be
reached among stakeholders on the plan for the risk assessment before proceeding, while
certainly a goal, seems too restrictive; there may be cases where the risk assessment must
proceed to meet the needs of most of the stakeholders, even if some disagreement remains
of the scope or approach.
The document could benefit from further discussion of who are potential
stakeholders (perhaps with some examples of how stakeholders may differ for different
kinds of cumulative risk assessments). The table on p. 17 seems to me to be deficient in
not clearly identifying "the affected industry" as a perennial stakeholder; "business
owners" and "trade associations" are too vague.
Regarding the role of stakeholders in the later phases of the risk assessment, some
ideas that could be considered, depending on the scope and approach agreed upon in
Problem Formulation, are the following. In the Analytic Phase, stakeholders can: (1)
suggest sources of data or provide data for the risk assessment; (2) help clarify issues
identified during Problem Formulation; (3) work alongside the risk assessment expert to
G-66
-------
Stephen Olin
see what data/assumptions are being used and why, and to understand better how the risk
assessment process works; (4) suggest alternate scenarios that may reflect more realistic
exposure conditions in the community. In the Interpretation Phase, stakeholders should:
(1) understand the outcome of the cumulative risk assessment; (2) ask questions about
how best to frame the interpretation; (3) confirm that the cumulative risk assessment has
met the goals set in Problem Formulation, or if not, why not; (4) participate in the
discussion of next steps and actions to be taken.
When should cumulative risk assessments be done? Recognizing that the scope
and nature of a cumulative risk assessment may range from a very limited qualitative
assessment of a local situation to a comprehensive assessment of the cumulative risk
patterns for a large community, the simple answer is whenever the combined impact of
multiple stressors needs to be considered. Only experience with these assessments over a
period of time will provide the wisdom needed to develop practical guidance on this
question.
Technical Issues
I have only a few additional comments on the technical issues discussed in the draft
Framework.
The use of biomarkers and biomonitoring in cumulative risk assessment is a good
idea, but its application is likely to be limited to some specific cases for the near future.
For example, information on the cumulative risks in a local population of a group of
chemicals that are toxic to the liver might be provided by selective liver function tests, but
causal inferences would have to take account of many other factors that may affect liver
function. For more disparate stressors, we will have to learn how to evaluate and interpret
broader measures of exposure and effect, a major challenge for cumulative risk
assessment. Also, as noted in Section 3.1.3 of the draft document, the use of biomarkers
as surrogates can lead to some loss of specificity in information.
The discussions in the document of approaches to cumulative risk assessment
probably could benefit from further input from the biostatistics community. There has
been some work done in recent years on the cumulative risk question and what is and is
not appropriate methodology, and some acknowledgment of the potential pitfalls and
problems would be helpful in the document.
The discussion of uncertainty needs to be carried through all three phases of
cumulative risk assessment, and the Framework should include a few sentences in each
Phase on how that might play out in practice. In addition, a brief discussion of different
approaches to uncertainty analysis for cumulative risk assessment would add substance to
the section already in the document. Also, as noted above, the significance of variability
and its consideration at each Phase in the assessment should be discussed.
G-67
-------
Stephen Olin
Vulnerability is, I think, a useful concept for cumulative risk assessment. Some
further examples/discussion of its application in cumulative risk assessment would be
helpful.
Two final thoughts:
• Cumulative risk assessment is not only pushing the boundaries of what we are
capable of doing technically in a risk assessment; it also is pushing the
boundaries of risk assessment into what some have called risk analysis, a
domain that begins to consider issues beyond the reach of science, issues of
policy and societal values. These are important issues in public health decision
making. But we should think about the potential effects of blurring the
boundaries. Do we really want to call the broadest applications of the process
described in this Framework "risk assessment"?
• A related cautionary reminder is that the planning, conduct, and characterization
of a cumulative risk assessment must be perfectly transparent. The process,
methodology, data, assumptions, and selection among alternate interpretations
must be very carefully documented and very clearly stated in the report of a
cumulative risk assessment. The complexity of the problem and the uncharted
territory of this new land will demand the highest level of clarity, integrity, and
openness if we are to reach our goal.
G-68
-------
Post-Meeting Written Comments From Bill Rhyne
G-69
-------
Bill Rhyne
Comments by W.R. Rhyne on the
Draft Framework for Cumulative Risk Assessment
Comment on whether the Framework adequately captures, describes, and reasonably
organizes the key issues for cumulative risk assessment? Include in this discussion whether
key terms have been identified and clearly defined.
The concept of a "stressor" is introduced on page 1, but the term is not described until page 17
and much clearer on page 34. A glossary will be added to the document, but I suggest that this
term deserves early, clear definition in the text. A discussion similar to that on page 34 is
recommended for page 1, particularly the range of physical hazards. Before I got to page 17, it
appeared that a stressor was limited to pollutants.
A statement is made on page 58 that the document devotes considerable time to the multiple
chemical aspect. I suggest that this statement be introduced much earlier, e.g., in Section 1.1,
including the statement that methods and data are not uniformly developed at the present time.
"Hazard Index" is mentioned on pages 25, 44, 45 and 52 but is not described.
What additional issues, if any, should be covered?
No comments in addition to those in several places below about physical and other stressors that
produce the immediate effects of injury and fatality.
Keeping in mind that the Framework is intended to address issues that are specific or
unique to cumulative risk assessment, comment on whether some of the topics presented in
the draft Framework are too generic for the intended purpose?
No comments in addition to those in several places below about physical and other stressors that
produce the immediate effects of injury and fatality.
The Framework attempts to identify where methods and data are currently lacking or are
underdeveloped for application to cumulative risk assessment. Based on a suggestion
during the meeting with other Federal scientists, research oriented issues have been
highlighted in an appendix to the Framework. Comment on whether the Framework
adequately conveys the state of the science with respect to cumulative risk assessment and
the areas that are in need of further research and development.
G-70
-------
Bill Rhyne
Appendix F is a top-level discussion of the shortcomings in the science. It is somewhat
abbreviated compared to the level of detail in Section 4.1, but seems to serve its intended
purpose.
Process Issues
Stakeholder Involvement Throughout the Cumulative Risk Assessment
Process [Framework Section 2; especially 2.1.3 and 2.3] Stakeholder
involvement is recognized as an important aspect of cumulative risk
assessment. The draft Framework highlights this involvement in the
Planning and Problem Formulation Phase of the assessment. Examples of
interested and affected parties are provided and the importance of defining
roles and responsibilities is discussed. In addition, the importance of
discussing the possible outcomes of the assessment with stakeholders is
emphasized. What suggestions do you have for engaging stakeholders early?
What kind of background or ground rules would be helpful? What roles and
responsibilities would be most appropriate for communities and interested
parties?
The stakeholder section is quite good but can be improved as suggested by panel
members. However we should recognize that it is difficult to force meaningful
stakeholder participation on an agency. If an agency wants it, they can and will
seek it. Conversely, if any agency wants to avoid it, then there is nothing that can
be added to this section of the document to force it. This document can heighten
the need for such involvement.
A role/responsibility that can be very helpful to meaningful stakeholder
involvement is to form subgroups of stakeholders to concentrate on various
specific issues in order to (1) understand the technical information, (2) elevate and
clarify the stakeholder issues as needed, (3) and convey the facts to their peers.
The third item is the key, and this role/responsibility works best in situations in
which the stakeholders are reacting to a lack of information and/or a belief that
their issues are not being considered.
My experience with stakeholder groups runs from excellent (all the positive
attributes discussed by the panel) to nonexistent stakeholder involvement. In the
most positive case, the stakeholders are very diverse and spread across the U.S.
The issue is not where facilities will be located, but what type of facility will it be.
Approximately 80 representatives of a like number of stakeholder groups met and
chose five of their members to attend technical meetings of the Army, the
G-71
-------
Bill Rhyne
competing contractors, and the National Research Council. The roles and
responsibilities described above seem to be working to everyone's satisfaction.
As noted on page 18, technical assistance is being provided. The stakeholders have
their own contractor, who as time passes is needed less and less. Substantial travel
funds are also provided.
The assistance of an independent review group is invaluable. In my experience the
citizens never fail to be impressed that the National Research Council committee is
composed of unpaid volunteers, i.e., not contractors paid to support a specific
position.
Using the Results of Cumulative Risk Assessment [Framework Section 4.5]
The intended use of the cumulative risk assessment should be considered at
each step of the process, from Planning and Problem Formulation through
completion of the assessment. It is important to recognize that the
cumulative risk assessment provides important information but is not the
only contributor to the decision making process. How the cumulative risk
assessment fits into the decision making process should be addressed when
discussing possible outcomes with stakeholders during the planning phase.
What kind of information could stakeholders contribute to cumulative risk
assessments? Under what circumstances do you feel stakeholder
participation is most effective?
See the above comment.
Cumulative Risk Assessment in the Broader Context of Environmental
Analysis and Decision Making [Framework Sections 1.4 and 1.5 and Section
4; especially 4.4.2 and 4.5]
An issue related to using the results of cumulative risk assessment is
understanding how these assessments fit, both conceptually and practicably,
into broader contexts. For example, cumulative risk assessment may be a
component of assessing cumulative effects under the National Environmental
Policy Act (NEPA). Further, these assessment may support Community-
Based Environmental Protection efforts. What are the most appropriate
situations for conducting cumulative risk assessments?
An example of the need for a community-based assessment in my community is the
addition of another chemical/radionuclide processing facility, although perhaps
G-72
-------
Bill Rhyne
small, to an industrial park. NEPA does not apply in this case, so cumulative
effects will not likely be addressed by any group
Technical Issues
Approaches to Cumulative Risk Assessment [Framework Section 3, especially
introduction]
There are a number of ways to approach cumulative risk assessment, either
starting with the NRC paradigm or using a different approach. Each
approach may present its own challenges in methods, data, and analysis.
Four example approaches are discussed in the draft Framework for
Cumulative Risk Assessment: combining toxicity before calculating risk,
calculating risk factors, using biomarkers and biomonitoring, and calculating
other types of probability statements. Are important approaches missing
from this list? Comment on whether there is a better way to organize this
discussion on approaches to cumulative risk assessment.
The above list omits "calculating risks for individual stressors and then combining
them" that is discussed on page 24.
I found Section 3 to be a little difficult to digest. Section 3.2 is 17 pages long and
the three parallel sections (3.3, 3.4 and 3.5) are to be provided later. Some of the
material in Section 3.2 is likely to apply to the missing sections, and it is likely that
Section 3 will have to be recast for this reason alone. At a minimum, Section 3.2
needs a "tell 'um what you are going to tell 'um" introduction.
"Other" is too broad for the title to Section 3.1.4. "Reduction in Life Expectancy
or Quality" would be better.
The penultimate sentence of Section 3.1.4 is confusing. It is not clear whether
"this" refers to QALY or to both LLE and QALY. Also the "second approach"
should be clarified to be the second of two approaches in Section 3.1.1 not the
second general approach presented in Section 3.1.2.
Using Biomarkers of Exposure or Effect [Framework Section 3.1.3]
Among the approaches to cumulative risk assessment discussed in the draft
Framework, use of biomarkers and biomonitoring information shows promise
for providing an integrated measure of where an individual falls on the
continuum between exposure and effect. However, existing methods are
G-73
-------
Bill Rhyne
inadequate for assessing complex situations involving a large number of
stressors. Further, experience using biomarkers and related information in
decision-making is limited. Comment on the discussion of biomarkers of
exposure and effects? Given the current state of the science, discuss the
utility of this approach in the near term and in the future. Does the
Framework include the appropriate level of discussion of this topic? How can
the discussion be improved (consider, for example, whether the Framework
discussion adequately characterizes how biomarkers can be used with more
traditional risk estimates)?
No comment.
Uncertainty in Cumulative Risk Assessments [Framework Section 4.3]
Uncertainty analysis is an important aspect of risk assessment (and policy
analysis in general). However, historically, dealing with uncertainty has been
a short-coming of many assessments. Cumulative risk assessments present
new challenges for uncertainty analysis. For example, assessing cumulative
risks will involve combining data of varying quality. Perhaps more
important, assessing cumulative risks will involve the use of "soft"
assumptions. These are assumptions which may have a high degree of
uncertainty that is difficult (or not possible) to quantify. Comment on
whether the Framework adequately characterizes the importance of
uncertainty analysis in cumulative risk assessment. What additional
discussions of uncertainty should be included in the Framework (and in what
sections of the document)?
I thought this section was a very good overview and contained about the right
amount of detail. One way to improve it would be to give examples of each end of
the uncertainty spectrum: (1) large but unknown uncertainty and (2) the ability to
express various uncertainties as mathematical distributions.
Vulnerability [Framework Section 3.2.1.6]
As applied to cumulative risk assessment, it is useful to think of four
components to vulnerability: the susceptibility or sensitivity of the human or
ecological receptors; the differential exposures of the receptors; the
differential preparedness of the receptor to withstand the insult from
exposure; and the differential ability to recover from the effects. The issue
for cumulative risk assessment is how to consider these aspects of
vulnerability and their potential impacts on risk. Comment on the discussion
G-74
-------
Bill Rhyne
of vulnerability in the draft Framework. Has the state of the science been
captured in this discussion? How can the discussion of this issue be
improved?
This is not my area of expertise, but the discussion seemed reasonable.
Combining Chemical and non-Chemical Stressors [Framework Sections
3.2.1.5 and 3.2.2.5]
Viewing cumulative risk assessment as an evaluation of the accumulation of
stressors presents many challenges. These may be seen when attempting to
combine, in some meaningful way, the risks from multiple chemicals that
may act as synergistic, antagonistic, or additive doses leading to a single
effect. The situation is exacerbated when non-chemical stressors (e.g.,
radiation, biological agents, and psychological stress) are considered.
Comment on the Framework's discussion concerning the combining of
disparate environmental stressors. In commenting, consider the state of the
science with respect to understanding the effects of different stressors acting
together (e.g., chemical exposure and viral infection). What can be added to
the Framework to adequately convey the state of the science in this area?
These sections address a complex problem. Radiological effects can be, but are
only rarely, large enough to be immediate, rather it takes years for cancer to
develop. In this sense radiological risk is similar to chemical risk.
Physical hazards, on the other hand, are of concern due to their immediate effects:
injury and fatality. Physical hazards include mechanical forces from accidents (both
transportation and processing accidents) and the forces due to fireballs, explosions,
and so forth. Chemical spills directly on the skin and airborne exposure greater
than IDLH (immediately dangerous to life or health) are additional stressors that
produce an immediate effect. The exposure can be occupational and public. I
recommend providing the reader with a better understanding of this type of "other
" stressor. A discussion similar to that on pages 48 and 49 for quality-of-life
stressors is needed.
Analysis of immediate effects is relatively straightforward, whereas, as noted in
Section 3.2.2.5, stress and quality-of-life issues are not straightforward to evaluate.
Appendix A is referenced in Section 3.2.2.5 as further reading relevant to
determining exposures to non-chemical stressors. The last paragraph mentions the
increased likelihood of vehicular accidents, but it is not clear whether the context is
G-75
-------
Bill Rhyne
transportation-induced chemical spills or the occurrence of physical stressors. It
should be clarified.
Appendix A.2 addresses resources relevant to exposures to non-chemical stressors
and is incomplete, i.e., no mention is made of physical stressors that result from
transportation or process accidents. Data is generally available to address these
stressors. References for process safety and risk could be added from Appendix D
of 29 CFR 1910.119, Process Safety Management of Highly Hazardous
Chemicals.
Transportation accidents may produce multiple stressors: small spills, large spills,
and physical. The physical stressors can result in property damage, injuries, and/or
fatalities. (Note that a spill can produce a physical stressor, e.g., an LNG spill can
result in a fireball producing heat and mechanical forces.) To put these stressors in
context, the conditional probabilities, given a heavy truck accident, are
approximately: 0.2 for an injury, 0.05 for a small spill, 0.02 for a large spill, and
0.01 for a fatality. The truck injury and fatality data are good on a global large
truck basis, but the spill values are soft (many packaging variables). The point is
that transportation physical stressors can produce (1) an immediate end point, an
injury, that is about the same probability as a small spill and (2) an immediate end
point, a fatality, that is about the same probability as a large spill.
Process accidents can also produce small spills, large spills, and physical stressors;
see 29 CFR 1910.119.
Combining Different Types of Risk [Framework Section 4.1.3]
Conveying the combined risks from multiple chemical and non-chemical
stressors, in a meaningful way, is the ultimate challenge for cumulative risk
assessment. Experience in this area is extremely limited. Indices, common
metrics (e.g., Disability Adjusted Life Years - DALYs) and graphical (e.g.,
GIS) approaches have been explored but much methods development work
remains to be completed. Cumulative risk assessment can be a valuable part
of the decision making process, but only if the results are conveyed in a
meaningful way. Comment on the Framework's discussion concerning the
combining of disparate measures of risk. Do the example approaches
discussed in this section capture the state of the science in the area? In
particular, consider the role of valuation (i.e., the assignment of societal
values to disparate health outcomes) implicit in some of the approaches.
Suggest changes or additions that may improve this discussion.
G-76
-------
Bill Rhyne
The problems are stated at a level appropriate for this top level document. Some
details on the indices in Section 4.1.3.2 would be helpful, but references are given
for the curious.
Limits on the Information Provided by Cumulative Risk Assessment
[Framework Section 2.3]
There are many factors that place limits on what can be learned from a
cumulative risk assessment. These include: how the problem is defined
during the Planning and Problem Formulation phase, what data are
available for the assessment, what models are available for the assessment,
and what resources are available for the assessment (e.g., expertise, time,
financial support, etc.). The adequacy of existing data may be a key factor
limiting the results of a cumulative risk assessment. One outcome of the
assessment may be the identification and prioritization of data gaps and
research needs. This also applies to the availability of models for the
assessment. Comment on the Framework's attempt to characterize the limits
that may be encountered when using cumulative risk assessment in the
decision making process. What can be added to improve this discussion?
No comment.
Distinguishing between Cumulative Risk and Cumulative Impacts
[Framework Sections 4.4 and 4.5]
The Framework attempts to clarify how cumulative risk assessment relates to
community assessments and decision making. In particular, the Framework
emphasizes that cumulative risk assessment is only a portion of the
information needed for decisions. Other factors such as employment and
quality of life are considered for many decisions. When these other factors
are brought into the assessment, the approach is often called cumulative
impact assessment. Comment on how well the framework distinguishes
between cumulative risks and cumulative impacts. Are the two concepts well
characterized? How can the discussion be improved to clarify the role of
cumulative risk assessment in the assessment of cumulative impacts?
This section is quite good, but it does not discuss cumulative impacts - only
cumulative risk.
G-77
-------
Post-Meeting Written Comments From Jennifer Sass
G-78
-------
Jennifer Sass
Peer Consultation Workshop ~ Review Draft Framework for Cumulative Risk Assessment
(NCEA-F-1098; August 2, 2001)
Meeting held August 22, 23 2001.
Arlington, VA
Comments by: Jennifer Sass
Natural Resources Defense Council
1200 New York Avenue, NW
Washington, DC
Tel: 202-289-2362
Email: jsass@nrdc.org
Submitted to: David Bottimore
Versar
Bottimoredavid@aol. com
EPA deserves much credit for tackling this tremendous task, and as a draft document, the Framework
is an extremely important initial step. The Framework is a very inclusive document, with an
appropriately large amount of flexibility which will necessarily give it great utility in its applications.
It draws from a number of previous documents, and provides a useful overview of the historical
development of cumulative risk analysis. I commend the writers of this document for both tackling
such a grandiose task, and for including peer review input so early in the process.
I felt that the Peer Consultation Workshop was extremely beneficial, and provided expertise drawing
from technical, theoretical, and practical knowledge. Surprisingly, there was much consensus during
the two day discussion, and many concrete suggestions for improvement. I feel confident that with
the incorporation of the panel suggestions, and further work by EPA, the document will take on the
shape of a Framework. I thank the EPA for including me in the consultation process, and look
forward to future versions of the Framework.
/. Overarching Issues
The document needs to provide a framework that is clear. As it stands, the document provides a
review of existing risk methodologies, but does not lay out a clear framework which is specific to
cumulative risk assessment, and tangible. A schematic (flowchart, matrix, etc.) would be extremely
helpful, possibly with some examples which might help the reader to grasp the utility of the
framework. A framework must be something that both frames the methodology, and can be worked
with.
The framework, when defined and presented, should be able to accommodate large and small scale
cumulative risk assessments, such as community projects, chemical family assessments, etc. Within
G-79
-------
Jennifer Sass
the framework, each step could include both the question, the criteria which would constitute an
answer to the question, and some examples of answers.
Variability and uncertainty should be defined at the beginning of the framework, and then carried
throughout the framework, with the goal of providing a risk assessment which considers the
protection of the population at all stages.
Given the Agency's desire to move towards a pro-active approach to Cumulative Risk Assessment,
it would be prudent to establish a database or National registry in which information on morbidity
and mortality were cross-referenced with information on occupation. This would require all death
certificates to list primary occupation as standard information. After all, science has identified almost
every known carcinogen from the death certificates of workers. This information, while so obviously
important in identifying risk factors leading to disease and death, is noticeably absent from all cancer
and occupation registries.
//. The Process
It is important to include stakeholders, particularly community groups, environmental activists, and
public health professionals, early in the process of Cumulative Risk Assessment, to help define the
problem, and to help identify stressors, sources, pathways, etc. However, while including
stakeholders throughout the risk assessment process, it is important that EPA maintain discretionary
power regarding the incorporation, if at all, of stakeholder input. Without this caveat, there is a
concern that powerful stakeholders, with financial interests in the outcomes of the risk assessment,
would have the potential to bias the assessment outcome towards profit goals rather than health or
environmental protective goals.
It is important to provide a "level playing field" whereby all stakeholders have equal access to the
resources required to play a meaningful role in the risk assessment. These resources would include
access to software, technical expertise, consultants, etc.
The risk characterization step would be better placed early in the process, to ensure that the desired
outcome of the assessment is identified early, and then the analysis process is designed/chosen to
obtain the desired outcomes. The risk characterization process needs to focus on a given question,
for a given application, early in the cumulative risk assessment. To this end, the time frame
(deadlines, etc.) and criteria for acceptable outcome decisions should also be defined, to the level that
this is possible. This would allow the cumulative risk assessment to be useful as a planning tool
consistent with the EPA goal of moving towards more pro-active community planning.
///. Technical issues
Some discussion regarding methods for harmonizing cancer and non-cancer approaches is needed.
Some discussion regarding the use of epidemiology for use with analysis is needed. Ecological
epidemiology might be particularly useful in providing estimates of population risks, and should be
discussed. The use of ecological data for decision-making should be discussed. This may include
its use in identifying and quantifying background exposure levels, from all sources, within a
population.
G-80
-------
Jennifer Sass
Much discussion, and examples, regarding methods for assessing qualitative stressors (nutritional
status, access to health care, etc.) is needed. Examples from European methods could be discussed.
Methods and approaches for the inclusion of value judgments and qualitative data should be a key
component of the Framework.
The importance of background or baseline exposure data should be more explicitly discussed,
including methods for assessing it, and for incorporating this information into a Cumulative Risk
Assessment Framework.
The Framework should include methodology and approaches for identifying uncertainty. This could
be included within the schematic Framework, such that the limitations of the methods are identified
initially, and then, given the limitations, what can be learned, and what cannot, with the given
methods.
An expanded discussion of biomarkers is warranted, given their increasing usefulness. Their use
early in the risk assessment process, to obtain a "basic health indicator" of the actual cumulative
assault on a person, regardless of source, warrants discussion. Biomarkers may be used for
surveillance and biomonitoring on a chronic basis, and can help define background exposures.
Likewise, their current limitations deserves discussion.
G-81
-------
Post-Meeting Written Comments From Margo Schwab
G-82
-------
Margo Schwab
Peer Consultation on Technical Issues Associated with Developing
a Framework for Cumulative Risk Assessment
Margo Schwab's Comments, Ideas, Vision for Document
I. Over Arching Issues
A. Title
In light of the extended process for achieving this framework, both from a
process and from a technical perspective, consider changing the title to "An
Evolving Framework for Cumulative Risk Assessment."
B. Goals
The goal(s) of the document should be stated clearly in the first section (see
suggested reorganization below), and then reiterated in the concluding section
of the document. As it currently stands, a variety of different goals are stated
throughout the report. The conflict among the goals is confusing, and is
reflected in the changing tone of the text in throughout the document. Keep
the goals narrow.
C. The Framework
Currently it is not clear what the "framework" is. See comments under
"proposed reorganization."
Proposed Reorganization
1. Introduction.
i. Begin by defining what cumulative risk assessment is, rather
than with history [the current beginning focus on the evolution
of risk through various EPA documents is distracting and not
useful until the concept is defined clearly.] Instead, focus
defining cumulative risk assessment as a tool to inform the
decision-making process. Discuss what is special about it: it is
designed to take into account a broad variety of stressors. See
proposed Figures 1 and 2 as aids in laying out the array of
stressors that are of concern and the array of types of
G-83
-------
Margo Schwab
cumulative risk assessments that might be conducted, from
narrow to broad. In defining the concept of cumulative risk
assessment, clearly state its goals and limitations (in a general
way). The current discussion attempts to make cumulative risk
assessment "all things to all people" and to answer to too many
of our problems.
ii. Next, set the need for this framework in its policy and
historical context. Set up the framework as the vision for the
future. The movement in society toward a more inclusive
(holistic) approach to defining problems, in part as a result of
our realization that public values are as important as science; in
part due to our improved scientific understanding of multiple
routes of exposures, effects at low levels of exposure, and
common mechanisms of action as well as understanding that
many diseases are multifactorial, thus a multi-risk factor
approach is needed. Talk about our increasing understanding of
the importance of stress in the body's ability to fight disease,
thus the importance of including multiple stressors.
It is here that the various EPA and NAS documents might be
sited as evidence of the evolution in both science and process.
It is here that the 1983 NAS Red Book paradigm should be
introduced as being pivotal to defining the risk assessment
process until now, but that conditions have changed, and the
needs (stated here) require an evolving approach. In essence,
emphasize that the cumulative risk framework is charting the
future of risk assessment. Introduce the idea that the goal of
setting up a "framework" is to show how the concept of
cumulative risk assessment can be applied in a wide variety of
contexts (give a few examples, like FQPA and superfund sites,
as well as community driven concerns).
iii. Next state the goal(s) of the document. Keep the goal
narrowly focused on setting up a framework. Again, highlight
what the document will cover (how one approaches a
cumulative risk assessment) and will not cover (e.g.
implementation/protocols/decision making frameworks).
iv. Present the framework. What is referred to as "The
Framework" can be presented as having three levels of detail:
The first level is depicted in Figures 1 and 2 - the concept
that multiple stressors define the context of a problem. .
Give examples of classes of stressors: chemical, biological,
G-84
-------
Margo Schwab
physical, economic, social/political, cultural, etc. Introduce
the concept of vulnerability (Kasperson's four types).
The second level is depicted in the three steps: problem
formulation, analysis, and interpretation/characterization
(proposed Figure 3). Outline the key purpose of each of
the three steps.
The third level depicts what is included in each one of
these steps with respect to decisions that must be make and
potential stakeholder roles (proposed Figure 4)
In this subsection, discuss why each component of the framework is
important. Set this up as a guide for the next three sections of the
report.
2. Problem Formulation Phase
Walk through this section as a series of decisions that must be made, the
purpose of those decisions, examples of decisions, and the implications of
those decisions for the subsequent steps in the process. The current
discussion of the importance of addressing the desired outcome during this
phase is essential, and should be highlighted earlier. Specifically, it is
important to emphasize that the purpose of the analysis determines the
type of characterization that is appropriate, which in turn influences the
type analysis approach chosen.
The current document talks about defining a conceptual model - this is
very good; keep it and give some graphic examples of such models.
The current focus on process is good, but is sometimes distracting.
Perhaps tie it all together by beginning the section by saying that there are
three sets of issues that will be covered:
i. the decisions to be make.
e.g., the definition of the problem, the goals of the analysis,
which stressors to included, the time frame for the entire
process, etc. Most of these are included in the current
discussion, however they get lost in too much text and not
enough "to do" lists.
ii. the process for going about making the decisions.
e.g., discuss the definition and importance of stakeholder
involvement and how to make determination regarding
who it is appropriate to include in this situation, the role of
G-85
-------
Margo Schwab
the stakeholders, and the need to provide technical and/or
financial resources to allow this process (see comments
below under "process issues").
iii. the criteria that should be used in making those decisions
e.g., level of uncertainty that is acceptable, implications of
vulnerability and variability, the state of science, how the
results will be used, etc.
I'd like to see the detailed discussions of vulnerability and
uncertainty that appear in the "Analysis" section of the
current document moved up to this section. It should be
made clear that both of these concepts/concerns must be
addressed in this first phase, and their significance carried
throughout subsequent phases.
3. Analysis Phase
This section is currently twice as long as it needs to be. Keep the focus on
the approaches available and their strengths and weaknesses of each .
Discuss how these approaches advance the more traditional one-pollutant
and one-stressor paradigm. Stay away from equations and jargon. See
detailed comments under "technical issues."
4. Characterization and Interpretation Phase
This section should be shortened to about one-third of its current length.
It is highly redundant with other parts of the document, and contains too
much speculation. Rather, tie it strongly the problem formulation phase -
bringing the significance of the decisions made full circle. The specifics
should reflect the organization of the "approach" section. If there is little
technical detail/protocol in the analysis section (as I recommend
above/below), then much of the technical detail can be removed from this
section.
E. Level of Detail
As commented upon by many of the reviewers, the level of detail is
uneven. Suggestions include:
1) Remove much of the detail from Section 3.2 (see
comments under technical issues).
2) Throughout the document, each of the issues discussed
in "generic" terms (i.e., not specifically limited to
G-86
-------
Margo Schwab
cumulative risk assessment), should be discussed in
terms of how the issue is particularly relevant to
cumulative risk assessment. An example here is with
"stakeholder involvement" (see comments under
process issues).
3) Finally, it is important throughout the document to use
examples to illustrate each of the points made. You
might choose to use two examples each time - one for a
limited scope assessment (e.g., an FQPA type
assessment) and one from a broad-based assessment
that includes many types of stressors. This would show
the range of applicability of the concepts. There should
be few concepts in this document that are limited to
only one end of this spectrum or the other.
F. State of Science
The state of science in unevenly documented. Areas that need
improvement include:
1) Most of the science documented is quantitative. Little
emphasis is put on the value of qualitative methods or the
approaches available.
2) The value of community-based health studies and the role
of epidemiology in general is missing.
3) Our recent understanding of how non-chemical stressors
influence our ability to fight disease, thus the dose-
response curve is not highlighted.
G-87
-------
Margo Schwab
II. Process Issues
A. Stakeholder Involvement
The baseline information on stakeholder involvement is appropriate - i.e., the
process is set out in a manner consistent with the Presidential/Congressional
Commission Report (1997). However, I encourage you to articulate more
clearly the particular importance of stakeholder involvement within the
context of "the cumulative risk assessment framework." Specifically, that
without stakeholder involvement in determining the problem that the
assessment will address; which stressors to include; the level of uncertainty
considered acceptable; the type of results (quantitative versus qualitative); and
the schedule for completion of the assessment, the "cumulative risk
assessment framework" is not only inadequate, it is destined for failure.
Given that this is a framework intended to provide EPA with guidance, the
extent and limitations of stakeholder involvement should be specifically
addressed. The issues identified below can help guide a positive involvement
process rather than simply insuring involvement:
Who is Involved: The key issues associated with determining who is a
stakeholder should be discussed: give examples of cumulative risk
assessments in which the stakeholders do not include the "general public,"
for instance permitting processes, and other examples in which the
stakeholders include multiple, competing community groups, multiple
government agencies and levels of jurisdiction, etc. The point here is that
the breadth of stakeholders invited to the table should be tailored to suit
the situation.
It might also be worthwhile to provide a little guidance about which
individuals are invited - that in order for the process to be productive, it
is necessary to choose those who have a genuine interest in solving the
problem and who are willing to engage in a dialogue.
Extent of Involvement: The trade offs between extent of involvement
and delay in producing information should be discussed. Examples
should be given to highlight situations in which there is an urgency for a
decision, and thus extensive involvement can hinder the process and, at
the other end of the spectrum, situations in which the schedule is more
flexible and the stakeholders have the technical knowledge and
understanding to participate at a very deep level.
The decisions which are most important for stakeholder involvement
should be identified, and examples given as to why this involvement is
essential.
G-88
-------
Margo Schwab
The extent of involvement possible by community groups and the general
public will depend upon their technical and financial resources. If a level
playing field is sought, the Agency must provide a mechanism to support
and/or facilitate contributions from groups without the necessary
resources.
It should be explicitly stated that EPA is inviting stakeholders to provide
input, but that the final decisions regarding each step in the process are the
purview of the Agency.
B. Using the Results of Cumulative Risk Assessment
I agree with the gist of the discussion on pages 60-61 - that if the problem
must be formulated to address a specific question; if this has been done, the
results will provide guidance in addressing that problem. Currently, however,
the section is not succinct; it is redundant. Since this section also serves as
the conclusion to the report, it might be set up as Section 5 (rather than the
last subsection of Section 4). As such, the format would be to link back into
the points discussed in the first chapter, and the schematics used therein. The
text currently does this - I a suggesting making the link even stronger and
clearer. Similarly, the point about the "use of the results" should be made in
Section 1, thereby ending the "framework" by going back to the "scope" set
out at the beginning of the document.
C. Cumulative Risk Assessment in the Broader Context of
Environmental Analysis and Decision Making.
See discussion under "overarching issues."
D. Add A Section:
Discuss the resources that the Agency must develop to facilitate the
implementation of this framework, including expertise in epidemiology and
social science, as well as collaboration with Federal and State agencies. Also
discuss the need for a targeted research agenda to address specific questions
raised in this document (gaps).
G-89
-------
Margo Schwab
III. Technical Issues
A. Approaches to Cumulative Risk Assessment
Section 3, the description of the "analysis phase," might be reorganized as
follows:
Begin with a one-page discussion of how the traditional "red book" paradigm
is used to "analyze" risk [note: per my comments under organization in the
"over arching issues section, the paradigm would have originally be
introduced under the "historical context" subsection of Section 1.] Opposite
this one-page description of how one analyzes risk per the red book paradigm
should be two figures: one of the "four step" process, and one that lists the
key activities and decisions associated with each of those four steps. The text
would focus on describing these two figures. Concepts such as dose-
response, extrapolation, exposure, pathways and sources would be introduced
and defined in this document, but actual methods such as the RfD, the linear
no-threshold model, BMD, etc would not be introduced or defined in this
document. This page would end with a description of the limitations of this
analysis approach, both for combining multiple chemical and for combining
multiple stressors. These limitations would highlight the state of science
regarding mechanism of action as well as combining chemical and non
chemical stressors, etc. A brief introduction to the issues listed in Section 3.2
can be included here (e.g., temporal trends). Reiterate issues of uncertainly
here - that the traditional red book paradigm is full of uncertainty; it is
compounded each time another chemical and stressor is added.
The next section would begin with general approaches to the shortcomings of the
traditional approach to risk assessment.
1) Qualitative Approaches
A variety of qualitative approaches are available. Below I note two; a
comprehensive review of the social ecology literature is likely to turn up
more such indicators. Also see the Health and Environment Analysis for
Decision Making project of the WHO/UNEP, discussed in "Linkage
methods for environmental health analysis" WHO/ENG/95.26 ed by D.
Briggs, C. Corvalan, MNurminen. 1996.
The attached adaptation of Figure 2 from Sexton et al 1992 (see ref
below) shows the conceptual paradigm underlying the prioritization of
environmental health problems. The highest ranking problems are those
with elevated exposures and very toxic pollutants; conversely, the lowest
rankings are for contaminants with low exposures and low toxicities. This
matrix approach can be expanded to include any number of additional
G-90
-------
Margo Schwab
stressors, including underlying health status (e.g., based upon cancer
statistics), demographic characteristics that influence sensitivity and
susceptibility (e.g. % elderly), economic status, etc.
Look at the entire issue of Archives of Environmental Health from
November/December 1992; volume 47, No. 6. Especially: Sexton et al
"Estimating Human Exposures to Environmental Pollutants: Availability
and Utility of Existing Databases, pg 398 and Burke et al., Role of
Exposure Databases in Risk Management, pg 481
The qualitative indicators provided by such matrices can be
communicated using geographic information systems. Displaying
complex multi-dimensional matrices in a map can help visualize locations
of areas with multiple stressors.
Furthermore, geographically based measures of hazard are potentially
useful cumulative measures - although they do not provide information on
the risks, the locations of hazards can be used as an indictor of cumulative
exposures, thus risks from all of the potential chemicals associated with
that site. The environmental justice literature has used this approach.
Within this context, a semi-quantitative approach to cumulative risk
assessment was demonstrated by Dr. Jill Litt in her recent dissertation.
She demonstrated that it is possible to identify high risk populations using
information on the location of brownfields, in conjunction with health and
demographic statistics, [additional information on this dissertation is
available from her advisor, Dr. Thomas Burke].
2) Inherently Cumulative Measures
There are a variety of measures that are inherently cumulative. These
include biomarkers (they give the full effect or full exposure, regardless of
source) and measures of the incidence and prevalence of disease in a
community. The latter give an indication of the total effect of multiple
sources of exposure. In light of our understanding of the multifactorial
basis of disease, a public health approach that says "regardless of the
cause, a community has x level of disease can be very informative. Such
statistics can be compared across geographical areas that have different
sources or different groups that have different levels of vulnerability. The
approach is based strongly in the field of epidemiology. Indeed, the most
often heard critique of epidemiology - that it is the prevalence or
incidence of disease documented is a function of the combined effect of
many exposures (over time and/or space), is exactly what makes it so well
suited for cumulative risk assessment. It is likely that epidemiological
concepts will figure prominently in cumulative risk assessment, both in
G-91
-------
Margo Schwab
identifying the underlying vulnerability of a population and by generating
hypotheses regarding the determining relative contributions of multiple
stressors.
Sources of data include cross sectional analyses that determine prevalence
levels, as well as basic surveillance techniques. With respect to the latter,
The Pew Environmental Health Commission
(http://pewenvirohealth.jhsph.edu/html/home/home.html then click on
"reports") has recently completed a series of reports that document the
extent of national and state level resources for chronic disease
surveillance. Reports focus on the type of surveillance systems needed, as
well as the status of registries for birth defects and asthma. Health Track
(http://health-track.org/ and http://healthyamericans.org/)is the outgrowth
of that research, and is devoted to tracking and monitoring of chronic
disease that would help communities begin to identify patterns of health
problems.
Biomarkers are inherently cumulative risk measures, thus they would be
discussed under this category.
3) Common Denominators
This category of measures would include those associated with assigning a
common measurement to all costs and benefits associated with a particular
situation. Common metrics include money, time, and effort. Use of a
monetary metric is based on an implicit faith in the markets. Examples of
approaches to quantifying non-monetary costs and benefits include:
expressed values (willingness to pay or willingness to accept); human
capital costs (year potential life lost); or costs to society (friction costs).
The current discussions of the DALYs, etc. would feature in this section.
These measures need to be described more clearly, and linked to the
concept of cumulative risk.
4) Mathematical Models
Discuss options that range from the simple additive model of toxicity that
uses traditional risk assessment methods, to complex synergistic
approaches.
The recent dissertation of Dr. Mary Fox suggests that you can use
cumulative toxicology to characterize community health. Dr. Fox uses a
multiple health end point data base to characterize cumulative risk. She
found that her total hazard ratio risk score, created from this data base was
correlated with estimates to Years Potential Life Lost (YPLL) and
G-92
-------
Margo Schwab
mortality rates at the census tract level, [information on this project is
available from her advisor, Dr. Thomas Burke]
Another subtopic under mathematical models would be the use of
epidemiologic studies to predict effects from multiple exposures. This is
the "risk factor" method discussed in the current document. The current
discussion requires a re-write by an epidemiologist, with particular
emphasis on differentiating between conclusions at the individual level
versus the population level. See my comments in the text of the
document.
5) Others?
Emphasize here the need for new, creative approaches - that one should
not feel tied to these approaches, as the science and process will only
progress with innovation.
Each of the approaches set out here should be linked to a separate box that lays
out the strengths and weaknesses for dealing with uncertainty, cumulating across
types of stressors, vulnerability, variability, interpretation, and quantification of
that approach. The text associated with each section (a maximum of two pages
each) would just enough to provide the information necessary for choosing an
approach, not implementing that approach. An example should be presented of a
situation in which each of the approaches might be of value.
All of these approaches are amenable to application to a broadly defined
assessment with multiple stressor or a narrowly defined assessment with a single
stressor.
General comments:
The most of Section 3.2 should be eliminated. This section is too detailed for a
framework document (the person reading this document should not have to
understand dose response at this level of detail). This information is better
presented elsewhere (in other guidance documents). Furthermore, this
presentation puts too much emphasis on quantitative results and on using the
techniques traditionally used instead of setting out a framework that provides the
leeway for creative new approaches.
G-93
-------
Margo Schwab
B. Combining: Chemical and Non-Chemical, Risks and Impacts, Types
of Risks,
This should not be a separate section; rather the framework should be generic
enough to apply to any such combination of stressors, risks, or impacts. Rather,
what the document should do is identify approaches that are more likely to be of
value in combining different sorts of information (e.g., if one is doing an FQPA
type analysis, a mathematical model would be more appropriate than a qualitative
matrix, whereas a qualitative approach may be the best option for an assessment
designed to set priorities.) See above discussion of qualitative approaches; need
for innovation, etc.
Note: Sixteen pages of unnumbered attachments follow. Next numbered page is G-111.
G-94
-------
iQ
A
J-
b o.
»
r .b^- ~
V)
^'NA
J \
-------
WX03V
Hl
ex.
V^x^&^^S.
-------
r \,*y/V>i_ J>
i)
v^
»rs»JUs«i
r^Y^H
w
-------
Human Exposure
(Level and Number of People)
HIGH
LOW
HIGH
Health Effects
(Toxicity)
LOW
Highest
Priority
Medium
Priority
Medium
Priority
Lowest
Priority
Exposure Estimators
• Production Volumes
• Emissions
• Environmental Concentration
• Human Contact
• Internal Dose
Fig. 2. Conceptual framework for setting environmental health priorities.
Effects Estimators
1 Structure - Activity Relationships
1 In Vitro Data
1 Animal Toxicology
• Human Clinical Studies
1 Epidemiology
»*W.
5
-------
Estimating Human Exposures to Environmental
Pollutants: Availability and Utility of Existing Databases
KEN SEXTON
Office of Health Research
SHERRY G. SELEVAN
Human Health Assessment Group.
U.S. Environmental Protection Agency
Washington, D.C.
DIANE K. WAGENER
Centers for Disease Control
National Center for Health Statistics
Hyattsville, Maryland
JEFFREY A. LYBARGER
Division of Health Studies
Agency for Toxic Substances and Disease Registry
Atlanta, Georgia
ABSTRACT. Information about human exposures to environmental agents is a crucial com-
ponent of informed decisions about protection of public health. Results from an inventory
of exposure-related databases are used to examine the value of exposure information for
risk assessment, risk management, surveillance of status and trends, and epidemiologic
studies. Findings indicate that current and future exposure-related databases should in-
clude (1) standardized procedures for the collection, storage, analysis, and reporting of
data; (2) an enhanced ability to compare data over time, i.e., conduct comparison studies
of "old" and "new" methods; (3) mechanisms for coordination and cooperation among
public and private-sector organizations with respect to the design, maintenance, ex-
change, and review of information systems; (4) measurements of actual exposures and
dose for relevant human populations; and (5) data collection, storage, and retrieval meth-
ods that permit easy manipulation of information for both model building and testing.
DURING normal day-to-day activities, everyone
comes into contact with environmental pollutants, i.e.,
breathing of air, drinking of water, consumption of
food, and encounters with soil or dust. This contact be-
tween people and pollutants, termed exposure, re-
quires the simultaneous occurrence of two events: the
presence of a pollutant in an environmental medium
(e.g., air, water, soil, food) and contact between a per-
son and that medium. More precisely, environmental
exposure is defined as contact between the outer
boundary of the human body (e.g., skin, nose, and
throat) and a pollutant or pollutant mixture. It is quanti-
fied by reporting the concentration of the pollutant and
the time of contact.1
Four basic characteristics describe exposure: (1)
Route—do exposures occur by inhalation, ingestion, or
dermal absorption? (2) Magnitude—what is the pollu-
tant concentration (e.g., parts per million, micrograms
per cubic meter, milligrams per liter)? (3) Dura-
tion—what is the duration of exposure (e.g., minutes,
hours, days, lifetime)? and (4) Frequency—how often
do exposures occur (e.g., daily, weekly, seasonally)?
Although magnitude (i.e., concentration) is the most
commonly reported parameter, exposure data are
398
Archives of Environmental Health
-------
more useful when expressed as a concentration over
some specified time (e.g., ppm/h, mg/l • d).
Exposure is a key element in the chain of events that
leads from release of pollutants into the environment to
a concentration of the pollutant in one or more envi-
ronmental media, to actual human exposure, to inter-
nal or delivered dose, and ultimately to environmental-
ly induced disease or injury. This series of events (Fig.
1) serves as the conceptual basis for understanding and
evaluating environmental health. Actions taken by so-
ciety to protect its members from the harmful health
consequences of pollution are based on established or
postulated links between pollution sources, human ex-
posures, and adverse health effects.
Estimation of health risks associated with environ-
mental pollutants is composed of two primary activi-
ties: (1) exposure assessment and (2) effects assess-
ment. During exposure assessment, the initial portion
of the event chain is evaluated (Fig. 1), i.e., sources of
pollutants, media concentrations, exposures, and dose.
A major goal is to estimate exposure levels and the
number of persons exposed (e.g., the population ex-
posed to nitrogen dioxide concentrations that exceed
the National Ambient Air Quality Standard). In addi-
tion, the relative contributions of all important sources
and exposure pathways to the associated target dose
are determined as part of exposure assessment.
Health effects assessment focuses on the final portion
of the event sequence, i.e., exposure, dose, and ad-
verse effects. The goals are twofold: (1) determination
of intrinsic health hazards associated with the pollu-
tant, including cancer and noncancer effects; and (2)
quantification of the relationship between target dose
or exposure and health effects (i.e., dose-response) in
human populations.
The overlap between exposure and effects assess-
ment, shown in Figure 1, reflects the importance of ex-
posure and dose information to both activities. Expo-
sure determination, a critical component of epidemio-
logic studies, is needed to examine associations be-
tween environmental exposures and potential health
consequences. Measurement of internal dose is crucial
for relating exposure to dose (i.e., pharmacokinet-
ics—what the body does to the pollutant) and for relat-
ing dose to effects (i.e., pharmacodynamics—what the
pollutant does to the body). Moreover, measurements
of environmental pollutants or their biological conse-
quences after contaminants have crossed one of the
body's boundaries and have entered human tissues or
fluids are increasingly used as "biomarkers" of expo-
sure, effects, and susceptibility.2
Although often overlooked, exposure is an equal
partner with effects for the determination of health risks
of environmental pollutants. The simple two-by-two
matrix in Figure 2 represents the conceptual paradigm
that underlies prioritization of environmental health
problems. The highest ranking problems are those that
embody elevated exposures (i.e., elevated levels expe-
rienced by a significant number of people) and very
toxic pollutants (i.e., harmful at low exposure/dose).
Conversely, the lowest priority problems exhibit rela-
tively low exposure and toxicity.
Despite their obvious importance, human exposure
data are not collected in a systematic or comprehen-
sive manner. Only limited information is available;
therefore, understanding historical trends, estimating
current levels, and predicting future directions for envi-
ronmental exposures to populations and population
subgroups is difficult. The available data tend to be an-
ecdotal, fragmented, and focused narrowly on specific
pollutants, media, and routes of exposure. Recently,
several groups, including the Environmental Protection
Agency's Science Advisory Board and the National Re-
search Council, have pointed out the value of realistic
exposure information; the current lack of adequate and
appropriate data; and the concomitant need for better
methods, measurements, and models.3"7
Exposure Assessment
• Level
• Distribution
• # of People
• Source Apportionment
• Target Dose
Emission Source(s)
• Pollutant Type
• Amount Released
• Geographic Location
Environmental
Concentrations
• Air
• Water
• Soil
• Food
>,
Human Exposures
• Route
• Magnitude
• Duration
• Frequency
Internal Dose
• Absorbed Dose
• Target Dose
• Biomarkers
>,
Health Eftect(s)
• Cancer
• Noncancer
- Damage/Disease
- Signs/Symptoms
V
Effects Assessment
• Intrinsic Hazard
• Type of Effect
• Dose-Response
Fig. 1. Relationship of exposure assessment and effects assessment to the Environmental Health Paradigm.
November/December 1992 [Vol. 47 (No. 6)]
399
-------
Human Exposure
(Level and Number of People)
HIGH
LOW
HIGH
Health Effects
(Toxicity)
LOW
Exposure Estimators
• Production Volumes
• Emissions
• Environmental Concentration
• Human Contact
• Internal Dose
Fig. 2. Conceptual framework for setting environmental health priorities.
Effects Estimators
1 Structure - Activity Relationships
1 In Vitro Data
Animal Toxicology
Human Clinical Studies
Epidemiology
Concerns about the adequacy of available exposure-
related information prompted a joint inventory (EPA,
Centers for Disease Control [CDC]-National Center for
Health Statistics [NCHS], and Agency for Toxic Sub-
stances and Disease Registry [ATSDR]) of federally
sponsored databases and a consensus workshop that
targeted the use of databases in risk assessment, risk
management, epidemiology, and status and trends
analyses. The objectives were to examine the existing
exposure-related databases and to identify needed im-
provements. The following discussion highlights impor-
tant issues that relate to the estimation of exposures,
briefly examines results of the database inventory, and
summarizes recommendations of the consensus work-
shop.
Estimating human exposures
Pollutants can move along many different pathways
between the emission point and subsequent contact
with people. Some of these exposure pathways are dis-
played in Figure 3 for airborne and waterborne pollu-
tants emitted from a hypothetical che.mical plant. Al-
though Figure 3 may appear complicated, it does not
capture the complexity of the real world in which a ple-
thora of sources (e.g., motor vehicles, industrial wastes,
agricultural runoff, consumer products) release literally
thousands of pollutants (e.g., chemicals, microorgan-
isms, radionuclides) in a multitude of settings (e.g., resi-
dential, occupational, recreational, transportational) in
which people can be exposed.
A more generic overview of the major pathways by
which people come into contact with pollution is pro-
vided in Figure 4. Although Figure 4 does not explicitly
show "nontraditional" pathways, e.g., infant expo-
sures to contaminated breast milk or indoor air pollu-
tion exposures from the volatilization of water pollu-
tants during showering, it does demonstrate the many
possible ways that people and pollution connect.
Historically, these pathways have been examined in-
dividually. It is clear from Figure 4, however, that multi-
ple pathways can contribute to exposures and dose for
a single pollutant (e.g., lead in drinking water, indoor
and outdoor air, paint, house dust, soil, and food—all
of which contribute to blood-lead levels). Many human
exposures (e.g., metals, dioxins and related com-
pounds, pesticides, polycyclic organic matter) occur
through a wide spectrum of environmental pathways
and by different routes (e.g., inhalation, ingestion, and
dermal contact). Acknowledgment of this reality and
the need to account for it in exposure assessment are
embodied in the "total human exposure" concept.5
The total human exposure approach to assessment
accounts for all relevant pathways and routes when en-
vironmental exposures experienced by an individual or
group are estimated. The information gained from ap-
plication of this methodology can be used to construct
and examine the "human exposure topology" (Fig. 5)
for the pollutant of interest.8 The total exposure ap-
proach to design, collection, analysis, and presentation
of information provides a more realistic picture of ex-
posure patterns.
Ideally, all exposure assessments would be based on
data similar to that displayed in Figure 5. But in reality,
the costs and, in some cases, the technical feasibility
limit our ability to make the necessary measurements.
Nevertheless, the only way to determine whether and
to what extent people come into contact with specific
environmental contaminants is to measure their expo-
sure. This may be accomplished in three different ways:
(1) use of microenvironmental samplers (e.g., passive
400
Archives of Environmental Health
-------
f M * f> t *
)* * * fc f * ,"d**3!i^J<>S»
CONTACT: Donna Rohrer
(202) 828-9762
Protect Public From Chronic Diseases, Pew Commission Urges
Diseases, Environmental Exposures Must Be Tracked to Save Lives
WASHINGTON, DC -Sept. 6, 2000 -- The Pew Environmental Health Commission
at the Johns Hopkins School of Public Health today called on Congress and the White
House to protect Americans from chronic diseases - the No. 1 cause of death in the U.S.
- by tracking where and when these health problems occur and possible links to
environmental factors.
In its third report, the Commission charged that the nation faces an environmental
health gap and proposed a Nationwide Health Tracking Network to provide the critical
information now lacking in nearly every state. Without this Network, the Commission
said the U.S. will remain unable to mount effective prevention efforts for asthma, birth
defects, developmental disabilities, cancers, and neurological disorders such as
Alzheimer's and Parkinson's, among other chronic diseases.
"We responded quickly to the threat of West Nile virus, tracking and monitoring
every report of infected birds and people, but 20 years into the asthma epidemic this
country is still unable to track where and when attacks occur and what environmental
links may trigger them," said Lowell Weicker Jr., the Commission chairman and former
U.S. senator and Connecticut governor.
The recommended Network would include the following components:
1. Nationwide baseline tracking of priority diseases - asthma and chronic respiratory
diseases; birth defects; developmental disorders; cancers, especially childhood
cancers; and neurological diseases such as Alzheimer's, multiple sclerosis and
Parkinson's - and priority exposures such as PCBs, and dioxin; heavy metals such
as mercury and lead; pesticides and water and air contaminants.
2. Monitoring of immediate health crises such as heavy metal and pesticide
poisonings to serve as an early warning system.
3. Establishing 20 state pilot tracking programs to address regional environmental
health concerns.
4. Developing a federal, state and local rapid response capability to investigate
clusters, outbreaks and emerging threats.
-------
5. Supporting community interests and scientific research to further health tracking
efforts.
The Commission estimated that the Network will cost $275 million annually - less
than one-tenth of a percent of the $325 billion that chronic disease costs the U.S. annually
in health care and lost productivity.
"It is time to make the investment in our public health that matches the threat from
chronic disease," Weicker said.
As examples of the environmental health gap, the nation's failure to deal with chronic
diseases and their potential links to environmental hazards, the Commission noted:
• More than half the states (27) lack ongoing tracking and monitoring of asthma
even though it is a rapidly growing national epidemic;
• Most states fail to track developmental disabilities such as autism and mental
retardation despite an estimated 50 percent rise nationwide in these disabilities
in the last decade, and research indicating that 25 percent are related to
environmental exposures. California, District of Columbia, Kentucky,
Mississippi, Nebraska, New Jersey, North Carolina, South Carolina and South
Dakota reported tracking at least some of these disorders.
• Only four states reported tracking autoimmune diseases such as lupus even
though rates for these diseases are rising. The four states are: Arizona,
Massachusetts, New Mexico and South Dakota.
• Less than half the nation's population is covered by birth defect registries
even though birth defects are the leading cause of infant mortality in the U.S.
and rates for certain birth defects and related conditions are increasing.
"This is Public Health 101, and as a nation we are flunking," said Louis Stokes,
former 15-term U.S. House member from Ohio and a member of the Pew Environmental
Health Commission.
In calling for the Nationwide Health Tracking Network, the Commission has the
support of many voluntary disease support groups and national public health
organizations, including: Allergy and Asthma Foundation of America; Allergy and
Asthma Network, Mothers of Asthmatics; American Academy of Pediatrics; American
Autoimmune and Related Diseases Association; American Cancer Society, Inc.;
American Lung Association; American Public Health Association; Association of State
and Territorial Health Officials; Candlelighters Childhood Cancer Foundation; Council of
State and Territorial Epidemiologists; Joint Council of Asthma, Allergy and
Immunology; National Association of City and County Health Officials; and Public
Health Foundation.
"The support for health tracking is broad and deep. It should command the attention
of every candidate for President and every candidate for Congress," Weicker said.
-------
"Tracking on a nationwide basis will enable our medical system to focus on a critical
area of health care that has long had insufficient emphasis, the prevention of the diseases
that cause the most suffering and death," said Dr. Neil Schlackman, the senior corporate
medical director for Aetna US Healthcare and a member of the Commission.
The Commission's recommendations were based on an analysis conducted under the
guidance of Thomas A. Burke, associate professor at the Johns Hopkins School of Public
Health and chairman of the Director's Advisory Committee to the National Center for
Environmental Health at the Centers for Disease Control and Prevention (CDC). The
analysis examined public health tracking capacity at the national, state and local levels.
"We need to know when and where diseases occur, what are the environmental
hazards, and what are the actual exposures we have to those hazards," said Burke. "This
is a basic right to know issue for our communities and our public health professionals."
"Americans have a right to this information about the health of their communities,"
said the Rev. Michael D. Place, STD, president and chief executive officer of the
Catholic Health Association of the United States, representing the nation's largest group
of not-for-profit healthcare systems, facilities and related organizations, and a member of
the Commission. "These recommendations will enlighten our research and make
possible disease prevention efforts that will save lives and spare much suffering."
Later this year, the Commission will release its final recommendations on improving
the ability of the public health system to combat environmental threats to health. Its two
previous reports examined the increases in birth defects and related conditions and the
asthma epidemic, focusing on the need for national environmental health tracking.
The Pew Environmental Health Commission was launched in May 1999 to develop
recommendations on improving the nation's ability to track and prevent health problems
linked to environmental conditions. Its members include leaders from the public policy,
health industry, environment, government, academic and nonprofit communities.
The Commission is funded by a grant from the Pew Charitable Trusts to the Johns
Hopkins School of Public Health. The Pew Charitable Trusts support nonprofit activities
in the areas of culture, education, the environment, health and human services, public
policy and religion. Based in Philadelphia, the Trusts make strategic investments to help
organizations and citizens develop practical solutions to difficult problems. In 1999, with
approximately $4.9 billion in assets, the Trusts granted over $250 million to 206
nonprofit organizations.
For the full report and appendices, visit the Commission's Web site at
httv://vewenvirohealth. ihsvh. edu
-------
EMBARGOED FOR RELEASE
Until Noon EST Nov. 16,1999
Contact: Donna Rohrer or Sara Sigelman
(202) 659-0330
Commission Study Finds Increases in Certain Birth Defects
Nearly 59 Million Americans Not Covered by Birth Defects Tracking
NEW YORK - Nov. 16,1999 - A major analysis of national data on birth defects and the environment has
found unexplained increases in certain birth defects and related conditions that point to the need for strengthening
the public health system.
Even though birth defects are the nation's leading cause of infant death, one-third of the states, the
District of Columbia and Puerto Rico - with a total population of nearly 59 million - fail to track birth defects, and
25 more states have systems that need improvement to better protect the public.
The study, conducted by the Pew Environmental Health Commission at the Johns Hopkins University
School of Public Health, reported that only about 20 percent of birth defects have known causes, while the causes
of the majority await further research. However, there is increasing evidence that environmental factors, including
diet, personal behavior and exposure to toxic substances and pollutants, may play an important role in the
development of birth defects and related conditions.
Commission researchers concluded that the unanswered questions about wide variations in certain birth
defect rates reported over an eight-year period may be the result of improved diagnoses, variations in state
tracking procedures or increases that are actual.
Researchers evaluated each state's monitoring effort and graded each on its quality. One-third of the
states, the District and Puerto Rico - or 19 - received an "F" grade because they do not currently track birth
defects at all. Only eight states were awarded an "A" for systems that can be helpful to future research on the
causes of birth defects. The remaining 25 states have tracking systems that are in need of improvements.
OF STATED 1RTII t*rBCT« SUR VEI1.LAKCE
. MOVSMBBB 1 W9
-------
"Without this information, public health officials are working in the dark," said Lowell Weicker, Jr., former
Connecticut governor, three-term U.S. Senator and chairman of the commission. "We lack the key tool needed to
identify emerging disease clusters and trends, making it tougher to tackle the environmental threats that may
cause sickness and death in our children."
While federal and state investments in birth defects monitoring over the past two decades have produced
some excellent systems, as observed in the commission's report, much more must be done.
"What we need is full funding of the Birth Defects Prevention Act of 1998," said Dr. Jennifer L. Howse,
president of the March of Dimes and member of the commission. "This legislation called for more surveillance,
research, education and services to prevent birth defects and protect children, but without full funding the job just
won't get done."
The analysis examined data from the Centers for Disease Control and Prevention and states gathered
from 1989-1996. Increases were found by analyzing information from those states with tracking systems for birth
defects.
Among the study's findings about birth defects and related conditions:
• Rates of babies born with low birth weight and pre-term conditions have been rising steadily since the
micL198Qs despiteincreased prevention efforts. LowLbirth weight and pre-temi conditions in infants
contribute to infant deaths and often accompany birth defects and related conditions such as cerebral
palsy and mental retardation.
• The rate of infants born with one serious heart defect rose 2 _ times in less than a decade among
states that track this defect.
• The rate of infants born with a blockage in the urinary tract rose more than 1 _ times in less than a
decade among states that track this defect.
As a reflection of the variations in birth defects monitoring systems, researchers found inconsistencies
even among the 33 states collecting data.
"Gaps in quality and consistency in state data prevent public health officials from answering the many
disturbing questions raised by our analysis," said Dr. Lynn R. Goldman, principal investigator on the birth defects
study.
"We know more about how much pollution we put in our physical environment than we know about the
levels of those pollutants in our bodies and how they can affect the health of a developing baby," she said.
The Pew Environmental Health Commission was launched in May to develop recommendations to
improve the nation's ability to track and prevent health problems linked to environmental conditions. Commission
members include leaders from the public policy, health industry, government, academic and nonprofit
communities.
In the coming months, the Commission will study two other children's environmental health issues,
asthma and childhood cancers, and examine the quality of state tracking systems and other issues relating to the
capacity of the public health system to combat new threats to health.
The Pew Environmental Health Commission is funded by a grant to the Johns Hopkins School of Public
Health by The Pew Charitable Trusts. The Pew Charitable Trusts support nonprofit activities in the areas of
culture, education, the environment, health and human services, public policy and religion. Based in Philadelphia,
the Trusts make strategic investments to help organizations and citizens develop practical solutions to difficult
problems. In 1998, with approximately $4.734 billion in assets, the Trusts granted over $213 million to 298
nonprofit organizations.
For the full report and state-by-state analysis, visit the Commission's website at http://Dewenvirohealth.ihsDh.edu
###
-------
The Pew Charitable Trusts: Investing in Ideas
Page 1 of 1
TRUSTS
HOME • ABDtlTUS | GHAUTS IHFOIMgtQN \ mm BOOM | iPHBUCCTONS |
HWfi
Articles about Pew
Facts and Stats
Good News/Grantee
Awards
Grant Highlights
Grantee Press
Releases
Grantee Reports
Newspaper/Magazine
Op-eds (grantee)
Pew Press Releases
Pew-produced
Publications
Strategy Paper
Summaries of Reports
and Publications
White Paper
1 more ideas... J>j
If I Sign
on the topic of your choice
Tracking disease in Americasearch this issue
Chronic diseases are the leading cause of death
in the U.S., yet the systems in place to combat
this major public health problem are woefully
inadequate. For example, more than half the
states have no tracking and monitoring of
asthma, which afflicts 17 million Americans,
including 5 million children.
Existing efforts to gather information on asthma
and other chronic diseases and their potential
environmental links are highly fragmented and
inadequate for collecting the kind of nationwide
data that are needed to truly understand how
often, in what locations and why these diseases
occur. This information could also help to
eventually prevent the incidence of some of
these diseases in the future. Lacking such
information, it is difficult for health care
professionals, the public and government
officials to grasp the magnitude of the problems
we face and how to prevent them.
Home > Tracking disease
Web Policies
http://www.pewtrasts.org/ideas/index.cfm?issue=14
8/29/2001
-------
? 2 » HI Market Place, Suite 850
EBVIRGKMBHfAL Baltimore, MD 21202
B 35 A i T B COHMXSSIQH 410.659.2690 phone
,. . .. 410.659.2699fax
EMBARGOED UNTIL 10 AM EDT MAY 16. 2000
Contact: Donna Rohrer (202) 659-0330
Dave Woodruff
ASTHMA CASES PROJECTED TO DOUBLE BY 2020;
WILL HIT 29 MILLION AMERICANS; 1 IN 5 FAMILIES
Commission Charges Federal Government Is Failing to Stop Asthma Epidemic
WASHINGTON - May 16, 2000 - The nation is in the grip of a rapidly growing asthma epidemic
whose victims will more than double by 2020, according to a study released today by the Pew Environmental
Health Commission at the Johns Hopkins School of Public Health.
An analysis by the Johns Hopkins researchers concludes that new cases of asthma among Americans
are increasing so rapidly that by the time another generation of children is born, asthma will strike 29 million
Americans - more people than the projected 2020 population of New York and New Jersey combined.
The report sharply criticized the federal government for shortchanging both research efforts to find
the...causes of asthma and public health strategieslo prevent its, spread. Two weeks ago, the federal
government released its action plan on asthma, 15 years into the epidemic.
"While the federal government acknowledges that there is an asthma epidemic, it has so far failed to
devote sufficient resources to tracking and preventing asthma," said Lowell Weicker, Jr., former governor
and three-term U.S. senator from Connecticut and chairman of the Commission.
"The real tragedy of this asthma epidemic is that the federal government's response has been too
little, too late," said Weicker. "If asthma were an infectious instead of a chronic disease, we wouldn't be
having this discussion, because the full power of the federal government would be aimed at preventing it."
The Johns Hopkins researchers concluded that in fiscal year 1999, less than 10 percent of federal
asthma research spending - approximately $14 million - is directed at monitoring or preventing the disease.
This year, asthma will cost the U.S. economy at least $11 billion and afflict more than 14 million Americans.
The Commission determined that unless the rates are slowed, asthma will strike 1 in 14 Americans
and 1 in 5 U.S. families by the year 2020.
Total number of persons with asthma, computed and projected, 1980-2020
30 million
25 million
20 million
15 million
10 million
5 million
1980 1990 1995 2010 2020
-------
-2-
Asthma rates are going up dramatically across all demographic groups, all across America. For
example, between 1980 and 1994, they rose by 75 percent - and by 160 percent for those under age 4. In
just one year, 1995, asthma accounted for 1.8 million visits to hospital emergency rooms, and more than 10
million missed school days.
That makes asthma the number 1 cause of school absenteeism.
" Asthma affects some groups in our society more than others, including the very young, very old and
very poor," said Dr. Lynn Goldman, principal investigator and adjunct professor at the Johns Hopkins School
of Public Health. "This is why tracking is so very important - it will help us understand why some groups
are hit harder by asthma and help find the keys to preventing the disease."
"While a higher percentage of African Americans have asthma, the disease is rapidly rising
everywhere and is affecting everyone," said Louis Stokes, former 15-term U.S. congressman from Ohio and
member of the Commission. "Despite what you may have heard, it is America's problem."
"It's time for the federal government to declare a War on Asthma," Stokes said.
Asthma is a chronic^lisease~whose causes~are lafgelyHnknownTyut genetic predisposition and
environmental-triggers Jn4>othindoor-and^autdoor air-are taaown contributors to^sthmaVdevelopment and
severity. " " " "' """" '
To win this War on Asthma, the Commission called for a five-part campaign for Congress and the
Administration to:
• Establish a 15-state tracking system within two years - and a nationwide system in five years - to
provide public health officials with the information necessary to analyze trends, determine causes,
better understand the role of the environment, and prevent asthma.
• Deploy a "9-1-1 Force" at the Centers for Disease Control and Prevention within one year to conduct
thorough, targeted investigations into asthma deaths in order to identify causes and learn how to
prevent asthma.
• Begin short- and long-term environmental health studies to determine what causes asthma, the best
strategies for prevention, and why asthma rates are higher in low income and minority communities.
• Begin a comprehensive public education program within two years on the immediate steps we can
take to combat asthma, beginning with the environmental factors that contribute to the development
or severity of the disease.
• Give the U.S. Surgeon General the authority to marshal all federal efforts on behalf of the War on
Asthma.
"The findings of the Pew Environmental Health Commission's report underscore the importance of
dramatically escalating the nation's War on Asthma," said John M. Coruthers, Jr., President of the American
Lung Association. "The American Lung Association strongly supports the Pew Commission's
recommendations for improving our national response to this public health crisis."
"We can win the War on Asthma in the United States by investing in the public health infrastructure in
response to this epidemic-with an emphasis on prevention," said Mary Worstell, executive director of the
Asthma and Allergy Foundation of America. "The Pew Environmental Health Commission's report is a call
to action in this vital life-saving effort."
The Pew Environmental Health Commission was launched in May 1999 to develop
recommendations on improving the nation's ability to track and prevent health problems linked to
-------
-3-
environmental conditions. Commission members include leaders from the public policy, health industry,
environment, government, academic and nonprofit communities.
The Commission released a report in November 1999 on birth defects and the environment that
concluded that although there are unexplained increases in certain birth defects nationwide, a third of all
states fail to track birth defects, and half of the states have monitoring systems that need improvement.
In the coming months, the Commission will examine the quality of state and federal health tracking
systems in more detail, and issue its final recommendations on the capacity of the public health system to
combat environmental threats to health.
The Pew Environmental Health Commission is funded by a grant to the Johns Hopkins School of
Public Health from The Pew Charitable Trusts. The Pew Charitable Trusts support nonprofit activities in the
areas of culture, education, the environment, health and human services, public policy and religion. Based in
Philadelphia, the Trusts make strategic investments to help organizations and citizens develop practical
solutions to difficult problems. In 1999, with approximately $4.9 billion in assets, the Trusts granted over
$250 million to 206 nonprofit organizations.
For the full report and appendices, visit the Commission's website at http://pewenvirohealth.jhsph.edu
###
-------
Post-Meeting Written Comments From Ken Sexton
G-lll
-------
8-25-01
Comments on an early draft of
EPA's Framework for Cumulative Risk Assessment
by Ken Sexton
General Comments
The document is a reasonable early draft that does (1) lay out a framework for thinking
about cumulative risk assessment, (2) highlight many of the important issues that must be
addressed, and (3) begin the process of building a common language. EPA is to be
commended for both the attempt and the quality of this work in progress.
The Framework should state clearly that cumulative risk assessment is a tool for
organizing and analyzing information in a way thatt will be useful for the decision at hand.
It is not a decision rule that constrains decisionmakers' discretion.
As I understand it, the "essence" of the framework involves (1) 3 phases -formulation,
analysis, interpretation, (2) an analytic-deliberative process, (3) up front consideration of
key issues including identification of the "risk management team" and the "risk assessment
team," (4) explicit consideration of both values (value judgments about scope, combining
risks, etc) and facts (data, data quality, uncertainty) throughout the process, (5)
specification of the vital activities that are performed as part of each phase, (6)
identification and brief discussion of key technical and science policy issues that are
intrinsic to cumulative risk assessment, and (7) definitions for important terms, concepts,
and principles. However, it took several readings and substantial effort to discern this
framework.
A major problem with the current version has to do with style (as opposed to substance).
It desperately needs a professional editorial presence to (1) punch up the verbiage
describing the key ideas, concepts, and principles, (2) provide suitable diagrams that
summarize significant elements of the framework and provide the reader with a better road
map of the document, and (3) restructure and rearrange the narrative so as to make
crucial points early and to reduce redundancy throughout (for example, several statements
describing the goals are spread randomly throughout the current version). I strongly
encourage EPA to retain a good editor and work with him/her to make the document more
focused, more concise, and less redundant. As it is, the draft framework is uneven (much
detail about some important issues and almost none about others), inappropriately even in
G-112
-------
Ken Sexton
tone/style (doesn't draw attention to the essential elements of the framework because all
issues, major and minor, get about the same level of description/discussion), and
unbalanced (meant to be generic but virtually all examples are human health, at the very
least it needs a couple of ecological examples; it also emphasizes quantitative approaches
and gives short shrift to qualitative methods). It is important, to the extent practicable, that
the discussion and descriptions relate directly to cumulative risk assessment (much of the
current version is too generic and non-specific).
I question whether the Framework needs to be so historical in terms of describing the wide
range of EPA documents/guidelines (many still in draft), some of which are only
tangentially related in one way or another to the Framework. For me this detracts from the
impact of the Framework itself and needlessly adds to the length of the narrative. I
recommend putting this sort of review of earlier EPA work (unless it relates directly to
cumulative risk) in an appendix. An explicit decision should be made about whether to
reference these draft documents at all.
One area where I think a formal peer review will find fault is the reference list. The
selection of references is always a difficult one, especially in this context. It is obvious that
a comprehensive literature review is inappropriate, so to the extent that the scientific
literature is cited at all the authors must be selective. The key question for future
reviewers is "have the authors cited the "best" primary or secondary source for the point
being made"? It is not apparent to me that the answer is necessarily "yes".
Process Issues
The Framework makes it clear that stakeholders are important and that they need to be
involved, as appropriate. However, virtually all of the discussion of stakeholders is found
in the formulation phase, and it is not easy to understand what roles they play, if any, in the
analytical and interpretation phases. The Framework could do a better job of making it
clear that (1) stakeholder involvement creates the need for specific resources to be set
aside up front for this purpose (e.g., funding for transportation, child care, reimbursement
for time), (2) there is a variety of different roles stakeholders might play (e.g., listeners,
respondents, advisors, partners, joint decisionmakers) and it is critical to decide at the
outset which role(s) are appropriate, (3) regulatory agencies need effective methods to
gather information, test ideas, and obtain opinions from stakeholders without necessarily
relinquishing their statutory role and responsibility to make the final decision; and
furthermore, that they cannot allow the decision-making process to be sidetracked or
delayed by a futile search for unanimous agreement (the reality is that regulators should
not feel obliged in every instance to offer stakeholders either a vote or veto). It would be
helpful if the Framework would be more specific about the value of stakeholder
involvement. Among the many reasons to involve stakeholders are the following: (1)
stakeholder acceptance (buy in) is critical for some cumulative risk assessments, (2)
G-113
-------
Ken Sexton
stakeholders often have unique knowledge of local conditions and can complement
specialized knowledge of risk assessment experts, and (3) including stakeholders as
equals in cumulative risk assessments is an important social value that builds social capital
and promotes the concept of environmental democracy.
In order for community members to participate meaningfully, the Framework should
explicitly acknowledge that they often need (1) financial support to attend meetings (e.g.,
parking, transportation, child care, time off from work), (2) independent technical/scientific
assistance, and (3) help understanding their rights and responsibilities as equal
participants in a cumulative risk assessment.
There are several prosaic yet critical questions that must be discussed and resolved during
the formulation phase, including:
When is the cumulative risk assessment to be completed? (timeliness issues)
How much money is available to conduct it? (fiscal/budgetary issues)
Which government agencies (federal, state, local) will be involved?
What are their respective roles and responsibilities?
Which stakeholders (business, NGOs, community) will be involved?
What are their respective roles and responsibilities?
How are decisions going to be made? (unanimity vs consensus vs other)
The Framework should do a better job of emphasizing the importance in the formulation
phase of directly confronting the almost overwhelming complexity associated with
cumulative risk assessment, and deciding what compromises need to be made so that the
agreed upon task can be completed within budget and time constraints.
The Framework should be more specific about the challenges of successful stakeholder
participation, and the fact that we don't know yet what works and what doesn't across the
spectrum of possible cumulative risk situations. Although there is much anecdotal
information, it should be made clear that this issue is as much art as science. I suggest
that the document reference a couple of key overview/summary articles and books, such
as:
Chess, C. and K. Purcell (1999). Public Participation and the Environment:
Do We Know What Works? Environmental Science & Technology 33(16): 2685-
2692.
Frewer, L. (1999). Risk Perception, Social Trust, and Public Participation in
Strategic Decision Making: Implications for Emerging Technologies. Ambio 28(6):
569-574.
Thomas, J. (1995). Public Participation in Public Decisions: New Skills and
Strategies for Public Managers. San Francisco, CA, Jossey-Bass Publishers.
It would be helpful if the Framework explicitly recognized the literature on the different
dimension of risk (e.g., Slovic, Covello, Sandman, NRC) and reminded the reader that
G-114
-------
Ken Sexton
experts and community stakeholders are likely to approach risk-related issues from
different frames of reference. This relates directly to decisions by the risk management
team about (a) combining chemical and non-chemical stressors and (b) combining
different types of risk.
Technical Issues
The issue of qualitative versus quantitative cumulative risk assessment deserves more
attention. As currently structured, the Framework focuses almost exclusively on methods
to arrive at a quantitative outcome, while giving the impression that qualitative methods are
somehow less useful and less desirable. I believe that qualitative approaches are valuable
for cumulative risk assessment and, in the near-term, the only practical way to address the
complexities involved. Rather than being discouraged, they should be advocated as a way
to overcome the complexity and data deficiencies that hinder quantitative approaches.
The discussion should describe qualitative methods (giving examples) and mention related
techniques for dealing with uncertainty/variability. It is vital for the Framework to make
clear that risk is not an inherently quantifiable variable.
Relevant topics like Expert Judgment techniques, Focus Groups, Opinion Surveys, Citizen
Juries, Alternative Dispute Resolution, etc. are not mentioned.
The Framework should more clearly state the obvious: cumulative risk assessments, either
qualitative or quantitative, are likely, on average, to require more time, more money, and a
broader skill mix than traditional risk assessments. There will be an expanding need for
social scientists, urban planners, philosophers, and ethicists, among others, to become
involved.
Variability should be discussed along with uncertainty in the document, and the importance
of each in all three phases should be made clear. The need to be consistent about how
uncertainty/variability is addressed (qualitatively or quantitatively) should be stressed. It is
important to state that this attention to uncertainty/variability is an essential element of
characterizing our ultimate "confidence" in the final product(s).
It should be stated explicitly that it might be either impossible or inadvisable to combine (1)
chemical and non-chemical stressors and (2) different types of risk, depending on the
situation and circumstances of the cumulative risk assessment.
Instead of talking about "Four General Approaches to Cumulative Risk Assessment"
(Section 3.1), it would be better to talk about "Four Examples ...." It would also help to add
one or two qualitative approaches as examples.
The distinction between cumulative risks and cumulative impacts is blurred and confusing.
The Framework should do a better job of defining these terms, using them consistently,
G-115
-------
Ken Sexton
and discussing how they are related or unrelated for the purposes of cumulative risk
assessment.
Much has been published on the strengths and weaknesses of biomarkers (e.g., NRC),
and at least a summary needs to be included. The most glaring need, however, is for a
concise description (with examples) of the role of biomarkers in cumulative risk
assessment, including data collection, analysis, and interpretation issues.
All of the forgoing comments have ramifications for describing related research needs;
therefore, it will be necessary to revise and enhance this section. Also, it is necessary to
comment on the equally important implications of the Framework for exposure and effects
surveillance activities (e.g., routine monitoring, determination of background levels).
A key point that comes through in the Framework is the need to consider, blend, and
balance both facts and values throughout the 3 phases of cumulative risk assessment.
This is significantly different that the conventional risk assessment paradigm described by
the National Research Council (NRC, 1983, 1994), although more similar to the paradigm
for ecological risk assessment and impact assessment for NEPA. The document should
confront these differences directly and comment on their ramifications for conducting
cumulative risk assessment. It is also important for the framework to put cumulative risk
assessment in the broader context of its relationship to (1) risk management, (2) risk
communication, and (3) risk-related research and surveillance.
G-116
-------
APPENDIX H - DRAFT FRAMEWORK FOR CUMULATIVE
RISK ASSESSMENTS
-------
REVIEW DRAFT NCEA-F-1098
August 2, 2001
Framework for
Cumulative Risk Assessment
IMPORTANT NOTICE:
This document does not represent official Agency policy and, therefore, is not for citation
or quotation. It is a work-in-progress developed by the Cumulative Risk Assessment
Technical Panel of the Risk Assessment Forum and may provide a basis for development of
an Agency-wide Framework for Cumulative Risk Assessment. This draft is being provided
for the purpose of external peer consultation.
The mention of commercial products is for illustration only and in no way implies EPA
endorsement of these products.
H-l
-------
DRAFT- Risk Assessment Forum Review Draft - August 2, 2001 - Do Not Quote or Cite
Table of Contents
Acknowledgments v
Foreword vi
Preface vii
Authors, Contributors, and Reviewers ix
List of Abbreviations and Acronyms x
Executive Summary xii
1. INTRODUCTION 1
1.1. Purpose and Scope of the Framework Report 4
1.2. Intended Audience 6
1.3. Key Definitions in Cumulative Risk Assessment 6
1.4. The Cumulative Risk Assessment as a Tool for a Variety of Users and Purposes . . 9
1.5. The Broader Decision-Making Context for Cumulative Risk Assessment 10
1.6. Organization of this report 11
2. THE PLANNING AND PROBLEM FORMULATION PHASE 13
2.1. Planning and Scoping 13
2.1.1. Defining the Purpose of the Assessment 14
2.1.2. Defining the Scope of Analysis and Products Needed 15
2.1.3. Agreeing on participants, roles and responsibilities 17
2.1.4. Agreeing on the Depth of the Assessment and the Analytical Approach
19
2.1.5. Agreement on the Resources Available and Schedule 19
2.2. Problem Formulation, Conceptual Model, and Analysis Plan 19
2.2.1. Problem Formulation 20
2.2.2. Developing the Conceptual Model 21
2.2.3. Constructing the Analysis Plan 22
2.3. The Final Step Before the Analysis Phase: Discussion of Possible Outcomes .... 23
3. THE ANALYSIS PHASE 23
3.1. General Approaches to Cumulative Risk Assessment 24
3.1.1. Combining Individual Stressor Risks 24
3.1.2. Use of Risk Factors Developed from Epidemiologic Associations 25
3.1.3. Biomarkers and Biomonitoring 25
3.1.4. Other Types of Probability Statements 26
H-2
-------
DRAFT- Risk Assessment Forum Review Draft - August 2, 2001 - Do Not Quote or Cite
3.2. Issues Related to the Approach of Combining Individual Stressor Risks 26
3.2.1. Characterization of Hazard Identification and Dose-Response 27
3.2.1.1. Current Methods for Assessing Noncancer Effects 28
3.2.1.2. Current Methods for Assessing Cancer Risks 30
3.2.1.3. Time-Related Issues with Dose-Response Curves for Cumulative
Assessments 31
3.2.1.4. Issues Associated with Assessing Mixtures of Stressors 32
3.2.1.5. Hazards Other than Chemical Hazards 34
3.2.1.6. Vulnerability 35
3.2.2. Characterization of Exposure 36
3.2.2.1. The Time Dimension of Exposure 38
3.2.2.2. Variation of Mixtures 41
3.2.2.3. Sources and Pathways of Exposure 41
3.2.2.4. Subpopulations with Special Exposures 42
3.2.2.5. Exposures to Non-chemical Stressors 43
3.3. Issues Related to the Approach of Using Risk Factors 43
3.4. Issues Related to the Approach of Biomarkers and Biomonitoring 44
3.5. Issues Related to Other Approaches 44
4. THE INTERPRETATION (RISK CHARACTERIZATION) PHASE 44
4.1 Risk Estimation 44
4.1.1. Methods for Combining Chemical Risks 44
4.1.2. Other Impacts or Effects 46
4.1.2.1. Stress-Induced Risks 46
4.1.2.2. Quality-of-Life Risks 47
4.1.3. Combining Different Types of Risk 49
4.1.3.1. Converting Adverse Effects to a Common Metric 50
4.1.3.2. Translating Adverse Effects into an Index 52
4.1.3.3. Other Approaches 53
4.1.3.4. General Issues with Combining Risks 54
4.2. Risk Description 54
4.3. Uncertainly Analysis 56
4.4. The Information Provided by Cumulative Risk Assessment 58
4.4.1. Making Sense of Multiple Stressor Effects 58
4.4.2. Cumulative Risk Assessments in a Public Health Context 59
4.4.3. How Scope and Purpose of the Assessment Affect Results 59
4.5. Using the Results of the Assessment 60
5. GLOSSARY 62
6. REFERENCES 62
H-3
-------
DRAFT- Risk Assessment Forum Review Draft - August 2, 2001 - Do Not Quote or Cite
APPENDIX A: A RESOURCE LIST FOR METHODS RELEVANT TO EXPOSURE
ASSESSMENT 75
A. 1. Resources Relevant to Chemical Exposures 75
A.2. Resources Relevant to Exposures to Non-Chemical Stressors 76
APPENDIX B: ASSESSING ACCIDENTAL CHEMICAL RELEASE EXPOSURE 77
APPENDIX C. DATA QUALITY ISSUES IN MONITORING AND OTHER EXPOSURE-
RELATED DATA 79
APPENDIX D: SOME THOUGHTS ON QUALITY OF DATA IN VARIOUS WIDELY-
USED DATA BASES 80
APPENDIX E: SOME THOUGHTS ON BACKGROUND EXPOSURES 82
APPENDIX F: RESEARCH AND DEVELOPMENT NEEDS 83
H-4
-------
DRAFT- Risk Assessment Forum Review Draft - August 2, 2001 - Do Not Quote or Cite
1 Acknowledgments [to be completed later]
2
3
H-5
-------
DRAFT- Risk Assessment Forum Review Draft - August 2, 2001 - Do Not Quote or Cite
1 Foreword [to be completed later]
2
3 Several reports have highlighted the importance of understanding the accumulation of
4 risks from multiple environmental stressors. These include the National Research Council's 1994
5 report Science and Judgment in Risk Assessment and the 1997 report by the
6 Presidential/Congressional Commission on Risk Assessment and Risk Management entitled Risk
1 Assessment and Risk Management in Regulatory Decision-Making. In addition, legislation such
8 as the Food Quality Protection Act of 1996, has directed the Environmental Protection Agency to
9 move beyond single chemical assessments and to focus, in part, on the cumulative effects of
10 chemical exposures occurring simultaneously. Further emphasizing the need for EPA to focus on
11 cumulative risks are cases filed under Title VI of the 1964 Civil Rights Act. These cases have
12 emphasized the need for a population-based approach to assessing human health risks from
13 environmental contaminants.
14
15 In response to the increasing focus on cumulative risk, several EPA programs have begun
16 to explore cumulative approaches to risk assessment. In 1997, The EPA Science Policy Council
17 issued a guidance on planning and scoping for cumulative risk assessments
18 (http://www.epa.gov/ORD/spc/2cumrisk.htm). More recently, the Office of Pesticide Programs
19 has developed draft cumulative risk assessment guidance focused on implementing certain
20 provisions of FQPA. The Office of Air Quality Planning and Standards has applied cumulative
21 exposure models in its analyses for the National-Scale Air Toxics Assessment (NATA). In
22 addition, community-specific cumulative risk assessment has been explored through the Agency's
23 Cumulative Exposure Project.
24
25 The EPA Science Policy Council has asked the Risk Assessment Forum to begin
26 developing Agency-wide cumulative risk assessment guidance that builds from these ongoing
27 activities. As a first step, a technical panel convened under the Risk Assessment Forum has been
28 working to develop a Framework for Cumulative Risk Assessment. Building from the Agency's
29 growing experiences, this Framework is intended to identify the basic elements of the cumulative
30 risk assessment process. It should provide a flexible structure for the technical issues and define
31 key terms associated with cumulative risk assessment.
32
33 [This preliminary draft of the Framework for Cumulative Risk Assessment is being made available
34 at this time for the purpose of peer consultation. At the completion of the peer consultation
35 process, the document will be revised and then reviewed by the Agency's Science Advisory Board
36 (SAB). The final framework document will reflect the SAB comments and will require review
37 and approval by the Agency's Science Policy Council.]
38
39 William P. Wood
40 Director, Risk Assessment Forum
41
42
H-6
-------
DRAFT- Risk Assessment Forum Review Draft - August 2, 2001 - Do Not Quote or Cite
1 Preface
2
3 In the past several years, cumulative risk assessment, aggregate exposure assessment, and
4 research on chemical mixtures has taken on increased importance. This is underscored by recent
5 reports such as the National Research Council's 1993 report Pesticides in the Diets of Infants and
6 Children, (NRC, 1993) the 1994 NRC report Science and Judgment in Risk Assessment, (NRC,
7 1994), the 1995 National Academy of Public Administration report Setting Priorities, Getting
8 Results (NAPA, 1995), the 1997 report by the Presidential/Congressional Commission on Risk
9 Assessment and Risk Management titled Risk Assessment and Risk Management in Regulatory
10 Decision-Making (PCCRARM, 1997), and the EPA Science Advisory Board report Toward
11 Integrated Environmental Decision-Making (USEPA, 2000a). There also have been several
12 recent pieces of legislation that mandate the consideration of cumulative risk and variability
13 factors in the risk characterization process. Specifically, the Food Quality Protection Act of 1996
14 (FQPA) [PL 104-170, August 3, 1996] directs EPA in its assessments of pesticide safety to focus,
15 in part, on the cumulative effects of pesticides that have a common mechanism of toxicity,
16 considering aggregate dietary and non-occupational pathways of exposure.
17
18 Assessment of cumulative risk through complex exposures is one of the high priorities of
19 the Agency, especially in light of FQPA mandates, and is germane and of great interest to all
20 program and regional offices. This area of research is also directly applicable to children's risk
21 issues. This Framework is meant to lay out broad areas where analysis might be done if needed.
22 It does not suggest that cumulative risk assessment is a tool that should be used with every issue,
23 nor does it suggest that when cumulative risk assessment is applied, that all areas of analysis
24 outlined or discussed here must or even should be done in every assessment. The scope of the
25 assessment will define the areas to be analyzed. In some areas discussed in this Framework, the
26 methodology for doing the risk analysis may not yet exist.
27
28 According to the expert panel report Safeguarding the Future: Credible Science,
29 Credible Decisions (USEPA 1992a), a key role of science at EPA is to reduce uncertainties in
30 environmental decision-making. The report points out that while EPA historically has focused on
31 chemical-specific impacts, methods to assess or control the effects of chemical mixtures and
32 general stressors on human health and ecosystems remained to be developed. In Pesticides in the
33 Diets of Infants and Children, (NRC, 1993) the NRC recommended that all exposures to
34 pesticides—dietary and nondietary—need to be considered when evaluating the potential risks to
35 infants and children. Estimates of total dietary exposure should be refined to consider intake of
36 multiple pesticides with a common toxic effect. Further, the report identifies important
37 differences in susceptibility with age. NRC in Science and Judgment in Risk Assessment (NRC,
38 1994) states that health risk assessments should generally consider all possible routes by which
39 people at risk might be exposed, and recommends this approach universally in the assessment of
40 hazardous air pollutants regulated by EPA under the Clean Air Act Amendments of 1990 [P.L.
41 101-549, November 15, 1990]. Regarding variability, the NRC report recommended that EPA
42 assess risks to infants and children whenever it appears that their risks might be greater than those
H-7
-------
DRAFT- Risk Assessment Forum Review Draft - August 2, 2001 - Do Not Quote or Cite
1 of adults. Public criticisms documented in this report note that EPA does not often consider the
2 possibility of synergistic interactions when multiple chemical exposures occur, nor does it
3 consider extreme variability among individuals in their responses to toxic substances. A related
4 issue is the problem of how risks associated with multiple chemicals are to be combined. Finally,
5 the FQPA [P.L. 104-170, August 3, 1996], requires research on the influence of complex
6 exposures on non-cancer human health effects of pesticides and other toxic substances.
7
8 The issue of cumulative risk is also an important issue with the general public. In public
9 meetings of Superfund stakeholders, held in late 1996 in San Francisco and Washington, DC, and
10 in early 1998 in Atlanta, the issue of cumulative risk was raised several times in each session
11 (USEPA 1996a, USEPA 1998a).
12
13 There are over 20,000 pesticide products on the market (USEPA, 200Id), and over
14 80,000 existing chemicals on the TSCA inventory (USEPA, 2001e). Each year, an additional
15 number of chemicals are added. The question of how to assess the cumulative effect of these
16 chemicals on the population will be a great challenge to the Agency in the coming decade. This
17 issue may well become the primary issue in the risk assessment field in the next ten years.
18
19
H-8
-------
DRAFT- Risk Assessment Forum Review Draft - August 2, 2001 - Do Not Quote or Cite
1 Authors, Contributors, and Reviewers [to be completed]
2
3
H-9
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
List of A
ACGffl
AFS
AIChE
AfflA
AIRS
AMTIC
APCA
APEX
ARE
ATSDR
CARES
CBEP
CEQ
CFR
CHIEF
COHb
CRIA
DALY
DOT
EPA
FIFRA
FQPA
GAO
GIS
HAP
HEC
HRS
HUD
IED
ILSI
LADD
LDP
LLE
LOAEL
MOE
MSDS
NAAQS
NAPA
NATA-
NEPA
DRAFT- Risk Assessment Forum Review Draft - August 2, 2001 - Do Not Quote or Cite
List of Abbreviations and Acronyms
- American Conference of Government Industrial Hygienists
- AIRS Facility Subsystem
- American Institute of Chemical Engineers
- American Industrial Hygiene Association
- Aerometric Information Retrieval System
- Ambient Monitoring Technology Information Center
- American Crop Protection Association
- Air pollution exposure model
- Acute reference exposure
- Agency for Toxic Substances and Disease Registry
- Cumulative and Aggregate Risk Evaluation System
- Community-based environmental protection
- Council for Environmental Quality
- Code of Federal Regulations
- Clearinghouse for Inventories and Emissions Factors
- Carboxyhemoglobin
- Cumulative Risk Index Analysis
- Disability-adjusted life year
- United States Department of Transportation
- United States Environmental Protection Agency
- Federal Insecticide, Fungicide, and Rodenticide Act
- Food Quality Protection Act
- United States General Accounting Office
- Geographical Information System
- Hazardous air pollutant
- Human equivalent concentration
- Hazard Ranking System
- United States Department of Housing and Urban Development
- Integrated Environmental Decision-making
- International Life Sciences Institute
- Lifetime average daily dose
- Locational Data Policy
- Loss of life expectancy
- Lowest observed adverse effect level
- Margin of exposure
- Materials Safety Data Sheet
- National Ambient Air Quality Standards
- National Academy of Public Administration
National-Scale Air Toxics Assessment
- National Environmental Policy Act
H-10
-------
DRAFT- Risk Assessment Forum Review Draft - August 2, 2001 - Do Not Quote or Cite
1 NHEXAS - National Human Exposure Assessment Survey
2 NIOSH - National Institute for Occupational Safety and Health
3 NOAEL - No observed adverse effect level
4 NRC - National Research Council
5 OAR - Office of Air and Radiation (EPA)
6 OECA - Office of Enforcement and Compliance Assurance (EPA)
7 OPP - Office of Pesticide Programs (EPA)
8 OPPTS - Office of Prevention, Pesticides, and Toxic Substances (EPA)
9 ORD -Office of Research and Development (EPA)
10 OSWER - Office of Solid Waste and Emergency Response (EPA)
11 P.L. - Public Law
12 PAH - Polycyclic Aromatic Hydrocarbon
13 PCB - Poly chlorinated biphenyl
14 PCS - Permit Compliance System
15 PM-10 - Particulate matter of 10 micrometer size or less
16 pNEM - Probabilistic NAAQS Exposure Model
17 QALY - Quality-adjusted life year
18 RfC - Reference Concentration
19 RfD - Reference Dose
20 SAB - Science Advisory Board
21 SAP - Science Advisory Panel
22 SAR - Structure-activity relationship
23 SCRAM - Support Center for Regulatory Air Models
24 SHEDS - Stochastic Human Exposure and Dose Simulation model
25 SPC - Science Policy Council
26 TEAM - Total Exposure Assessment Methodology
27 TEMRAP - The European Multi-Hazard Risk Assessment Project
28 TIA - Transient ischemic attack
29 TRI - Toxic(s) Release Inventory
30 TRIM.Expo - Total Risk Integrated Methodology, exposure module
31 U.S.C. - United States Code
32 UF - Uncertainty factor
33 USEPA - United States Environmental Protection Agency
34
35
H-ll
-------
DRAFT- Risk Assessment Forum Review Draft - August 2, 2001 - Do Not Quote or Cite
1 Executive Summary [to be completed last]
2
H-12
-------
1
2
3
4
5
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
DRAFT- Risk Assessment Forum Review Draft - August 2, 2001 - Do Not Quote or Cite
1. INTRODUCTION
During much of its early history, EPA focused its efforts on cleaning up the overt
pollution problems of the 1960s and 1970s. Until EPA was established in 1970, relatively
uncontrolled air emission, water effluents, and dumping of wastes had led to pollution of the
environment that was easily detected by the five senses. The most effective and efficient way to
approach these overt problems of the 1970s was to find the entry point of the pollution into the
environment, and to keep it from entering the environment by controlling it there. Looking back,
we see a strategy that moved to control stack emission, industrial and municipal effluents,
pesticide application, land applications, burial of chemical wastes, and other so-called "sources" of
pollution. In addition, criteria and standards were established as goals for cleanup of the various
environmental media. By the 1980s, this so-called "command and control" strategy was well
established in environmental laws and regulations, but was reaching the point of diminishing
returns from a cost-benefit viewpoint.
The development of risk assessment methodology during the 1970s and early 1980s
closely followed the Agency's strategy for control of pollution, since risk assessments were being
used as one of the factors in EPA's decision-making for regulations to control pollution. The
focus on sources led naturally to analysis of what types of pollutants were in effluents, air
emissions, and waste sites. These were chemical,
biological, and sometimes radiological agents. By the
1970s, the links between some chemicals and certain
diseases such as cancer had been established through
a series of bioassays, or in the cases of chemicals like
vinyl chloride and asbestos, through epidemiological
studies. New analytical techniques of the 1970s also
made it possible to detect very minute concentrations
of chemicals for the first time. The focus of the EPA
strategy to control pollution (and the risk assessment
methodology being used to partially support
decisions) gradually leaned toward assessing and
controlling the individual chemicals. Congressional
legislation tended to underwrite this approach by
focusing on controlling sources and even including
lists of individual chemicals to be controlled.
Figure 1. Chemical (or stressor) focused
assessment starts with a source and
evaluates how the chemical gets to
various populations or ecological targets.
Individual assessments may choose to
pursue some or all pathways, media, or
population segments.
Risk assessment methodology of the 1970s
and early 1980s, for this reason, tended towards
single chemical assessments (see Figure 1). The 1983
National Research Council report Risk Assessment in
H-13
-------
DRAFT- Risk Assessment Forum Review Draft - August 2, 2001 - Do Not Quote or Cite
1
2
3
4
5
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
the Federal Government (NRC, 1983) was largely focused on the single chemical risk assessment
approach when it spoke of the four parts of a risk assessment: hazard identification, dose-
response assessment, exposure assessment, and risk characterization. EPA's 1986 Risk
Assessment Guidelines (USEPA 1986a), with the exception of the mixtures guidelines (USEPA,
1986b), were also largely focused on single chemical assessment.
Research done or sponsored by EPA in the
early 1980s, however, was taking the first steps
toward a different type of risk assessment
methodology, one that focused on the persons
exposed rather than the chemicals (Figure 2). The
goals of this second, population-based, approach
were much more useful to decision-makers who were
focusing on public health or ecological health
questions, rather than controlling sources of
pollution. The approach for the chemical-focused and
population-focused approaches depicted in the two
figures are quite different, even though some of the
tools to do the assessment may be the same.
Figure 2. Population-based assessments
start with the receptors, and determine
what chemicals, stressors, or other risk
factors are affecting them.
The challenges posed by the population-based
assessment can be daunting. Taken to the extreme,
Figure 2 represents a concept of "total risk" for the
population or population segment being evaluated,
with each chemical, biological, radiological, or other stressor adding some fraction of the total
risk. Looking at the problem from an individual stressor viewpoint, to do this type of assessment
would require not only evaluating each individual stressor, but developing a way to add up all the
risks among stressors across a population of individuals with different exposures, susceptibilities,
etc. In the early 1980s, the state of the science was unready for virtually any part of the methods
for doing this type of assessment.
But progress was being made toward developing a population-based methodology.
Starting in the late 1970s, a group of EPA researchers and contractors began developing what
would become the Total Exposure Assessment Methodology (TEAM) study (USEPA 1987).
TEAM measured the concentrations of a number of chemicals simultaneously at the point of
exposure. This led to a larger study, the National Human Exposure Assessment Survey
(NHEXAS) in the 1990s (Sexton, et, al. 1995). Both TEAM and NHEXAS were population-
based exposure assessment approaches which developed analytical tools and methodologies to do
this type of exposure assessment.
H-14
-------
DRAFT- Risk Assessment Forum Review Draft - August 2, 2001 - Do Not Quote or Cite
1 Also in the early 1980s, some progress was being made toward the question of how to add
2 the risks from different chemicals or stressors. The 1986 Risk Assessment Guidelines (USEPA,
3 1986a) included a guideline on chemical mixtures (USEPA, 1986b), which discussed how the
4 risks from multiple chemicals could be evaluated as a whole. The work on this guidance has
5 continued most recently with a draft chemical mixtures guidance document (USEPA, 200la)
6 which expands and supplements the 1986 beginnings.
7
8 About the same time the Agency made some progress on single chemical and chemical
9 mixture risk assessment with the 1986 Guidelines, some different kinds of risk assessment
10 problems began to catch the Agency's attention. In 1986, eleven Chicago-area community
11 groups joined together to file a petition under Section 21 of the Toxic Substances Control Act
12 asking for a community assessment in Southeast Chicago. A series of community-based actions
13 which started in 1982 and grew throughout the 1980s focused on disparities of risk among
14 various population subgroups, calling specific attention to cumulative effects of pollution on
15 minority subgroups (GAO, 1983; Lee, 1987). This series of community-based actions, chronicled
16 in the 1990 book Dumping in Dixie: Race, Class and Environmental Quality (Bullard, 1990)
17 eventually became known as the Environmental Justice movement. The issues raised by the
18 Environmental Justice movement were the basis of a 1994 Presidential Executive Order
19 [Executive Order 12898, February 11, 1994] which told Agencies, among other things, that
20 "Environmental human health analyses, whenever practicable and appropriate, shall identify
21 multiple and cumulative exposures." In the 1990s, Environmental Justice cases, including the
22 cases which have been filed under Title VI of the 1964 Civil Rights Act, [P.L. 88-3 52, July 2,
23 1964] have added to the demand that a population-based human health risk assessment
24 methodology be developed.
25
26 Even before Executive Order 12898 was issued, it was apparent that population-based
27 assessments were going to be needed, in addition to the chemical-based assessments, if EPA was
28 going to be able to answer the questions and issues being raised by the public. Community
29 spokespersons and other "stakeholders," as well as scientific panels, were increasingly coming to
30 the Agency with problems that demanded a multi-stressor, population-based approach (e.g., NRC
31 1994). Ecological problems, especially, were demanding a "place-based" context (such as the
32 Chesapeake Bay watershed) in which the various populations within the area were looked at from
33 a "total system" viewpoint. This place-based focus was a part of the 1992 Framework for
34 Ecological Risk Assessment (USEPA 1992b) and the 1998 Guidelines for Ecological Risk
35 Assessment. (USEPA 1998b)
36
37 Also by the early 1990s, it was becoming clear that the population-based assessments
38 being contemplated for EPA's cumulative risk needs and the type of assessments done under the
39 National Environmental Policy Act of 1969 (NEPA) were related. NEPA [P.L. 91-190, 42
40 U.S.C. 4321-4347, January 1, 1970, as amended by P.L. 94-52, July 3, 1975, P.L. 94-83, August
H-15
-------
DRAFT- Risk Assessment Forum Review Draft - August 2, 2001 - Do Not Quote or Cite
1 9, 1975, and P.L. 97-258, §4(b), Sept. 13, 1982], which was passed at about the same time EPA
2 was established, requires assessments on the impacts of federal or federally-funded projects (such
3 as roads, dams, power lines, military projects, and infrastructure development) on natural
4 ecosystems, endangered species, habitats, and opportunities for public enjoyment and natural
5 resource use. A primary concern for NEPA is "cumulative effects analysis," defined as "the
6 incremental impact of the action when added to other past, present, and reasonably foreseeable
7 future actions... Cumulative impacts result from individually minor but collectively significant
8 actions taking place over a period of time" (CEQ, 1997). Much of the NEPA cumulative effects
9 analysis is qualitative, but risk assessments and cause-and-effect relationships are key parts of the
10 analysis process for controversial projects.
11
12 The projects or actions that NEPA addresses can be viewed as sources of stressors.
13 Environmental impact assessment under NEPA contains a description of the affected environment
14 that contains four types of information: (1) data on the status of important natural, cultural, social,
15 or economic resources and systems; (2) data that characterize important environmental or social
16 stress factors; (3) a description of pertinent regulations, administrative standards, and
17 development plans; and (4) data on environmental and socioeconomic trends. In addition to health
18 effects on populations and susceptible individuals as part of the affected environment, the NEPA
19 cumulative effects analysis would consider effects on historic and archaeological resources,
20 socioeconomic factors like employment, human community structure, and quality of life changes.
21 These may be among the types of effects EPA may be asked to include in future cumulative risk
22 assessments. As EPA moves toward cumulative risk assessment, there is some parallel with the
23 NEPA methods for cumulative impact analysis, which may be applied to cumulative risk
24 assessments.
25
26 By the first decade of the twenty-first century, cumulative risk assessment needs have
27 become relatively common, especially in EPA's Regional Offices and in the Office of Civil Rights.
28 Much like the "place-based" ecological assessments, communities are asking for community-based
29 assessments which include human health risk assessments, ecological risk assessments, and
30 sometimes, assessments of "quality of life" factors. It is the demand for population-based human
31 health risk assessments that has driven the need for research into cumulative risk assessment,
32 aggregate exposure assessment, and risk from chemical mixtures.
33
34
35 1.1. Purpose and Scope of the Framework Report
36
37 An understanding of the finite purpose and scope of this Framework Report is important.
38 EPA, other regulatory agencies, and other organizations need detailed, comprehensive guidance
39 on methods for evaluating cumulative risk. Before such detailed Agency-level guidance can be
40 developed on a relatively new field of risk assessment, it has been the recent policy of the Agency
H-16
-------
DRAFT- Risk Assessment Forum Review Draft - August 2, 2001 - Do Not Quote or Cite
1
2
3
4
5
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
to first develop a simple framework as a foundation for later comprehensive guidance. This
Framework for Cumulative Risk Assessment will emphasize chemical risks to human health in its
discussion, but will do so in the context of the effects from a variety of stressors, including non-
chemical stressors. Some important topics that could be characterized as "cumulative risk", such
as global climate change, are beyond the scope of this Framework.
With this background, the Framework has two simple purposes, one immediate and one
longer term. As a broad outline of the assessment process, the Framework immediately offers a
basic structure and provides starting principles for EPA's cumulative risk assessments. The
process described by the Framework provides wide latitude for planning and conducting
cumulative risk assessments in many diverse situations, each based on common principles
discussed in the Framework. The process also will help foster a consistent EPA approach for
conducting and evaluating cumulative risk assessments, for identifying key issues, and for
providing operational definitions for terms used in cumulative risk assessments.
In the longer term, the Framework
offers the basic principles around which to
organize a more definitive set of
Cumulative Risk Assessment Guidance.
With this in mind, this report does not
provide substantive guidance on certain
things that are integral to the risk
assessment process (but see box at right).
These include specific analytical methods,
techniques for analyzing and interpreting
data, and guidance on issues influencing
policy. Rather, on the basis of EPA
experience and recommendations of peer
reviewers, EPA has reserved discussion of
these important aspects of cumulative risk
assessment for future Guidance, which will
be based on the risk assessment process
described in this Framework.
This Framework is meant to lay
out broad areas where analysis might be
done if needed. It does not suggest that
cumulative risk assessment is a tool that
should be used with every issue, nor does
it suggest that when cumulative risk
EPA's Risk Assessment Guidelines
Carcinogen Risk Assessment (USEPA1986d)
Mutagenicity Risk Assessment (USEPA 1986c)
Chemical Mixtures (USEPA 1986b)
Developmental Toxicity Risk Assessment (USEPA 1991b)
Exposure Assessment (USEPA 1992c)
Reproductive Toxicity Risk Assessment (USEPA 1996b)
Ecological Risk Assessment (USEPA 1998b)
Neurotoxicity Risk Assessment (USEPA 1998e)
Proposed Carcinogen Risk Assessment (USEPA 1996c)
Selected Policy and Guidance Documents
Risk Assessment Guidance for Superfund (USEPA 1989a)
Community Involvement in Superfund RA (USEPA 1999c)
Locational Data Policy (USEPA 199la)
Framework for Ecological Risk Assessment (USEPA 1992b)
Application of Refined Dispersion Models (USEPA 1993a)
Policy /Guidance for Risk Characterization (USEPA 1995ab)
Handbook for Risk Characterization (USEPA 2000c)
Cumulative Risk Planning and Scoping (USEPA 1997a)
Chemical Emergency Risk Management (USEPA 1998c)
Draft Comparative Risk Framework (USEPA 1998f)
Guideline on Air Quality Models (USEPA 1999a)
Framework for Community Based Env. Prot. (USEPA 1999b)
Guidance for Offsite Consequence Analysis (USEPA 1999d)
Handbook for Peer Review (USEPA 2000b)
Supplementary Guidance for Conducting Health Risk
Assessment of Chemical Mixtures (USEPA 2001)
Guiding Principles for Monte Carlo Analysis (USEPA 19997c)
H-17
-------
DRAFT- Risk Assessment Forum Review Draft - August 2, 2001 - Do Not Quote or Cite
1 assessment is applied, that all areas of analysis outlined or discussed here must or even should be
2 done in every assessment. The scope of the assessment will define the areas to be analyzed. In
3 some areas discussed in this Framework, the methodology for doing the risk analysis may not yet
4 exist.
5
6 1.2. Intended Audience
7
8 The framework is primarily intended for EPA risk assessors, EPA risk managers, and
9 other persons who either perform work under EPA contract or sponsorship or are subject to EPA
10 regulations concerning risk assessments. The terminology and concepts described here also may
11 be of assistance to other Federal, State, and local agencies as well as to members of the general
12 public who are interested in cumulative risk assessment issues. The style and language used in
13 this Framework document are chosen to be understood by as wide a variety of interested parties
14 as possible, from the policy maker to the risk assessment scientist to the concerned non-scientist
15 member of the general public. It is hoped that this Framework will be the first step in developing a
16 broad scientific consensus about cumulative risk assessment, and that further guidelines and
17 guidance will build upon this foundation.
18
19 1.3. Key Definitions in Cumulative Risk Assessment1
20
21 According to common English usage, "cumulative" means (Random House, 1966):
22
23 1. made up of accumulated parts; 2. increasing by successive additions; 3. tending to prove
24 the same point (e.g., cumulative evidence); 4. additional rather than repeated (e.g.,
25 cumulative legacy); 5. taking effect upon completion of another penal sentence (e.g.,
26 cumulative sentence); 6. increasing in severity with repetition of the offense (e.g.,
27 cumulative penalty); 7. formed by the addition of new material of the same kind (e.g.,
28 cumulative book index); 8. summing or integrating overall data or values of a random
29 variable less than or equal to a specified value (e.g., cumulative normal distribution or
30 cumulative frequency distribution)
31
32 The key concepts in the definitions are that of accumulation (gathering into a mass,
33 collecting, or heaping up) and integrating the accumulated parts into a whole. Cumulative Risk
34 Assessment, then, would examine the accumulation (over time, across sources, across routes,
35 etc.) of stressors that can cause adverse effects, and then integrate the effects these stressors
36 cause into a picture of the risk caused to the whole (individual or population) by the stressors
In this section, a few basic definitions related to cumulative risk assessment will be discussed. For a glossary of
terms, the reader is directed to Section 5.
H-18
-------
1
2
3
4
5
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
DRAFT- Risk Assessment Forum Review Draft - August 2, 2001 - Do Not Quote or Cite
acting together. Some examples of types of cumulative risk assessments are listed below. Each of
these presupposes a defined individual or population2:
1. Risks can be added or accumulated over time for a single agent or stressor across sources,
environmental pathways, or exposure routes. [This is consistent with "aggregate risk" in the
FQPA terminology in the box below.] A cumulative risk assessment of this type integrates effects
by considering differences and interactions related to routes, sources, and time patterns of
exposure. This is contrasted to a single chemical assessment which merely adds up exposures
across sources, routes, and time as if they all were equal, without regard to how or when they
occur, or how these differences affect the final risk result.
2. Risks can be accumulated over time (and
pathways, sources, routes, etc.) for a number
of agents or stressors causing similar types of
effects, e.g., a number of carcinogenic
chemicals or a number of threats to habitat
loss. Again, a cumulative risk assessment for
multiple stressors will take into consideration
the interactions among stressors, and attempt
to address the risks from the combined or
integrated insult, not merely list risks from
individual stressors separately in a table.
3. Risks can be accumulated across different
types of stressors causing different types of
effects, for example chemical, biological,
radiological, and physical stressors, causing
human health, ecological health, and "quality
of life" effects. This is considerably more
complex methodologically and
computationally than the types of cumulative
risk assessments in examples 1 and 2, above.
A cumulative risk assessment with multiple
types of stressors will address how these
stressors can be integrated into the overall
FQPA's Terminology Interpretations
The Food Quality Protection Act of 1996 [P.L. 104-170]
discusses the addition of exposure for a single chemical
across sources, pathways, routes, and time as aggregate
exposure. To be consistent with that terminology, the
Agency has elected to speak of multiple source/pathway/
route single stressor exposures and risks as "aggregate
exposures" and "aggregate risks." The EPA Science
Policy Council's Cumulative Risk Subcommittee has
developed the following working definitions for single-
chemical or single-stressor situations:
Aggregate exposure: The combined exposure of
an individual (or defined population) to a specific agent or
stressor via relevant routes, pathways, and sources.
Aggregate exposure assessment: An analysis,
characterization, and possibly quantification of exposure
of an individual (or defined population) to a specific agent
or stressor via relevant routes, pathways, and sources.
Aggregate risk: The risk resulting from
aggregate exposure to a single agent or stressor.
The Food Quality Protection Act also discusses
"cumulative effects" from different pesticides which act
by the same mechanism of action (or as interpreted, mode
of action).
Populations can be defined by geophysical boundaries, such as a watershed, geopolitical boundaries, such as city or
county limits, or by cultural, racial, economic, or other criteria within a certain geographic boundary such as a neighborhood.
The definition of a population needs to be clear enough so that it can be agreed upon whether any specific individual is "in" the
population or "out."
H-19
-------
DRAFT- Risk Assessment Forum Review Draft - August 2, 2001 - Do Not Quote or Cite
1
2
3
4
5
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
estimate of risk for the individual or population. For example, if one were doing a cumulative risk
assessment focusing on a wide variety of stressors to a certain (human) community's health, one
might also look at how changes in ecological health or quality-of-life in the area affect human
health risk.
As a note on #3, individual and community health status and the corresponding health
statistics are reflective of all stressors in the lives of the population, across all types of effects.
When attempting to compare health statistics from a certain area with the results of a narrower
cumulative risk assessment, this should be kept in mind. Combining different types of risks will
also require more than just an analytical process; it also requires a deliberative process. This will
be discussed more fully in Chapter 4.
We have used the key concepts of accumulation and integration to craft the following
definition:
cumulative risk assessment: The
examination of the accumulation (over
time, across sources, across routes,
etc.) of stressors or exposures that can
cause adverse effects, and then the
integration of the effects these
stressors or exposures cause into an
estimate and characterization of the
risk caused to the individual or
population by the stressors acting
together.
We believe that this is a broad
definition, but not so broad that any risk
assessment will fit. These definitions clearly
exclude assessments which examine a single pathway for a single chemical (no accumulation over
various sources, routes, time, different stressors), or even assessments which look at a number of
stressors but merely list stressor risks separately (no integration of how they act together to affect
overall risk).
We also believe that the definition used here is consistent with the sense of most
definitions of "cumulative" such as are included in NEPA, FQPA, or defined by other groups such
NEPA's "Cumulative Impact" Definition
CEQ Regulation 1508 for Implementing the National
Environmental Policy Act of1969 [P.L. 91-190, 42 U.S.C.
4321-4347, January 1, 1970, as amended by P.L. 94-52,
July 3, 1975, P.L. 94-83, August 9, 1975, and P.L.
97-258, §4(b), Sept. 13, 1982] defines "cumulative
impact" as "the impact on the environment which results
from the incremental impact of the action when added to
other past, present, and reasonably foreseeable future
actions regardless of what agency (Federal or
non-Federal) or person undertakes such other actions.
Cumulative impacts can result from individually minor
but collectively significant actions taking place over a
period of time."
Source: CEQ, 1997
H-20
-------
1
2
3
4
5
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
DRAFT- Risk Assessment Forum Review Draft - August 2, 2001 - Do Not Quote or Cite
as the EPA Science Policy Council's Cumulative Risk Subcommittee3.
1.4. The Cumulative Risk Assessment as a Tool for a Variety of Users and Purposes
Cumulative risk assessment is conceptually an analytic-deliberative process (NRC, 1996).
It includes both analytic (i.e., rigorous, replicable methods, evaluated under the agreed protocols
of an expert community) and deliberative (i.e., stakeholder-value-and-judgment based) parts.
Much of what is discussed in Chapter 2, the Planning and Problem Formulation Phase, is
deliberative in nature, which means it depends on input from experts other than those who know
how to do risk assessments. These include persons who are knowledgeable about a community
and its values. Although much of Chapter 3, the Analysis Phase, is given over to the analytic
process where risk assessment experts apply science to a problem, the deliberative aspect returns
in Chapter 4, the Interpretation Phase, especially where risks of different types are being evaluated
and combined.
Cumulative risk assessment, because of this
analytic-deliberative process, can be applied to
a variety of different problems where analysis
of the overall impacts of multiple sources,
stressors, chemicals, pathways, or routes is
necessary. It can be used as a regulatory
analysis tool, such as in reviewing the overall
impact of several different pesticides that all act
by the same mode of action (ILSI, 1999), or in
NEPA analyses (CEQ, 1997). It can be used to
analyze the overall impacts of permit decisions
or the results of compliance with permits in a
given community.
The Core Principles of Community-Based
Environmental Protection (CBEP)
1. Focus on a definable geographic area.
2. Work collaboratively with stakeholders.
3. Assess the quality of all resources in a place.
4. Integrate environmental, economic, and social
objectives.
5. Use the most appropriate tools.
6. Monitor and redirect efforts through adaptive
management.
Source: USEPA, 1999b
Cumulative risk assessment can also be
used in a community-based assessment approach, such as is outlined in EPA's Framework for
Community-Based Environmental Protection (USEPA, 1999b). The CBEP approach (see box
above) encompasses both ecological and human health assessments, and Cumulative risk
assessment, being a population-based or place-based analytic-deliberative process, is ideal for
CBEP-type applications.
The Council Cumulative Risk Subcommittee has developed the following working definitions for cumulative risk,
which incorporate both the accumulative and integrative aspects of cumulative risk assessment: Cumulative Risk: The
combined risks from aggregate exposures to multiple agents or stressors. Cumulative risk assessment: An analysis,
characterization, and possible quantification of the combined risks to health or the environment from multiple agents or
stressors.
H-21
-------
DRAFT- Risk Assessment Forum Review Draft - August 2, 2001 - Do Not Quote or Cite
1
2 Cumulative risk assessment is also applied in ecological assessments. The definition of
3 cumulative ecological risk assessment, as given in the EPA's 1998 Guidelines for Ecological Risk
4 Assessment is: A process that involves consideration of the aggregate ecological risk to the target
5 entity caused by the accumulation of risk from multiple stressors (USEPA, 1998b). A recent
6 Society of Environmental Toxicology and Chemistry publication (Foran and Ferenc, 1999)
7 discusses multiple stressors in ecological risk assessment, and gives a good overview of the topic
8 of cumulative ecological risk assessment.
9
10 1.5. The Broader Decision-Making Context for Cumulative Risk Assessment
11
12 Although it is possible to use cumulative risk assessment for research, that is, to form
13 hypotheses and test them by analyzing data, it is far more likely that cumulative risk assessment
14 will be used as a tool in decision making.
15
16 Decisions can be at a wide variety of levels, from a neighborhood group evaluating ways
17 to improve or safeguard their health and environment, to a Federal official weighing options for
18 action at a much broader geographical level. Although the decision-making method is beyond the
19 scope of this Framework, such decisions usually involve more than the basic science and analysis
20 that make up the "scientific" part of risk assessment. Robert T. Clemen, in his book Making Hard
21 Decisions notes that in one type of decision-making approach (called decision analysis):
22
23 Managers and policy makers frequently complain that analytical procedures from
24 management science and operations research ignore subjective judgments. Such
25 procedures often purport to generate "optimal" actions on the basis of purely objective
26 inputs. But the decision-analysis approach allows the inclusion of subjective judgments. In
27 fact, decision analysis requires personal judgments: they are important ingredients for
28 making good decisions. (Clemen, 1996, page 5)
29
30 Regardless of the type of decision being made or the decision-making approach, a
31 cumulative risk assessment's analytic part is not the decision-making vehicle in itself, that is,
32 "cranking out the numbers" will not be the sole basis for a decision. Although in some cases, the
33 estimated risks can weigh heavily in the decision, understanding the risk estimate is but one factor
34 in a broader decision-making process. The U.S. EPA's Science Advisory Board (SAB) in their
35 August, 2000, publication Toward Integrated Environmental Decision-Making (USEPA, 2000a),
36 constructed a framework for what they termed Integrated Environmental Decision-making (LED).
37 They noted that "The LED Framework recognizes that risks often are experienced simultaneously
38 and are cumulative...". They speak of risk assessments in a very broad way, including human
39 health effects, ecological effects, and quality-of-life effects. The first two phases of the IED,
40 "Problem Formulation" and "Analysis and Decision-making" essentially correspond to the three
H-22
-------
DRAFT- Risk Assessment Forum Review Draft - August 2, 2001 - Do Not Quote or Cite
1 phases we discuss in this Framework for Cumulative Risk Assessment. The SAB's third phase,
2 "Implementation and Performance Evaluation," is beyond the scope of this framework.
3
4
5 The SAB's report (USEPA, 2000a) gives a good insight into the broader context for
6 cumulative risk assessment, and some of the aspects of the analytic-deliberative parts of the
7 assessment. These will be discussed in Chapters 2-4, as these phases of the cumulative risk
8 assessment process are examined.
9
10 The 1996 book Understanding Risk (NRC, 1996) also provided much information on the
11 analytic-deliberative aspects of a risk assessment, and devoted a great deal of discussion to risk
12 characterization. Needless to say, it is very important to apply cumulative risk assessment in the
13 context of the decision or decisions to be made. This is most efficiently done by early and
14 continued attention to the "risk characterization" step in the risk assessment process (NRC, 1996;
15 USEPA, 2000c). The box on the following page summarizes some of the points made in
16 Understanding Risk.
17
18
19 1.6. Organization of this report
20
21 This report is organized to follow the general process steps for a cumulative risk
22 assessment, namely a planning and problem formulation phase (Chapter 2), an analysis phase
23 (Chapter 3), and a synthesis and interpretation phase, where the Risk Characterization is
24 completed (Chapter 4). Chapter 5 is a glossary of terms, followed by References in Chapter 6. For
25 certain topics throughout this Framework, a more in-depth discussion was warranted than the
26 main text would conveniently allow; these have been placed in several Appendices.
H-23
-------
DRAFT- Risk Assessment Forum Review Draft - August 2, 2001 - Do Not Quote or Cite
Some Thoughts on Risk Characterization
The NRC book Understanding Risk (NRC, 1996) has
risk characterization as its primary focus. In their
conclusions, NRC states:
1. Risk characterization should be a decision-driven
activity, directed towards informing choices and solving
problems. The view of risk characterization as a
translation or summary is seriously deficient.... Risk
characterization should not be an activity added at the
end of risk analysis; rather, its needs should largely
determine the scope and nature of risk analysis.
2. Coping with a risk situation requires a broad
understanding of the relevant losses, harms, or
consequences to the interested and affected parties. A
risk characterization must address what the interested
and affected parties believe to be at risk in the
particular situation, and it must incorporate their
perspectives and specialized knowledge.
3. Risk characterization is the outcome of an analytic-
deliberative process. ... Analysis and deliberation can
be thought of as two complementary approaches to
gaining knowledge about the world, forming
understandings on the basis of knowledge, and reaching
agreement among people.
4. The analytic-deliberative process leading to a risk
characterization should include early and explicit
attention to problem formulation.
5. The analytic-deliberative process should be mutual
and recursive. ... A recurring criticism of risk
characterization is that the underlying analysis failed to
pay adequate attention to questions of central concern to
some of the interested and affected parties. This is not
so much a failure of analysis as a failure to integrate it
with broadly based deliberation: the analysis was not
framed by adequate understanding about what should
be analyzed. ... Structuring an effective analytic-
deliberative process for informing a risk decision is not
a matter for a recipe. Every step involves judgment,
and the right choices are situation dependent. Still, it is
possible to identify objectives that also serve as criteria
for judging success:
Getting the science right. The underlying analysis
meets high scientific standards in terms of
measurement, analytic methods, data bases used,
the magnitude and character of uncertainty...
Getting the right science. The analysis has addressed
the significant risk-related concerns of public officials
and the spectrum of interested and affected parties,
such as risks to health, economic well-being, and
ecological and social values, with analytic priorities
having been set so as to emphasize the issues most
relevant to the decision.
Getting the right participation. The analytic-
deliberative process has had sufficiently broad
participation to ensure that the important, decision-
relevant information enters the process, that all
important perspectives are considered, and that the
parties' legitimate concerns about inclusiveness and
openness are met.
Getting the participation right. The analytic-
deliberative process satisfies the decision makers and
interested and affected parties that it is responsive to
their needs: that their information, viewpoints, and
concerns have been adequately represented and taken
into account; that they have been adequately consulted;
and that their participation has been able to affect the
way risk problems are defined and understood.
Developing an accurate, balanced, and informative
synthesis. The risk characterization presents the state of
knowledge, uncertainty, and disagreement about the
risk situation to reflect the range of relevant knowledge
and perspectives and satisfies the parties to a decision
that they have been adequately informed within the
limits of available knowledge.
6. Those responsible for a risk characterization should
begin by developing a diagnosis of the decision
situation so that they can better match the analytic-
deliberative process leading to the characterization to
the needs of the decision, particularly in terms of level
and intensity of effort and presentation of parties. ...
Diagnosis of risk decision situations should follow eight
steps: diagnose the kinds of risk and the state of
knowledge, describe the legal mandate, describe the
purpose of the risk decision, describe the affected
parties and anticipate public reactions, estimate
resource needs and timetable, plan for organizational
needs, develop a preliminary process design, and
summarize and discuss the diagnosis with the
H-24
-------
DRAFT- Risk Assessment Forum Review Draft - August 2, 2001 - Do Not Quote or Cite
1 2. THE PLANNING AND PROBLEM FORMULATION PHASE
2
3 The first step in any risk assessment process is to define the problem to be assessed. This
4 step has been called "problem formulation" in the Framework for Ecological Risk Assessment
5 (USEPA, 1992b), the NRC book Understanding Risk (NRC, 1996), Toward Integrated
6 Environmental Decision-Making (USEPA, 2000a) and elsewhere (e.g., USEPA, 1997a). It is a
7 phase where, according to NRC, "public officials, scientists, and interested and affected parties
8 clarify the nature of the choices to be considered, the attendant hazards and risks, and the
9 knowledge needed to inform the choices" (NRC, 1996) . Planning and Scoping of the assessment
10 are often thought of as being part of the Problem Formulation phase, although the 1997 Planning
11 and Scoping guidance treats Planning and Scoping as a separate activity before problem
12 formulation begins (USEPA, 1997a). Whether it is considered a separate phase or not, it takes
13 place at the very start of the process of doing a cumulative risk assessment. For convenience, this
14 section incorporates both Planning and Scoping and Problem Formulation into the Planning and
15 Problem Formulation Phase.
16
17 2.1. Planning and Scoping
18
19 Risk assessments are done within some context, that is, they are usually done because of a
20 regulatory requirement, a community need, a health crisis, or some other "driving force." This
21 context generates individuals or groups with interest in having the assessment done. They may be
22 public officials, risk experts, community leaders, or any number of other "interested parties."
23 Planning and scoping begins with a dialogue among these interested parties.
24
25 Among these interested parties, there will be a person or a group of people charged with
26 making decisions about how a risk may be mitigated, avoided, or reduced. For the sake of
27 simplicity, we will call this person or group the "risk manager," and for ease of discussion, will
28 discuss the risk manager as if it were a single person.
29
30 Also among the interested parties is a person or group of persons expert at doing the
31 scientific part of risk assessment. Sometimes called the "risk experts," we have chosen here to call
32 this person or group the "risk assessor." The risk assessor is usually responsible for getting the
33 risk assessment done, by analyzing the probability of adverse effects from stressors. In fact, due
34 to the complex nature of cumulative risk assessments, the "risk assessor" in most cumulative risk
35 assessments will involve a multi-disciplinary team of scientists, engineers, economists, computer
36 experts, statisticians, and other experts.
37
38 As part of the initial discussions concerning the need for a risk assessment, other
39 "interested and affected parties" besides the risk manager and risk assessor may help define
40 purpose, scope, and approach. The risk manager, risk assessor, and interested and affected
H-25
-------
DRAFT- Risk Assessment Forum Review Draft - August 2, 2001 - Do Not Quote or Cite
1 parties, if any, make up the "risk assessment planning team." In the initial phase, "risk assessors,
2 risk managers, and interested and affected parties seek agreement through extensive dialogue and
3 discussion on what analytical and deliberative steps need to be taken by whom, by when, and why
4 -- if not how." (USEPA, 2000a)
5
6 In the SAB's report Toward Integrated Environmental Decision-Making, they explain
7 some of the roles of the various participants on the risk assessment planning team during the
8 Planning and Problem Formulation phase:
9
10 Scientists play an important role in [this phase] by collecting, analyzing, and presenting
11 data in such a way that all parties can appreciate the type and magnitude of the problem(s)
12 under discussion. This activity will generally involve all four parts of risk assessment,
13 including assessment of exposures experienced by special populations and/or ecological
14 resources. Planning, scoping, and screening — including selection of endpoints of concern
15 - also requires explicit input of societal values and stakeholder participation. For
16 instance, while some of the ecological endpoints may be chosen because of their role in a
17 valued ecosystem, there may also be ecological endpoints chosen because of their direct
18 significance to society. Examples of the latter include both economically important species
19 and "charismatic" species. Similarly, in integrated decision-making, judgments may have
20 to be made about diverse health endpoints, such as cancer risks in the general population
21 and the risk of reproductive/developmental risks in children. While scientists can help
22 characterize such risks, they are not uniquely qualified to set priorities among them and
23 broader deliberation is essential. Finally, decision-makers also play an important role
24 during problem formulation; in addition to bringing the scientific and other resources of
25 the Agency to bear on the problem, they also should help to identify the range of potential
26 decisions and viable management options, while examining economic, political, or other
27 constraints on those options. Decision-makers also serve as managers of the overall
28 process. (USEPA, 2000a)
29
30 Another role of the risk assessment planning team is documentation. The activities of the
31 following sections are important, and should be documented by the team for several reasons.
32 Written records can be referred to by assessors and people at public meetings. They can also help
33 prepare for response to comments, and begin establishing a peer-review record for any later
34 decisions or plans that need to be peer reviewed (USEPA, 2000b).
35
36
37
38
39
40
H-26
-------
1
2
3
4
5
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
DRAFT- Risk Assessment Forum Review Draft - August 2, 2001 - Do Not Quote or Cite
2.1.1. Defining the Purpose of the Assessment
As discussed in section 1.5 above, the risk assessment should be developed to inform the
risk management decision by constructing an appropriate, decision-relevant risk characterization.
After the risk assessment planning team is assembled, the risk manager must explain why the
assessment is being performed, and what questions need to be answered. If interested and
affected parties are part of the risk assessment planning team, it is especially important that the
entire team agree on the purpose of the assessment, since differing sense of purpose among the
team will lead to problems later.
The list of questions to be addressed
(and hopefully answered) may influence the
management goals (see box at right), risk
management options, key participants, data
sources, selection of assessment endpoints,
approach, and the schedule for developing the
assessment.
The manager and assessment planning team
must discuss any regulatory or legal basis for
the risk assessment, and what kind of
information is needed to satisfy such
requirements.
The previous discussion follows the
typical situation where the risk manager is
presented as an independent decision-maker,
such as a senior official in a regulatory agency
who is responsible for establishing permit
conditions for a facility of some type. There
are situations, however, where the risk
manager may be one of the interested parties, such as a local citizens' board. For example, the
risk assessment may indicate that mitigation of risks may not be significantly affected by any
permit decisions but will depend instead on local zoning decisions or on decisions which affect
traffic patterns in a community. This is one of the reasons why in the final step in the planning
and problem formulation phase, the discussion of possible outcomes (discussed in section 2.3), is
so important.
Possible Management Goals
The goals of risk management are varied. They may
be risk related, aiming to:
• Reduce or eliminate risks from exposure to
hazardous substances.
• Reduce the incidence of an adverse effect.
• Reduce the rate of habitat loss.
They may be economic, aiming to:
•Reduce the risk without causing job loss.
•Reduce the risk without reducing property values.
They may involve public values, aiming to:
• Protect the most sensitive population.
• Protect children.
• Preserve a species from extinction.
Source: Presidential/Congressional Commission..., 1997
H-27
-------
DRAFT- Risk Assessment Forum Review Draft - August 2, 2001 - Do Not Quote or Cite
1 2.1.2. Defining the Scope of Analysis and Products Needed
2
3 Scoping a cumulative risk assessment effort defines the elements that will or will not be
4 included in the risk assessment4 (USEPA, 1997a). These include the stressors, sources,
5 pathways, routes, and populations to be evaluated. Initially, the risk assessment planning team
6 needs to select the kind of risk information, exposure scenarios and assessment issues that need to
7 be covered. These should be directly linked to the risk-related questions being asked when
8 establishing the purpose. Limitations in scope can be geographical (such as political or ecological
9 boundaries), environmental (such as assessing only certain media), demographic (such as
10 assessing only risks to children or asthmatics), statutory, or by using other criteria such as data
11 limitations. The issue of "background" exposures should be discussed and agreements reached
12 (see Appendix E). An adequate assessment scope should make it clear what's included and what's
13 excluded from the assessment. Care must be taken to reconcile the limitations of scope with the
14 list of questions to be answered in the statement of purpose. If, for example, data limitations
15 preclude the addressing of certain of the questions outlined in the purpose, the list of questions to
16 be addressed must be modified and the risk assessment planning team agree to the narrower scope
17 of the assessment.
18
19 Reasons for choosing the particular scope of the assessment, and how it will address the
20 questions posed in the purpose statement, must be stated explicitly. Defining the scope of the
21 assessment should include details on limitations on resources, limitations of data, the impact of
22 risk elements on the risk estimate (i.e., some pathways may be seen as having negligible impact on
23 the risks related to the questions being addressed), and methods available. In cases where an
24 element of risk is likely to be important, but no valid data are available, the assessor must
25 highlight this deficiency or use judgment or assumed values to approximate the missing data.
26 Such judgments and approximations should be clearly documented, and explained to the manager
27 in the risk characterization.
28
29 Once the elements (sources, stressors, populations, etc.) have been identified through
30 brainstorming with all participants, the participants should discuss the need and availability of
31 technical information and how such information may affect the overall uncertainty of the
32 assessment. Using input from the risk assessor, the risk assessment planning team must determine
33 what elements will and will not (or, can and cannot) be included in the risk assessment.
34 Information gathered at this stage is preliminary and may be modified during the analysis phase.
35 Identification of potential stressors, populations to be assessed, and potential effects are all part of
36 the scoping process, and help define the method of approach.
37
An assessment which looks at all stressors over a period of time for a specific population would be a "total risk"
assessment, which is difficult to perform given our current methods.
H-28
-------
DRAFT- Risk Assessment Forum Review Draft - August 2, 2001 - Do Not Quote or Cite
1
2
3
4
5
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
As examples of some of these scoping elements, stressors can include physical (including
radiological) stressors or chemical or biological agents that may cause an adverse effect. The
sources of the stressors can be human activities in sectors of society (e.g., manufacturing,
transportation, agriculture, land development), personal human activities (e.g., smoking, other so-
called "lifestyle activities") or natural phenomena. Stressors that are not physical, chemical, or
biological, such as economic or other quality-of-life stressors may also be identified.
Possible population elements to be assessed usually focus on the entities that are at risk,
e.g., populations, communities, ecosystem functions, or vulnerable subpopulations such as
persons with certain diseases, or persons at vulnerable life stages, such as children. The more
specifically defined these can be, the more focused the analysis can be. This will be helpful in
interpreting results of the assessment.
2.1.3. Agreeing on participants, roles and responsibilities
The risk assessment planning team will usually recommend others who should participate
in the assessment's planning, scoping, and risk analysis phase. Depending on the schedule,
approach, and level of effort envisioned for the risk assessment, there may be no additional
participants, or there may be many. Assessments will usually require substantial technical
expertise in the analytic portions of the assessment. Some of the fields of science that may be
needed or helpful include toxicology, epidemiology, ecology, risk assessment, exposure
assessment, fate and transport modeling of various sorts (e.g., indoor and outdoor air, surface and
drinking water), computer science (including geographical information systems [GIS]), chemistry,
biology, various engineering fields (e.g., chemical, mechanical, industrial, civil), economics,
sociology, and others.
For the deliberative portions of the
assessment, there can be a number of
stakeholders and other interested parties that
should be considered for participation. The
box at the right lists some examples to choose
from among interested or affected parties for
the deliberative portions of the assessment.
For community-based assessments, in
particular, it is important that community
involvement be sought and encouraged. The
Presidential/Congressional Commission on
Risk Assessment and Risk Management
Examples of Possible Interested or Affected
Parties (Stakeholders) (adapted from USEPA
1999b)
State governments
Tribal governments
Local governments
Community groups
Grassroots organizations
Environmental groups
Consumer rights groups
Religious groups
Civil rights groups
Civic organizations
Business owners
Trade associations
Labor unions
Public health groups
Academic institutions
Cooperative ext. progs.
Impacted citizens
Other federal agencies
H-29
-------
DRAFT- Risk Assessment Forum Review Draft - August 2, 2001 - Do Not Quote or Cite
1
2
3
4
5
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
[hereafter, the "Commission"] (1997) suggests the following questions to identify potential
interested or affected parties (stakeholders):
• Who might be affected by the risk management decision? (This includes not only groups
that already know or believe they are affected, but also groups that may be affected but as
yet do not know it.)
• Who has information and expertise that might be helpful?
• Who has been involved in similar
risk situations before?
• Who has expressed interest in being
involved in similar decisions before?
• Who might be reasonably angered if
not included?
It has become increasingly recognized
as important that stakeholders be involved in
risk assessment (e.g., NRC 1996,
Presidential/Congressional Commission...
1997, USEPA 1996a, 1997a, 1998a, 1999b,
1999c, 2000a). The Commission suggested
guidelines for stakeholder involvement (see
box at right).
There are several issues concerning
the stakeholders' capacity to participate that
should not be overlooked by the risk
assessment planning team. First, some
stakeholders may need training to be able to
participate in technical and risk management
discussions. Second, as noted in the box at
right, some stakeholders may require
incentives such as travel funds or lodging at
sites of meetings outside the area where they
live. The risk assessment planning team,
along with the potential source of funds for
such incentives, should decide to what extent,
Guidelines for Stakeholder Involvement
• Regulatory agencies or other organizations
considering stakeholder involvement should be clear
about the extent to which they are willing or able to
respond to stakeholder involvement before they
undertake such efforts. If a decision is not
negotiable, don't waste stakeholders' time.
• The goals of stakeholder involvement should be
clarified at the outset and stakeholders should be
involved early in the decision-making process.
Don't make saving money the sole criterion for
success or expect stakeholder involvement to end
controversy.
• Stakeholder involvement efforts should attempt
to engage all potentially affected parties and solicit a
diversity of perspectives. It may be necessary to
provide appropriate incentives to encourage
stakeholder participation.
• Stakeholders must be willing to negotiate and
should be flexible. They must be prepared to listen
to and learn from diverse viewpoints. Where
possible, empower stakeholders to make decisions,
including providing them with the opportunity to
obtain technical assistance.
• Stakeholders should be given credit for their
roles in a decision, and how stakeholder input was
used should be explained. If stakeholder suggestions
were not used, explain why.
• The nature, extent, and complexity of stakeholder
involvement should be appropriate to the scope and
impact of a decision and the potential of the decision
to generate controversy.
H-30
-------
DRAFT- Risk Assessment Forum Review Draft - August 2, 2001 - Do Not Quote or Cite
1 if any, such incentives can be provided, based on the scope, level of effort, and financial
2 constraints of the risk assessment project.
3
4 Roles and responsibilities for technical and non-technical participants (i.e., ground rules
5 for participants) should also be proposed by the planning team, depending upon the schedule,
6 approach, and level of effort that is envisioned for the risk assessment. There will be several key
7 points in the risk assessment process where stakeholder input will be critical. Some of these are
8 the agreements on purpose, scope, and approach. Each project should define and agree upon a list
9 of critical points for stakeholder input.
10
11 In spite of increased emphasis on stakeholder participation, however, there are instances
12 where it may not be appropriate for large scale stakeholder involvement. EPA (as the decision
13 maker) must determine whether stakeholder involvement in a cumulative risk decision will be
14 useful and what objectives it may accomplish to plan the public involvement process. There is a
15 continuum of objectives that may apply to individual cases, from exchanging information on one
16 end, through obtaining stakeholder recommendations, to developing agreements for joint
17 activities at the other end (EPA, 1998). Sometimes citizens choose not to participate because
18 they feel it will not influence the outcome, the issue is too complex or technical, the effort is too
19 great, or because the decision process is unclear (EPA, 200Ib).
20
21 2.1.4. Agreeing on the Depth of the Assessment and the Analytical Approach
22
23 The analysis approach (discussed further in section 2.2.3 and chapter 3) may fall anywhere
24 on a continuum from relatively simple methods which rely heavily on conservative simple
25 assumptions, and consequently have greater uncertainty, to increasingly refined assessments in
26 which data are substituted for assumptions and uncertainty is reduced. Some of the factors that go
27 into deciding on the approach include what level of uncertainty in the risk estimates is acceptable
28 to the participants, the intended use and audience, time and money resources available, and the
29 amount, quality and accessibility of data. In making the decision on approach, there will need to
30 be an understanding of both the level of effort necessary for conducting an assessment of the sort
31 selected, with an insight to alternatives, and the features and limitations of the selected approach,
32 in comparison to other approaches.
33
34 2.1.5. Agreement on the Resources Available and Schedule
35
36 Schedule and resources are often interrelated. They may also affect whether the work is
37 performed in-house by the organization or team desiring the assessment, or by contractor or other
38 external source. The need to meet external deadlines or coordinate with schedules of other
39 organizations may become an overriding factor in defining what will be prepared. Assessments
40 requiring short-term, low budget efforts, or preliminary screening assessments, may not have the
H-31
-------
DRAFT- Risk Assessment Forum Review Draft - August 2, 2001 - Do Not Quote or Cite
1 scope, time or resources where extensive stakeholder involvement is necessary or beneficial. For
2 assessments, especially those where there is extensive stakeholder involvement, a budget and time
3 schedule should be developed and known by all participants.
4
5
6 2.2. Problem Formulation, Conceptual Model, and Analysis Plan
7
8 Problem formulation is an initial part of the cumulative risk assessment. The outcome of
9 the problem formulation is a conceptual model that describes how a given stressor might affect
10 health or ecological components of the assessment. The conceptual model serves as a basis for the
11 analysis plan, which is used to focus the analysis phase of the assessment. These three components
12 are discussed in the sections below.
13
14 2.2.1. Problem Formulation.
15
16 Problem formulation is a systematic planning step that identifies the major factors to be
17 considered in a particular assessment. It is linked to the regulatory and policy context of the
18 assessment. Problem formulation is an iterative process within which the risk assessor develops
19 preliminary hypotheses about why adverse effects might occur or have occurred. It provides the
20 foundation for the technical approach of the assessment. The outcome of the problem formulation
21 process is a conceptual model that describes the relationship between the stressors, the population
22 exposed, and the assessment endpoint that will be addressed in the risk assessment.
23
H-32
-------
DRAFT- Risk Assessment Forum Review Draft - August 2, 2001 - Do Not Quote or Cite
Conceptual Model
with Examples of Possible Elements and Linkages
Sources I Stressors
I Pathways
Receptors
SurfaceWater
Ingestion/
contact
Air/inside and
out side-
(inhalation/
contact)
Terrestrial/
Habitat Alter.
Groundwater/
Ingestion
h
^
Aquatic
Ecosystem
Terrestrial
Ecosystem
Wetlands
Ecosystem
Human Health
Infants
Sensitive
Populations
Occupational
General
Population
Minorities/
EJ Populations
h. Q^/>i^ T7/>/^«
System
i-j
-j
^-J
-j
U
k
Figure 3. Generic conceptual model (from USEPA, 2001c).
I Endpoints
Ecological Endpoints
Condition of Aquatic
Ecosystem
Condition of
Terrestrial/Wetlands
Ecosystems
• Populations
•environmental
processes
Critical species or
species of special
concern
Human/Societal Endpoints
Cancer
Heart Disease
Respiratory disease
Immunologic disease
Gastrointestinal dis.
Psych, condition
Farm culture change
Demographic change
Comm. Infrastructure
Aesthetics
Loss of recreation
Property values
1
2
3
4
5
9
10
2.2.2. Developing the Conceptual Model
A conceptual model is both a written description and a visual representation of actual or
predicted relationships between humans (or populations, population segments) or ecological
entities and the chemicals or other stressors to which they may be exposed. Conceptual models
represent many relationships, and may describe primary, secondary, or tertiary exposure
pathways. The model is developed by the risk assessor and may include input from other experts
(including stakeholders). The model needs to distinguish between what is known or determined
and what is assumed. Also, it needs to include a discussion of uncertainties in the formulation of
the assessment. In some cases, conceptual models will be submitted for peer review. A general
H-33
-------
1
2
3
4
5
9
10
11
12
13
14
15
16
17
DRAFT- Risk Assessment Forum Review Draft - August 2, 2001 - Do Not Quote or Cite
conceptual model is provided in Figure 3 (previous page).
The conceptual model and the associated narrative show the basic rationale for the
decisions made in pursuing a particular course of action in a cumulative risk assessment. It
provides a record of decisions for future reference during risk analysis, characterization, and
communication of the risk management decision. It is also valuable as a risk communication tool
both internally within the Agency and externally in interactions with the public. The conceptual
model provides a scientific or technical work product that includes: (1) the scientific rationale for
the selection of the stressors, sources, receptors, exposed populations, exposure or environmental
pathways, assessment endpoints, and measurement endpoints; (2) the scientific, technical,
economic, or sociologic basis for the construction of the conceptual model; and (3) the scientific
implications of additional data gathering.
The Science Advisory Board in their report Toward Integrated Environmental Decision-
Making (USEPA, 2000a) suggests a list of desired outputs from Problem Formulation. These
should not only be left to the visual presentation of the Conceptual Model Diagram, but should
also be explained in narrative form. They are listed in the box below.
Desired Outputs for Problem Formulation
The initial goals for the decision-making exercise, including environmental goals to be achieved
Which environmental problems/stressors/systems will be included and which will not, and the reasons for these decisions
The health, ecological, and quality-of-life effects of concern
The spatial, temporal, and organizational dimensions of the problem
Relevant data and models, and possible approaches to data analysis
Scoping of the uncertainties involved and research needed to significantly reduce critical uncertainties
Initial review of the range of options available to reduce risks, considering likely economic, political, or other constraints
The endpoints upon which the condition of the ecological, human health, or societal systems ultimately will be judged
The types of factors that will be considered when reaching a decision
From Toward Integrated Environmental Decision-Making (USEPA, 2000a)
18 2.2.3. Constructing the Analysis Plan
19
20 The analysis plan is in the final stages of planning and scoping before the risk assessment
21 analysis phase is performed. The analysis plan includes pathways and relationships identified
22 during planning and scoping that will be pursued during the risk analysis phase. Those hypotheses
23 considered more likely to contribute to risk are given priority. The rationale for selecting and
24 omitting risk hypotheses should be incorporated into the plan and included discussion of data gaps
25 and uncertainties. It also may include a comparison between the level of confidence needed for
26 the management decision with that expected from alternative analyses in order to determine data
H-34
-------
DRAFT- Risk Assessment Forum Review Draft - August 2, 2001 - Do Not Quote or Cite
1 needs and evaluate which analytical approach is best. When new data are needed, the feasibility
2 of obtaining them can be taken into account.
3
4 In situations where data are few and new data cannot be collected, it may be possible to
5 extrapolate from existing data. Extrapolation allows the use of data collected from other
6 locations or organisms where similar problems exist. When extrapolating from data, it is
7 important to identify the source of the data, justify the extrapolation method, and discuss
8 recognized uncertainties.
9
10 A phased, or tiered, risk assessment approach can facilitate management decisions in cases
11 involving minimal data sets. However, where few data are available, recommendations for new
12 data collection should be part of the analysis plan. When new data are needed and cannot be
13 obtained, relationships that cannot be assessed are a source of uncertainty and should be described
14 in the analysis plan and later discussed in risk characterization.
15
16
17 2.3. The Final Step Before the Analysis Phase: Discussion of Possible Outcomes
18
19 It is useful for the entire team to hold some preliminary discussions, before the analytical
20 efforts of the assessment are started, about the various possibilities of the cumulative risk
21 assessment results and their implications. What conclusions will be associated with various results
22 or risk levels? For example, for a risk assessment team with members from the community,
23 industry, and the local and other government entities, what would happen if the assessment shows
24 risk levels to be "low"? Would members accept this? Conversely, if "unacceptable" risks are
25 determined, will all team members accept the results and their possible responsibility to do
26 something about that risk?
27
28 Discussions like these will help determine if the assessment can really address the
29 questions of the team. If not, the assessment may not be worth doing as planned. If members of
30 the team will not accept the possibility of a range of results of the analysis, then it is important to
31 reopen the entire planning and scoping discussion before anything is done in the analysis phase,
32 since the planning and scoping phase has not been satisfactorily completed.
33
34
35
36 3. THE ANALYSIS PHASE
37
38 The risk assessment paradigm most widely used during the past two decades was first
39 documented by the National Research Council (NRC, 1983). It consists of four parts: hazard
40 identification, dose-response assessment, exposure assessment, and risk characterization. This
H-35
-------
DRAFT- Risk Assessment Forum Review Draft - August 2, 2001 - Do Not Quote or Cite
1 paradigm was developed when almost all risk assessments were being done on single chemicals.
2 Nevertheless, it is a useful place to start when considering cumulative risks. There are a number of
3 ways to approach cumulative risk, either starting with the NRC paradigm or using a different
4 approach. Each may present its own challenges in methods, data, and analysis. Four examples of
5 general approaches are described below. There will undoubtedly be others developed as the
6 science advances.
7
8 3.1. General Approaches to Cumulative Risk Assessment
9
10 There are at least four different general approaches to cumulative risk assessment. These
11 are briefly outlined here. Each of the approaches has advantages and disadvantages, and will likely
12 come up with independently-derived estimates of cumulative risk (i.e., each uses different data
13 upon which to base the estimate). Given the different strengths and weaknesses, it may be useful
14 to use several of these approaches concurrently to strengthen the analysis.
15
16 The remainder of Chapter 3 summarizes some of the issues, and the current state of
17 knowledge, for various aspects of these approaches.
18
19
20 3.1.1. Combining Individual Stressor Risks.
21
22 For assessors familiar with the 1983 NRC risk paradigm (NRC, 1983), the most
23 conceptually straightforward approach would be to evaluate stressors individually, then combine
24 the individual risks. This can be done either by (a) combining toxicities before calculating risk, an
25 approach sometimes referred to as "combination toxicology" (Carpy, et al., 2000), or by (b)
26 calculating risks for individual stressors and then combining them.
27
28 Combination toxicology develops an estimate of combined toxicity for a multi-component
29 stressors such as a mixture, then treats the mixture, for risk estimation purposes, as if it were a
30 single entity lexicologically. Under this approach, chemical mixtures can be evaluated for toxicity
31 addition, independence, synergy, or antagonism, and a risk evaluation done on the mixture using
32 the 1983 NRC risk assessment paradigm. Mixtures of chemicals acting by the same mode of
33 action are sometimes shown to be additive, which will allow several stressors to be "lumped,"
34 simplifying the number of different types of stressors which need to be evaluated.
35
36 Calculating individual stressor risks and then combining them presents largely the same
37 challenges as combination toxicology, namely, taking interactions into account. Toxicity addition,
38 independence, synergy, or antagonism still need to be evaluated, but since risk estimates for
39 various stressors are often presented as values on the same numeric scale (e.g., as probabilities or
40 as fractions of an RfD or RfC), cancer risks are often just added together, as are non-cancer risks
H-36
-------
DRAFT- Risk Assessment Forum Review Draft - August 2, 2001 - Do Not Quote or Cite
1 as part of a Hazard Index. Addition of cancer risk estimates or hazard indices without
2 consideration for how these stressors may interact is making an (implicit or explicit) assumption
3 of dose additivity, which requires explanation in an assessment. This will be discussed further in
4 Chapter 4.
5
6 Given the current state-of-the-science, combining individual stressor risk, especially by
7 calculating individual risks and combining them, is probably the best known approach to
8 cumulative risk assessment, although there are currently many data gaps (especially in the area of
9 toxicity of mixtures) (USEPA, 200la). One major drawback is that as the number of stressors
10 increases, the difficulty of determining how the toxicities of all components interact increases
11 exponentially, and it becomes difficult to perform the assessment without many simplifying
12 assumptions. Depending on the tools or robustness of the data set available, the results of this
13 approach may be presented either in the form of probabilities of getting certain adverse effects
14 (e-g-, cancer), or, in the case of evaluating exposures relative to a reference level (such as used
15 with the hazard index approach), it may provide a gauge of the potential for effects5.
16
17 3.1.2. Use of Risk Factors Developed from Epidemiologic Associations.
18
19 The medical profession has long used "risk factors" to predict the chances of particular
20 health effects in individual patients. In this approach, the characteristics of individuals within the
21 population are correlated with the incidence of specific diseases or effects. For example, the risk
22 factors for stroke are: increasing age, heredity (family history) and race, prior stroke, high blood
23 pressure, cigarette smoking, diabetes mellitus, carotid and other artery disease, heart disease,
24 transient ischemic attacks (TIAs), high red blood cell count, sickle cell anemia, socioeconomic
25 factors, excessive alcohol consumption, and certain types of drug abuse (American Heart
26 Association, 2000). Each of these factors can be correlated with stroke incidence, and then the
27 risk of stroke from various combinations of these factors can be explored. Physicians use models
28 containing effect-specific risk factors to advise patients of the probabilities of future effects (e.g.,
29 stroke, breast cancer) based on their medical history. Although the medical data upon which these
30 factors are based have been well developed for many effects in humans, there are substantial data
31 gaps remaining in terms of the role played by exposures to many chemicals in the environment in
32 the development of human disease. This approach may be built on links between stressors and
33 effects for more well-studied stressors, but may be limited in both capability for quantification,
34 and coverage for stressors with less robust health effects data bases.
35
36 3.1.3. Biomarkers and Biomonitoring.
37
38 This approach uses biological measurements - biomarkers - to determine prior exposures
At exposures increasingly greater than reference levels, the potential for adverse health effects increases.
H-37
-------
DRAFT- Risk Assessment Forum Review Draft - August 2, 2001 - Do Not Quote or Cite
1 (biomarkers of exposure) or the current health status of individuals (biomarkers of effect). Use of
2 biomarkers for a group of chemicals or stressors which act upon individuals in the same way can
3 give the assessor a picture of where an individual currently falls on the continuum from exposure
4 to effects, making it much easier to predict risks if additional exposure occurs. A few biomarkers
5 (or even a single one) can represent exposure to a suite of chemicals. Although this reduces the
6 analytical burden and simplifies the process of estimating cumulative risk, the approach loses
7 some of the advantages of single-chemical assessment (especially being able to quickly discern the
8 importance of different pathways and routes of exposure contributing to the risk). This may be
9 the approach of choice in the future, but the state-of-the-science is not developed enough to make
10 this practicable today in an assessment with large numbers of diverse stressors (although it may be
11 possible to do this for more simple cases). One of the benefits of this method, the development of
12 data which show the actual current exposure and risk status of a population, is also its major
13 impediment: it can require extensive (or for humans, possibly invasive) monitoring. This can be
14 not only costly, but difficult to obtain. This approach uses primarily measurement methods, and
15 also can develop statements of probability of adverse effects of additional incremental exposures.
16 This approavch holds great promise for simplification of a cumulative risk assessment, but few
17 methods exist at this time for applying this approach in a cumulative assessment.
18
19 3.1.4. Other Types of Probability Statements.
20
21 Not all statements of probability of harm are expressed as probabilities of specific health
22 effects. Bernard Cohen, in his Catalog of Risks Extended and Updated (Cohen, 1991), uses
23 mortality ratios to derive "loss of life expectancy" (LLE) estimates for a wide variety of risk-
24 related activities. For example, workers in all occupations have a 60 day LLE as a result of
25 working, but workers in agriculture have a 320 day LLE, construction workers a 227 day LLE,
26 etc., as a result of their particular occupation. These types of statements are empirically derived,
27 probability-based statements of harm that do not use "probability of adverse health effect" as the
28 basis for the risk statement. For estimates such as LLEs, one could theoretically add up the
29 various activities and the corresponding LLEs in days to estimate a cumulative risk in terms of
30 loss of life expectancy. Other bases for risk statements include the quality-adjusted life year
31 (QALY), which has been used extensively in the medical field for cost-benefit analysis and also
32 has been proposed for use in comparative risk analysis (USEPA, 1998f). In a sense, this approach
33 is similar to the second approach, where risks are added, but it differs qualitatively in the types of
34 risk statements derived. These "other" types of probability statements could conceivably be used
35 in cumulative risk assessment.
36
37
38 3.2. Issues Related to the Approach of Combining Individual Stressor Risks
39
40 The approach of combining toxicities or individual stressor risks to come up with an
H-38
-------
DRAFT- Risk Assessment Forum Review Draft - August 2, 2001 - Do Not Quote or Cite
1 estimate of cumulative risk is most similar to traditional health risk assessment for chemicals, and
2 provides a quantitative analysis, yet may include a large degree of uncertainty.
3
4 In evaluating the combined effects from different chemicals, there is often an assumption
5 made that chemicals which have the same mechanism or mode of action, and result in the same
6 effects, are additive at any level of exposure or dose (e.g., see ILSI, 1999, page 23). The EPA's
7 Office of Pesticide Programs has prepared guidance on cumulative risk assessments for chemicals
8 with the same mode of action, as required under the Food Quality Protection Act (USEPA,
9 2000i). In this guidance, risks are only added for chemicals having the same mode of action. In
10 some screening-level assessments, risks from individual stressors may be added without
11 consideration of any similarity in mode of action (USEPA, 1998J). The issue of how individual
12 stressor risks contribute to the cumulative risk is critical to this approach, and will be discussed
13 further in Chapter 4.
14
15 Among the steps in this approach, the method usually requires (1) some evaluation of
16 what may be important to the cumulative risk of the population, through risk screening or other
17 means, (2) working through an analysis of the individual risks of individual stressors or mixtures,
18 and (3) determining or estimating the way these individual stressors act in combination with one
19 another. The following sections provide a discussion of some of the issues which may be
20 encountered within these steps.
21
22 3.2.1. Characterization of Hazard Identification and Dose-Response.
23
24 An initial step in the effects assessment component of human health risk assessment is
25 identification of the potential adverse health effects causally linked to the stressors of concern.
26 This is the hazard identification. Factors such as the route of exposure, the type and quality of the
27 effects, the biological plausibility of findings, the consistency of findings across studies, and the
28 potential for bioaccumulation all contribute to the strength of the hazard identification statement.
29
30 Dose-response assessment is the characterization of the relationship between the
31 concentration, exposure, or dose of a pollutant or pollutant group and the resultant health or
32 environmental effects. The nature of quantitative dose-response assessment varies among
33 pollutants. Sufficient data exist for a few pollutants, such as the air pollutants ozone or carbon
34 monoxide, so that relatively complete dose-response relationships can be characterized. In such
35 cases, there is no need for extrapolation to lower doses because adequate health effects data are
36 available for humans at environmental levels. Such is not the case, however, for most pollutants.
37 Most epidemiologic and toxicologic data on toxic pollutants typically result from exposure levels
38 that are high relative to environmental levels. Consequently, dose-response assessment methods
39 for most toxic pollutants generally consist of two parts. First is the evaluation of data in the
40 observable range, and second is the extrapolation from the observable range to low doses/risks.
H-39
-------
DRAFT- Risk Assessment Forum Review Draft - August 2, 2001 - Do Not Quote or Cite
1 Recent terminology refers to the result of analysis in the observable range as the "point of
2 departure," from which extrapolation begins. The approaches used for evaluation in the
3 observable range are similar for all types of effects, while the Agency's current extrapolation
4 methods differ considerably for cancer and non-cancer effects. Efforts are underway to
5 harmonize these two methods.
6
7 Important to characterizing hazard and the dose-response relationship is consideration of
8 the processes of distribution, elimination, and metabolism. Specific characteristics of different
9 chemical and biological stressors dictate how they are distributed within the body, how they are
10 eliminated and via what processes, and how they may be metabolized. These may differ with
11 route or circumstances of exposure, as well as characteristics of the exposed population (e.g., life
12 stage, genetic disposition, etc). To the extent that hazard and dose-response characterization is
13 drawn from laboratory animal data, differences or similarities between animals and humans in
14 distribution, elimination and metabolism are critical to the presumption of relevance to humans.
15
16 The Agency has clearly defined methods for hazard identification and dose-response
17 assessment for human health. Those described here are largely relevant to the majority of
18 pollutants for which human effect data at environmental exposures are scarce. In multi pollutant
19 risk assessments, however, it is important to consider the role of other pollutants for which
20 exposures eliciting human effects are not uncommon (e.g., ozone, particulate matter and carbon
21 monoxide in ambient air, nitrates or lead in drinking water). As the Agency's methods for cancer
22 and noncancer assessment currently differ, they are summarized separately here.
23
24 3.2.1.1. Current Methods for Assessing Noncancer Effects.
25
26 Due to the wide variety of endpoints, hazard identification procedures for noncancer
27 effects are less formally described in EPA guidance than are procedures for the identification of
28 carcinogens. The EPA has published guidelines for assessing several specific types of noncancer
29 effects, including mutagenicity (USEPA, 1986c), developmental toxicity (USEPA, 1991b),
30 neurotoxicity (USEPA, 1998e), and reproductive toxicity (USEPA, 1996b).
31
32 For identification of long-term (chronic) hazards other than cancer, EPA reviews the
33 health effects literature and characterizes its strengths and weaknesses, using a narrative approach
34 rather than a formal classification scheme. Available data on different endpoints are arrayed and
35 discussed, and the effects (and their attendant dose/exposure levels) are described. Particular
36 attention is given to effects that occur at relatively low doses or that may have particular
37 relevance to human populations. Information is presented in a narrative description that discusses
38 factors such as the methodological strengths and weaknesses of individual studies (as well as the
39 overall database), the length of time over which the studies were conducted, routes of exposure,
40 and possible biological mechanisms. EPA considers the severity of effects, which may range from
H-40
-------
DRAFT- Risk Assessment Forum Review Draft - August 2, 2001 - Do Not Quote or Cite
1 severe, frank, effects that can cause incapacitation or death, to subtle effects that may occur at the
2 cellular level but are early indicators of toxic effects. Not all effects observed in laboratory
3 studies are judged to be adverse. The distinction between adverse and non-adverse effects is not
4 always clear, and considerable professional judgment is required in applying criteria to identify
5 adverse effects. All of these observations are integrated into a presentation that gives a concise
6 profile of the toxicological properties of the pollutant.
7
8 The inhalation reference concentration (RfC) and oral reference dose (RfD), established by
9 Agency consensus after external peer review, are the primary quantitative toxicity values for use
10 in noncancer risk assessment. The RfC and RfD are defined as estimates, with uncertainty
11 spanning perhaps an order of magnitude, of an inhalation exposure or oral dose, respectively, to
12 the human population (including sensitive subgroups) that are likely to be without appreciable
13 risks of deleterious effects during a lifetime. The RfC or RfD is derived after a thorough review
14 of the health effects data base for an individual chemical and identification of the most sensitive
15 and relevant endpoint and the principal study(ies) demonstrating that endpoint. The methodology
16 for the RfD derivation is discussed in Barnes and Dourson (1988); inhalation RfCs are derived
17 according to the Agency's Methods for Derivation of Inhalation Reference Concentrations and
18 Application of Inhalation Dosimetry (USEPA, 1994). The RfC or RfD should represent a
19 synthesis of the entire data array. The evaluation of and choice of data on which to base the RfC
20 or RfD derivation are critical aspects of the assessment and require scientific judgment. The
21 Agency, under the auspices of a Technical Panel under the Risk Assessment Forum, is currently
22 evaluating the RfC and RfD methodology as to the need for revisions and improvements.
23
24 Derivation of the RfC or RfD begins with identification of the critical adverse effect from
25 the available valid human and animal study data, followed by identification of a lowest-observed-
26 adverse-effect level (LOAEL) or, preferably, a no-observed-adverse-effect level (NOAEL). The
27 LOAELs or NOAELs from animal studies are converted to human equivalent concentrations
28 (HECs) using dosimetric methods (described in USEPA, 1994). The NOAEL [HEC] or LOAEL
29 [HEC] from one or a few studies that is representative of the threshold region of observable
30 effects is the key value gleaned from evaluation of the dose-response data.
31
32 The RfC or RfD is then derived by consistent application of uncertainty factors (UFs),
33 generally a 1, 3 or 10, to account for recognized uncertainties in the extrapolation from the
34 experimental data and exposure conditions to an estimate (the RfC or RfD) appropriate to the
35 assumed human lifetime exposure scenario (Barnes and Dourson, 1988; USEPA, 1994). The
36 standard UFs are applied as appropriate for the following extrapolations or areas of uncertainty:
37 1) Laboratory animal data to humans; 2) Average healthy humans to sensitive humans; 3)
38 Subchronic to chronic exposure duration; 4) LOAEL to NOAEL; and 5) Incomplete data base.
39 The composite UF will depend on the number of extrapolations required. RfCs have been derived
40 using composite UFs that range from 10 to 3,000, with most RfCs using factors of 100 to 1,000.
H-41
-------
DRAFT- Risk Assessment Forum Review Draft - August 2, 2001 - Do Not Quote or Cite
1 The use of order-of-magnitude uncertainty factors for RfCs and RfDs and the definition of the
2 RfC or RfD as having "uncertainty spanning perhaps an order of magnitude" are indications of the
3 general lack of precision in the estimates.
4
5 In addition to toxicity related to chronic exposures, many hazardous air pollutants (HAPs)
6 also can cause toxic effects after acute (short-term) exposures lasting from minutes to several
7 hours. Indeed, for some pollutants acute exposures are of greater concern than chronic
8 exposures. The hazard identification step for acute effects is comparable to that for chronic
9 effects, with the primary difference being the duration of exposure. Methods for dose-response
10 assessment of acute exposures are substantially similar to the approach for chronic exposure.
11 Risk assessment for acute inhalation exposure is complicated by the steep concentration-response
12 curves that are often observed, and because small differences in exposure duration (in some cases,
13 a few minutes) need to be taken into account. Because increased exposure duration increases the
14 incidence and severity of response, acute toxicity criteria or exposure guideline values are
15 developed for a specified duration (e.g., one hour). While several EPA offices have addressed
16 acute exposures across a variety of regulatory programs, we have only recently drafted Agency -
17 wide guidance on how to assess toxic effects from short-term inhalation exposures. This
18 guidance for acute reference exposure (ARE) levels, when completed, will assist Agency acute
19 risk assessment activities for inhalation exposures (USEPA, 1998d).
20
21 3.2.1.2. Current Methods for Assessing Cancer Risks.
22
23 The EPA's 1986 Guidelines for Carcinogen Risk Assessment (USEPA, 1986d) provide
24 guidance on hazard identification for carcinogens. The approach recognizes three broad
25 categories of data: (1) human data (primarily epidemiological); (2) results of long-term
26 experimental animal bioassays; and (3) supporting data, including a variety of short-term tests for
27 genotoxicity and other relevant properties, pharmacokinetic and metabolic studies,
28 physical/chemical properties, and structure-activity relationships (SARs). In hazard identification
29 of carcinogens under the 1986 guidelines, the human data, animal data, and "other" evidence are
30 combined to characterize the weight of evidence regarding the agent's potential as a human
31 carcinogen into one of several hierarchic categories or groups: A) Carcinogenic to Humans; B)
32 Probably Carcinogenic to Humans; C) Possibly Carcinogenic to Humans; D) Not Classifiable as to
33 Human Carcinogen!city; and E) Evidence of Noncarcinogenicity for Humans.
34
35 In 1996, EPA proposed major revisions of the carcinogen hazard identification scheme.
36 The proposed revision to the cancer risk assessment guidelines (USEPA, 1996c), which has
37 undergone subsequent revisions as a result of Scientific Advisory Board reviews (e.g., USEPA,
38 1999e), focuses on narrative statements describing the main lines of evidence and their
39 interpretation, replacing the current alphabetic designations. The proposed guidelines also replace
40 the system of stepwise consideration of different types of data with a single comprehensive
H-42
-------
DRAFT- Risk Assessment Forum Review Draft - August 2, 2001 - Do Not Quote or Cite
1 evaluation process that stresses the coherence of various data elements. The result is a single
2 scientific interpretation that evaluates, to the extent possible, how well the commonality of mode
3 of carcinogenic action between human beings and the various test systems has been established.
4 Emphasis is also placed on defining the qualitative conditions under which carcinogenic hazards
5 might be expected. If warranted, limitations to the finding of carcinogenic hazard can be drawn
6 based on route of exposure, existence of other factors needed for tumorigenesis, and doses below
7 which elevation of cancer risk is not expected.
8
9 EPA's 1986 Guidelines for Carcinogen Risk Assessment adopted a default assumption
10 that chemical carcinogens would exhibit risks at any dose (USEPA, 1986d). This is often called
11 the "no threshold" assumption, that is, unlike non-carcinogens, there is no concentration below
12 which there is no risk. Extrapolation of cancer risk using the linearized multistage model, which
13 results in a linear extrapolation of risk in the low dose region, was proposed as a reasonable
14 upper-bound on risk, and this approach has been used for most chemicals with adequate data
15 since then. The 1986 guidelines did allow that when data supported it, other models could be
16 used in addition to the default linearized model. The Proposed Guidelines for Carcinogen Risk
17 Assessment (USEPA, 1996c), however, stressed that when there are adequate mechanistic data to
18 suggest that other models would be more appropriate to estimate low exposure risk, they may be
19 used on a case-by-case basis in lieu of the linearized multistage model. In the absence of such
20 data, the assumption of response linearity is maintained, although the modeling scheme has been
21 simplified.
22
23 In cancer dose-response assessment, the evaluation of data in the observable range is
24 similar to that for noncancer effects. The method of extrapolation to lower doses from the point
25 of departure, however, differs depending on whether the assessment of the available data on the
26 mode of action of the chemical indicates a linear or nonlinear mode of action. For linear
27 extrapolation, a straight line is drawn from the point of departure to the origin (i.e., to the point
28 where dose and response are both zero on the dose-response curve), and the risk at any
29 concentration is determined by interpolation along that line. A linear mode of action serves as a
30 default when available evidence is not sufficient to support a nonlinear extrapolation procedure,
31 even if there is no evidence for DNA reactivity.
32
33 Nonlinear methods (where data support them) or a margin of exposure approach are
34 recommended when there is sufficient evidence to support a nonlinear mode of action. A
35 nonlinear mode of action could involve a dose-response pattern in which the response falls much
36 more quickly than linearly with dose, but still indicating risk at low doses. Alternatively, the mode
37 of action may theoretically have a threshold if, for example, the cancer response is a secondary
38 effect of toxicity or an induced physiological change which is a threshold phenomenon.
39
40 3.2.1.3. Time-Related Issues with Dose-Response Curves for Cumulative Assessments.
H-43
-------
DRAFT- Risk Assessment Forum Review Draft - August 2, 2001 - Do Not Quote or Cite
1
2
3
4
5
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
Cumulative risk encompasses repeated
exposures to a single stressor or exposures to
multiple stressors (see definition, section 1.3).
This has implications with regard to the dose-
response assessment method used. Most
exposure data used in developing a dose-
response relationship (with the exception of
some life-stage related effects such as
developmental toxicity) is usually treated as
"cumulative" for the duration of interest, but
may not match the exposure regimes seen in
actual assessments. Moreover, in the case of
non-cancer effects for many chemicals, there is
no explicit description of the dose-response
relationship for use in the risk assessment,
since the objective has usually been
development of an RfC or RfD (a level at
which effects are considered low probability)
to be compared with estimates of continuous
exposure or daily doses.
In the case of linear carcinogens, this
cumulative exposure assumption has been
carried into the risk assessment step.
Regardless of the details of the exposure
circumstances for the study on which the
cancer potency was based, it is assumed that
there is a linear relationship between amounts
of exposure and associated cancer risk For
non-linear carcinogens assessed in cumulative risk assessments6, the details and sequence of
exposure may be important, both in developing the dose-response relationship and in predicting
risk associated with exposures of interest.
As some chemicals may have the ability to affect an organism's response to other
chemicals, consideration of the time sequence of exposure may take on an additional layer of
Some Issues Concerning Time Sequence of
Exposures in Developing Dose-Response
Relationships for Cumulative Risk Assessment
• What types of chemicals are likely to function as
"promoters" or to cause damage that will make a
person more susceptible to other exposures later?
What is known about how they work?
• What are some of the ways time sequencing is
dealt with in considering risk from effects which are
thought to have a threshold? Can these methods be
adapted to cumulative risk assessment?
• What work has been done in looking at non-
chemical stressors which can cause a person to be
more susceptible to exposure to chemicals later? Is
anything known about the permanence or transitory
nature of the damage done by these non-chemical
stressors?
• What are the specific factors which need to be
known to properly evaluate risks from exposures to
different stressors at different times? What
circumstances, types of stressors, or non-chemical
stressors may be important?
• What new types of problems will cumulative risk
assessment present to the practitioner confronted
with a population exposed to a non-constant mixture
of stressors over a period of time?
The draft cancer guidelines (USEPA, 1996c) explicitly recognize the potential for non-linear dose response. It is
only in the case where non-linear response is modeled that time sequence of exposure can be considered in the risk assessment.
H-44
-------
DRAFT- Risk Assessment Forum Review Draft - August 2, 2001 - Do Not Quote or Cite
1 complexity in multiple chemical cumulative risk assessments.
2
3 3.2.1.4. Issues Associated with Assessing Mixtures of Stressors.
4
5 While some potential environmental hazards involve significant exposure to only a single
6 compound, most instances of environmental contamination involve concurrent or sequential
7 exposures to a mixture of compounds. These various components may induce similar or
8 dissimilar effects over exposure periods ranging from short-term to lifetime. Within EPA's
9 guidelines on assessing health risks from chemical mixtures, mixtures are defined as any
10 combination of two or more chemical substances regardless of source or of spatial or temporal
11 proximity that can influence the risk of chemical toxicity in the target population (USEPA,
12 1986b). In some instances, the mixtures are highly complex, consisting of scores of compounds
13 that are generated simultaneously as by-products from a single source or process (e.g., coke oven
14 emissions and diesel exhaust). In other cases, complex mixtures of related compounds are
15 produced as commercial products (e.g., PCBs, gasoline and pesticide formulations) and eventually
16 released into the environment. Another category of mixtures consists of compounds, often
17 unrelated chemically or commercially, that are placed in the same area for disposal or storage, and
18 have the potential for combined exposure to humans.
19
20 Multi-pollutant exposure scenarios can be extremely diverse. Moreover, the quality and
21 quantity of pertinent information available for risk assessment varies considerably for different
22 mixtures. Occasionally, the chemical composition of a mixture is well characterized, levels of
23 exposure to the population are known, and detailed toxicologic data on the mixture are available.
24 Most frequently, some components of the mixture are unknown, exposure data are uncertain or
25 vary over time, and toxicologic data on the known components of the mixture are limited
26
27 To address concerns over health risks from multi-chemical exposures, EPA issued
28 Guidelines for Health Risk from Exposure to Chemical Mixtures in 1986 (USEPA, 1986b).
29 Those Guidelines described broad concepts related to mixtures exposure and toxicity and included
30 few specific procedures. In 1989, EPA published guidance for the Superfund program on
31 hazardous waste that gave practical steps for conducting a mixtures risk assessment (USEPA,
32 1989a). Also in 1989, EPA published the revised document on the use of Toxicity Equivalence
33 Factors for characterizing health risks of the class of chemicals including the dibenzo-dioxins and
34 dibenzofurans (USEPA, 1989b). In 1990, EPA published a Technical Support Document to
35 provide more detailed information on toxicity of whole mixtures and on toxicologic interactions
36 (e-g-, synergism) between chemicals in a binary (two-chemical) mixture (USEPA, 1990a). The
37 concept of toxicologic similarity was also discussed, and is expanded upon in the recent Guidance
3 8 for Conducting Health Risk Assessment of Chemical Mixtures (USEPA, 2001 a).
39
40 The prediction of how specific mixtures of toxicants will interact must be based on an
H-45
-------
DRAFT- Risk Assessment Forum Review Draft - August 2, 2001 - Do Not Quote or Cite
1 understanding of the mechanisms of such interactions. It generally is recognized that toxicant
2 interactions may occur during any of the processes that take place with a single compound:
3 absorption, distribution, metabolism, excretion, and activity at the receptor site(s). Compounds
4 may interact chemically, yielding a new toxic component capable of causing a change in the
5 biological availability of the existing component. They may also interact by causing different
6 effects at different receptors sites. Because of the uncertainties inherent in predicting the
7 magnitude and nature of toxicant interactions, the assessment of health risk from chemical
8 mixtures must include a thorough discussion of all assumptions. No single approach is
9 recommended in the Agency Guidelines. Instead, guidance is given for the use of several
10 approaches depending on the nature and quality of the data. Accordingly, the most recent
11 Guidance describes procedures for assessment using data on the mixture of concern, data on a
12 lexicologically similar mixture, as well as data on the mixture component chemicals. The state of
13 science varies dramatically for these three approaches. The "whole mixture" procedures are most
14 advanced for assessing carcinogenic risk, mainly because of the long use of in vitro mutagenicity
15 tests to indicate carcinogenic potency. In vitro test procedures for noncancer endpoints are still in
16 the pioneering stage. In contrast, the component-based procedures, particularly those that
17 incorporate information on toxicologic interactions, are most advanced for noncarcinogenic
18 toxicity.
19
20 Risk assessment on mixtures usually involves substantial uncertainty. If the mixture is
21 treated as a single complex substance, these uncertainties range from inexact descriptions of
22 exposure to inadequate toxicity information. When viewed as a simple collection of a few
23 component chemicals, the uncertainties include the generally poor understanding of the magnitude
24 and nature of toxicologic interactions, especially those interactions involving three or more
25 chemicals. Because of these uncertainties, the assessment of health risk from chemical mixtures
26 must include a thorough discussion of all assumptions and the identification when possible of the
27 major sources of uncertainty.
28
29 3.2.1.5. Hazards Other than Chemical Hazards.
30
31 In addition to chemical stressors, there are other broad categories of stressors: biological,
32 radiological, physical, and various other types of hazards can also cause adverse effects. The
33 adverse effects of radiation, such as the radiation from radon gas which infiltrates into a home, are
34 well-known. Biological effects, such as bacterial infections and Cryptosporidium outbreaks in
35 drinking water, can have very serious adverse effects. Physical hazards include natural hazards,
36 such as earthquakes, hurricanes, and floods, or man-made hazards, such as traffic accidents. Other
37 types of stressors, including socioeconomic factors and lifestyle conditions, can also cause or
38 exacerbate harmful effects.
39
40 The context of a risk assessment might lead a risk assessor to consider the adverse effects
H-46
-------
DRAFT- Risk Assessment Forum Review Draft - August 2, 2001 - Do Not Quote or Cite
1 from exposure to a number of chemical, biological, physical, or other stressors which present
2 different types of hazards. Chemically, when two stressors cause similar effects (for example,
3 both are cholinesterase inhibitors), the interaction could lead to additive, synergistic, antagonistic,
4 or potentiated effects. Stressors causing different effects may interact in ways to potentiate either
5 or both the effects of the individual stressors, or dampen one or both the effects, or even operate
6 independently of one another (USEPA 200la). These possibilities also exist for the interactions of
7 chemical stressors with non-chemical stressors. Cumulative risk assessment could encompass the
8 interactions of chemical stressors with biological stressors, physical stressors, ecological stressors,
9 radiological stressors, and other stressors such as socioeconomic or lifestyle conditions (e.g., diet,
10 smoking, health care, housing).
11
12 One of the important processes in the development of cumulative risk assessment will be
13 the development of methodologies that can be used to compare and combine the risks from very
14 different types of hazards. The risk assessment methodology for single chemical exposure is well
15 developed, but methods for assessing risks from multiple chemicals, or for combining risks from
16 different types of hazards, such as biological and chemical, or biological and physical, are nowhere
17 near as robust and available.
18
19 Although the ultimate aim of cumulative risk assessment might include a combined risk
20 across many different types of hazards for a population, realistically, it will take a great deal of
21 research to develop methods to adequately combine risks across different types of hazards. This
22 will be discussed more in Chapter 4.
23
24 3.2.1.6. Vulnerability.
25
26 One of the concepts that can be used in risk assessments (both for human health and
27 ecological assessments) is that of vulnerability of the population or ecosystem. Vulnerability has
28 been a common topic in socioeconomic and environmental studies. The European Commission's
29 TEMRAP (The European Multi-Hazard Risk Assessment Project), studying vulnerability to
30 natural disasters such as floods, windstorms, fires, earthquakes, and others, defines "vulnerability"
31 as "the intrinsic predisposition of an exposed element to be at risk of suffering losses (life, health,
32 cultural or economic) upon the occurrence of an event of [a specific] intensity" (European
33 Commission, 2000).
34
35 Vulnerability of a population places them at increased risk of adverse effect. The
36 Agency's risk characterization policy and guidance (USEPA, 2000c) touches on this concept by
37 recommending that risk assessments "address or provide descriptions of [risk to] ... important
38 subgroups of the population, such as highly exposed or highly susceptible groups". Further, the
39 Agency's guidance on planning and scoping for cumulative risk assessments (USEPA, 1995b)
40 recognizes the importance of "defining the characteristics of the population at risk, which include
41 individuals or sensitive subgroups which may be highly susceptible to risks from stressors or
H-47
-------
DRAFT- Risk Assessment Forum Review Draft - August 2, 2001 - Do Not Quote or Cite
1 groups of stressors due to their age, gender, disease history, size or developmental stage". That
2 guidance also recognizes the potential importance of other social, economic, behavioral or
3 psychological stressors that may contribute to adverse health effects (e.g., existing health
4 condition, anxiety, nutritional status, crime and congestion). These same concepts may also be
5 discussed as a group in terms of "population vulnerability." The various ways in which a
6 population may be vulnerable are discussed below in four categories.
7
8 The first of these is susceptibility or sensitivity. Susceptible or sensitive individuals within
9 a population have a different or more pronounced dose-response relationship when confronted
10 with a stressor. Reasons for susceptibility may be related to any number of factors, including life
11 stage (e.g., children or the elderly may be more susceptible), prior exposure (e.g., developing
12 sensitization reactions, or having had exposures which compromise the immune system), genetic
13 polymorphisms (e.g., genetic susceptibilities which occur in a small but significant percentage of
14 the population), or existing disease state (e.g., asthmatics). Confronted with equal concentrations
15 of a chemical for equal durations, for example, a biologically susceptible individual may show
16 effects while the typical individual within the population would not. Although we generally do not
17 have a lot of data available on this topic, susceptibilities or sensitivities may also exist among
18 races or genders.
19
20 The second category of vulnerability is differential exposure. While it is obvious by
21 examining a dose-response curve that two individuals at different exposure levels may have
22 different likelihood of effects, this also extends to differences in historical exposure, body burden,
23 and background exposure, which are sometimes overlooked in an assessment.
24
25 The third category of vulnerability is differential preparedness to withstand the insult of
26 the stressor, and the fourth is the differential ability to recover from the effects of the stressor.
27 These last two are linked to what kind of coping systems and resources an individual, population,
28 or community has. Preparedness or recovery is often a crucial factor in ecological assessments. In
29 human health assessments, lack of access to health care, income differences, unemployment, or
30 lack of insurance, for example, may affect a community's ability to prepare or recover from a
31 stressor. One aspect of differential ability to recover is illustrated by differing survival rates for
32 the same disease (e.g., Lantz, et. al 1998).
33
34 Cumulative risk assessments may be uniquely suited to addressing the issues related to
35 vulnerability. In order to do that, however, there needs to be some relationship between the
36 factors discussed above and changes in risk. At the current state of the science, these factors have
37 not been extensively developed beyond correlations between mortality rates and several
38 socioeconomic factors such as income (e.g., Lynch, et al. 1998).
39
40 3.2.2. Characterization of Exposure.
41
H-48
-------
DRAFT- Risk Assessment Forum Review Draft - August 2, 2001 - Do Not Quote or Cite
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
Exposure generally refers to contact of an individual or the study population with the
stressor of interest. With regard to human exposure, the Agency defines exposure as taking place
at the visible external boundary of the person (e.g., skin, and openings into the body such as
mouth and nostrils) (USEPA, 1992c). Following exposure, a chemical or biological stressor may
be taken up into the body (e.g., inhaled, or ingested) leading to its availability for absorption into
the circulatory system, distribution to various sites within the body, elimination from the body and
metabolism or transformation. These processes following contact (exposure) are considered in
the hazard and dose-response characterization (see section 3.1).
The general approaches to quantitative exposure assessment are discussed in EPA's
Guidelines for Exposure Assessment (USEPA 1992c), which suggests three:
• Direct measurement. Measurement of exposure at the point of contact while the exposure is
taking place, measuring both the exposure concentration and the time of contact and integrating
them;
• Scenario evaluation. Estimation of exposure
by separately estimating the exposure
concentration and the time of contact, then
combining this information through modeling;
and
• Dose reconstruction. Estimating the
exposure from reconstructing the dose through
internal indicators such as biomarkers, body
burden, or excretion levels.
These same three approaches are useful
for evaluating exposure in a cumulative risk
assessment. The first approach, direct
measurement of exposure, requires personal
exposure measurements for individuals within a
population. The second approach, scenario
evaluation (most often employed by the
Agency), is usually done using environmental
source evaluations, fate and transport models,
population demographics, exposure models,
and by constructing exposure scenarios. (This
approach often is used by constructing a
conceptual model which uses monitoring data for
reconstruction, employs markers of exposure and
Components of Exposure Assessment
• Characterization of the Source in terms
of the pollutants/stressors released into the
environment, release rates, or amounts and
characteristics of the release.
• Environmental Fate and Transport
Characterization including how the
pollutant/stressor is transported, dispersed
and transformed over the area and media of
interest.
• Characterization of the Study
Population in terms of geographic
distribution and other characteristics relevant
to the exposure pathways or pollutant
effects of concern.
• Exposure Characterization is the spatial
integration of the pollutant
concentration/stressor intensity with the
study population.
calibration.) The third approach, dose
dose. The three different approaches use
H-49
-------
DRAFT- Risk Assessment Forum Review Draft - August 2, 2001 - Do Not Quote or Cite
1 different data for input to the exposure estimate, and for that reason, can be complementary for
2 verifying or validating estimates by either of the other approaches.
3
4 For most aspects of exposure assessment, the 1992 Guidelines for Exposure Assessment
5 provide a detailed discussion which can be used as the basis for exposure assessments within
6 cumulative risk assessments. Agency documents providing more in-depth discussion of
7 assessment methods for particular exposure routes or pathways (USEPA, 1999g) are also
8 available. There are several aspects of cumulative risk assessments which were not addressed by
9 the Guidelines. One of these, the concept of aggregate exposure (generally meaning the sum of
10 exposures for a stressor from multiple sources and routes over time), has been considered in the
11 development of some drinking water and air quality standards, and became a major focus of the
12 Food Quality Protection Act of 19961.
13
14 Although the concept of aggregate exposure focuses on a single chemical or stressor, it
15 does so from the standpoint of a defined receptor or population, and theoretically includes all
16 relevant pathways by which a chemical can reach the population. In Figure 2, a single circle
17 marked "chemical" or "stressor," with the arrow that connects it to the population, represents
18 aggregate exposure.
19
20 Across the various EPA programmatic areas, Offices are currently assessing or are moving
21 toward assessment of aggregate exposures, effects of mixtures, and cumulative risks. Several
22 other novel aspects of exposure assessment within the framework of a cumulative risk assessment
23 are discussed in the sections below.
24
25 3.2.2.1. The Time Dimension of Exposure.
26
27 As discussed in section 3.2.1.3, risk assessment for carcinogens has historically used a
28 linear, non-threshold theory8 which attaches the same risk of effect to a unit of exposure
The Food Quality Protection Act of 1996 used the term "aggregate exposure" eight times. Although it did not
define the term, it was used in the context of multiple exposures to a single pesticide chemical residue from a variety of
pathways. Typical of the wording is that on page 110 STAT. 1518 of the Act, which directs the Administrator to consider,
among other relevant factors, "available information concerning the aggregate exposure levels of consumers (and major
identifiable subgroups of consumers) to the pesticide chemical residue and to other related substances, including dietary
exposure under the tolerance and all other tolerances in effect for the pesticide chemical residue, and exposure from other non-
occupational sources."
Q
The linear, non-threshold theory is a convention applied when time data are not available. Although time-to-tumor
models and the Armitage-Doll model structure have been considered since the 1960s, their evaluation requires data that are not
routinely obtained in experiments.
H-50
-------
DRAFT- Risk Assessment Forum Review Draft - August 2, 2001 - Do Not Quote or Cite
1 regardless of when the exposure occurs during a lifetime9. For this reason, average exposures and
2 doses have been used extensively in risk assessments for cancer. For non-cancer effects, however,
3 a threshold is usually assumed, and effects may be contingent upon exposure at certain "critical
4 periods" in a person's lifetime (e.g., certain critical periods during pregnancy for developmental
5 toxicity, or exposure while a child during development). The initiator-promoter model of
6 carcinogenesis, first described with mouse skin studies almost 60 years ago10, provides an example
7 of the role that can be played by multiple chemicals and time sequence of exposure in eliciting a
8 physiological effect or disease such as cancer. This has implications for risk assessment. For
9 example, persons with relevant past exposures might have increased susceptibility to the effects of
10 a particular chemical due to a previous exposure to the same or a second chemical.
11
12 These considerations suggest that for cumulative risk assessment, chemical exposures
13 need to be characterized in terms of which other chemicals are present, and when. As noted in the
14 ILSI Framework for Cumulative Risk Assessment: "Data collected specifically to support a
15 cumulative exposure assessment should conserve the covariance and dependency structures
16 associated with the chemicals of concern.... For example, when residue levels of each chemical
17 within a set of chemicals are measured concurrently in an environmental matrix, such as food, the
18 probability of one chemical being absent when another is present is implicit in the analytical
19 results.... The combination of independent data sets to produce a cumulative assessment will
20 require inclusion of estimated covariance factors, and will necessarily reduce the reliability of the
21 analyses." (ILSI, 1999)
C)
A typical unit of dose used in exposure assessments for carcinogens is the Lifetime Average Daily Dose (LADD)
(USEPA, 1992c). By averaging dose over a lifetime, one is assuming that it doesn't make any difference to the ultimate
toxicity when the exposure takes place, or what the exposure pattern is. This assumption, a derivative of "Haber's
Rule"(Haber, 1924; a relationship developed in a study of mustard gas effects which showed that the effect - over a limited
range - was proportional to the product of concentration of the gas times exposure duration), fits well with the linear, non-
threshold approach to carcinogen risk assessment. Although the LADD is currently widely used in risk assessment, it will not
be accurate if the dose-response curve is not "linear non-threshold."
The model system for mouse skin carcinogenesis (Rouse and Kidd, 1941; Mottram, 1944; Berenblum and Shubik,
1947) involved the alteration (or "initiation") of individual cells as a result of a single dose of a chemical carcinogen.
Subsequently, the application of a second agent, which itself was not considered carcinogenic, elicited skin papilloma
(indicating "progression").
H-51
-------
DRAFT- Risk Assessment Forum Review Draft - August 2, 2001 - Do Not Quote or Cite
1
2
3
4
5
6
7
Some Examples of Exposure Models which Consider Time Aspects
Calendex (Novigen Sciences, Inc), integrates different pathways e.g., dietary (food and water), and
residential, and routes (oral, dermal, inhalation) of exposure using a calendar-based probabilistic approach .
One of the important factors of this approach is it provides estimates of risk which reflect aggregate and
cumulative exposure to discrete individuals with exposure pathways and routes appropriately linked for the
scenarios being assessed. Calendex also allows one to estimate exposure pre- and post use of a chemical, as
well as degradation periods. Calendar based assessments maintain the integrity of the individual by
capturing: the location of the exposed individual, the time of year in which he or she was exposed, and the
patterns of exposure. Calendex also allows for a variety of time-breakout options for analysis of exposure.
For example, specific, single day exposures which are multipathway (e.g., one could perform an assessment
on June 21 if one knew of specific exposure timing with which we were concerned).
APEX - The Air Pollution Exposure (APEX) model is based on the pNEM probabilistic National Ambient
Air Quality Standards model (pNEM) for carbon monoxide (Johnson, et al., 2000). This model mimics the
basic abilities of the pNEM/CO model; it calculates the distributions of human exposure to selected airborne
pollutants within a selected study area as a function of time. As a dose model (for CO), it calculates the
pollutant dose within the body, specifically summarized by the blood carboxyhemoglobin (COHb)
concentration. APEX is a cohort-microenvironment exposure model in that it combines daily activity diaries
to form a composite year-long activity pattern, which represent specific population cohorts and are tracked as
they move from one microenvironment to another. A cohort consists of a subset of the population that is
expected to have somewhat similar activity (and hence exposure) patterns; they are formed by combining
demographic groups and geographic locations (districts). Once each cohort has been modeled and its relative
size determined, an exposure distribution for the entire population can be assembled. A microenvironment
is a description of the immediate surroundings of an individual that serves as an indicator of exposure (e.g.,
inside a residence, school or car, outdoors, etc.). APEX has been developed as one of the inhalation exposure
models accessible in the Exposure Event Module of the Total Risk Integrated Methodology (TRIM.Expo) for
assessment of exposures to either criteria or hazardous air pollutants (USEPA, 1999J)
Other models include the LifeLine Model, developed under a cooperative agreement between EPA/OPP and
Hampshire Research Institute (Hampshire Research Institute, 1999, 2000); the Stochastic Human Exposure
and Dose Simulation Model (SHEDS), under development by EPA's Office of Research and Development
(Zartarian, et al., 2000), and the Cumulative and Aggregate Risk Evaluation System (CARES), under
development by member companies of the American Crop Protection Association (APCA, 1999).
Cumulative risk assessment presents challenges in matching exposure estimates with dose-
response relationships. In a cumulative risk assessment, the time sequence of exposure may be
particularly important. As discussed in section 3.1.3, ideally, the dose-response assessment will
indicate the importance of the time sequence for the chemical(s) of interest in the assessment. In
cumulative assessments involving these chemicals, in the same way, it may become important to
characterize the details and sequence of exposure to the exposed population (see box above), so
there will be a match in not only the form, but also the assumptions between the dose-response
relationship and the exposure or dose estimate.
H-52
-------
DRAFT- Risk Assessment Forum Review Draft - August 2, 2001 - Do Not Quote or Cite
1 3.2.2.2. Variation of Mixtures.
2
3 Unlike exposure assessments for single chemicals, the cumulative assessment is likely to
4 have to evaluate the exposure to mixtures, whether mixtures of chemicals or mixtures of
5 chemicals and other stressors. Evaluating exposure to mixtures requires characterizing the mixture
6 at the point of contact, but often the data are for the composition of the mixture at the source. In
7 these cases, the assessment may need to include a fate and transport modeling component to
8 predict how that combination of chemicals may have changed both over time and space, as the
9 chemicals move to the point where (human or ecological) receptors are exposed. The chemical
10 mixture at the point of contact with a receptor might be quite different from the original mixture
11 generated at the source, since chemicals move differently through the environment and can have
12 different rates of degradation in the environment (i.e., mixtures such as PCBs are sometimes said
13 to "weather" over time).
14
15 EPA has developed information to assist in determining what chemical mixtures people are
16 likely to be exposed to under different situations. While monitoring at the exposure point can be
17 the most accurate way to determine the specific nature of mixtures to which receptors are
18 exposed, there are also numerous modeling tools which can be used to predict the transport,
19 dispersion and transformation of chemicals in the environment. A variety of fate and transport
20 models as well as the parameters needed to run the models (such as vapor pressure, partitioning
21 coefficients, solubility measures, etc.) have been developed and activities continue to improve
22 them. These models can assist the assessor in predicting the nature of the mixture at the point of
23 contact.
24
25 The fact that mixtures can change or degrade over time and space makes exposure
26 assessment within the cumulative risk assessment a particular challenge when exposure
27 measurements are not available. Both exposure measurements at the receptor and predictive
28 approaches are applicable, and each pose its own challenges in implementation, including resource
29 requirements and uncertainty.
30
31 3.2.2.3. Sources and Pathways of Exposure.
32
33 Pathways of exposure within a cumulative risk assessment can be many and varied, as can
34 the sources of chemicals or stressors into the environment. Consider, as sources, for example,
35 consumer products or pesticides used, improper disposal of hazardous waste or hazardous
36 material, discharge of wastewater into surface or ground water, motor vehicle emissions, or
37 emissions from large (factories, power plants) or small (gas stations, dry cleaners, home heating)
38 point sources.
39
40 These chemicals can reach the receptor by a variety of pathways. For example,
41 application of an agricultural pesticide can potentially contribute to a farm-worker population's
H-53
-------
DRAFT- Risk Assessment Forum Review Draft - August 2, 2001 - Do Not Quote or Cite
1 exposure to that pesticide via inhalation of "drift" during and immediately after application,
2 ingestion of food or water to which the pesticide has been transported or directly applied,
3 ingestion of livestock who have been fed produce to which the pesticide has been applied, and
4 dermal absorption from contact with vegetation or clothing after application. An urban
5 population may be exposed to a volatile organic compound such as benzene from inhalation of
6 outdoor air receiving emissions from mobile sources and various large and small stationary
7 sources (e.g., petroleum refineries, bus stations, truck stops and gas stations), inhalation of air
8 while driving or riding in a car or bus, and inhalation of air inside the home, office or other
9 establishment frequented by a tobacco smoker. A population's exposure results from the
10 aggregate of all of the relevant pathways. The Agency considers the former example as part of
11 pesticide registration under the FQPA, and considers many of the latter in exposure assessments
12 conducted in support of the National Ambient Air Quality Standards and, more recently, in
13 priority setting for the air toxics program.
14
15 Sources and pathways for non-chemical stressors such as biological, radiological, and
16 other stressors can be even more varied. Many of these sources and pathways, with which
17 chemical risk assessors may be unfamiliar, were not routinely evaluated within the scope of
18 historical single-chemical risk assessments, but they may be of interest in some cumulative risk
19 assessments. Some of these sources are discussed in section 3.2.5. In many of the items below,
20 they are routinely evaluated in certain types of assessments, but not typically addressed in others.
21 For cumulative assessments, it is useful to have a list of sources of information. One such list can
22 be found in Appendix A.
23
24
25 3.2.2.4. Subpopulations with Special Exposures.
26
27 Certain subpopulations can be highly exposed to stressors based on geographic proximity
28 to sources of these stressors, coincident direct or indirect occupational exposures, their activity
29 patterns, or a combination of these factors. A cumulative risk assessment may need to include
30 special emphasis on identifying and evaluating these subpopulations.
31
32 Subpopulations at risk of high exposure due to geographic proximity could include
33 workers at a facility which is a source of a stressor or residents near such sources. Specific
34 examples might be people living in the plume from a coal burning power plant, those near and
35 using a polluted water body (for example, for fishing or recreation), or along roadways with high
36 levels of vehicular traffic.
37
38 Occupational exposures may be either direct (occurring in the workplace) or indirect
39 (occurring at home). Indirect occupational exposures include those experienced by family
40 members of those occupationally exposed, who may be exposed to occupational chemicals
41 brought into the house by the worker (e.g., on clothes, breath, etc). Thus, workers or family
H-54
-------
DRAFT- Risk Assessment Forum Review Draft - August 2, 2001 - Do Not Quote or Cite
1 members may be subject to greater exposures than others in the population without this additional
2 burden.
3
4 Examples of subpopulations at high exposure due to activity patterns may include people
5 who exercise heavily in polluted air, recreational or subsistence fishers or hunters who consume
6 large quantities offish or wild animals, farmers or others who get a large percentage of their food
7 from a location near a source of pollution and live in areas with high pesticide use, individuals
8 with long commutes in automobiles, or children (because they consume a larger amount of food,
9 drink, and air relative to their body weight, and because of additional exposure routes such as
10 incidental soil ingestion).
11
12 Two examples of the combined impact of high geographic exposure and high exposure
13 activity patterns are runners who run along heavily traveled roadways, and those who fish for
14 food in heavily polluted urban rivers.
15
16 It is important to recognize that some heavily exposed populations may also be
17 particularly vulnerable. Examples of those who could be particularly vulnerable to certain
18 stressors include children during certain stages of development, people with chronic respiratory
19 problems, the elderly, and those economically disadvantaged without access to medical care. A
20 cumulative risk assessment may need to take into account potential combinations of high
21 exposure and high vulnerability.
22
23 3.2.2.5. Exposures to Non-chemical Stressors.
24
25 Depending on the scope of the cumulative risk assessment, the analysis may include non-
26 chemical stressors which could cause adverse effects, or interact with chemical stressors to
27 potentiate or otherwise change the dose-response relationship of a chemical in a specific
28 population.
29
30 Assessing exposure to non-chemical stressors may be straightforward, such as in the case
31 of radon exposure. Radioactivity can be sampled, measured, and exposures estimated. Estimating
32 other exposures, such as stress induced by living near hazardous waste sites, or stress due to
33 impact on so-called "quality-of-life criteria," may not be straightforward at all. Partly, this is
34 because their evaluation moves away from a strictly analytical, scientific, process to a more
35 analytic-deliberative process. Exposure to psychological stressors and stressors that can affect
36 quality-of-life criteria are discussed in more detail in section 4.1.2. Appendix A suggests some
37 further reading on methods relevant to determining exposures to other non-chemical stressors.
38
39 3.3. Issues Related to the Approach of Using Risk Factors
40
41 [To be added]
H-55
-------
DRAFT- Risk Assessment Forum Review Draft - August 2, 2001 - Do Not Quote or Cite
1
2 3.4. Issues Related to the Approach of Biomarkers and Biomonitoring
3
4 [To be added]
5
6 3.5. Issues Related to Other Approaches
7
8 [To be added]
9
10
11 4. THE INTERPRETATION (RISK CHARACTERIZATION) PHASE
12
13 4.1 Risk Estimation
14
15 Risk estimation in a cumulative risk assessment will involve some combination of risks,
16 either risks from various stressors11 causing similar effects, or risks from various stressors causing
17 different types of effects. The stressors may be similar or widely different. Combinations of many
18 types of stressors with different endpoints in a single assessment will quickly cause the risk
19 estimation step to become very complex and difficult. Basic calculation techniques for various
20 single-chemical risks are covered in EPA's various Guideline documents (USEPA 1986c, 1986d,
21 1991b, 1992c, 1994, 1996b, 1996c, 1998b, 1998e). The following sections discuss how risk
22 estimation in cumulative risk assessments may differ substantially from single-stressor
23 assessments.
24
25 4.1.1. Methods for Combining Chemical Risks
26
27 One approach to assessing health risk from multiple stressors is to combine the individual
28 risks when the effects are similar (this was example #2 in section 1.3). The simplest example, and
29 one with the longest heritage, is the treatment of all cancer as one endpoint and the combination
30 of the single chemical probabilistic risks using the formula for statistical independence. The result
31 is one probability (risk) for cancer from all the chemical exposures. While this approach could
32 equally well be applied to other toxic endpoints, the differences in how the body reacts to non-
33 carcinogenic insults, and the consequent assumption of a toxicological threshold for many non-
34 cancer effects, has led to a weaker quantitative measure for general risk assessment, the Hazard
35 Index. The formula for the composite cancer risk is preferably applied for mixtures of chemicals
36 with different underlying toxic mechanisms. In contrast, the Hazard Index is best applied for
In some cases, it will involve a complex single-stressor exposure over a period of time where response to exposure
is not constant. This could be the result in changes in disease state, vulnerability, intervening exposure to different stressors, or
other factors which make the response profile change over time.
H-56
-------
DRAFT- Risk Assessment Forum Review Draft - August 2, 2001 - Do Not Quote or Cite
1 lexicologically similar chemicals and is specific to each target organ. The underlying principle,
2 called dose addition (or concentration addition), is also used when converting the multiple
3 exposure levels of the mixture components into the toxically equivalent exposure to one so-called
4 index chemical in the mixture. This latter procedure, called the Relative Potency Factor approach
5 (USEPA, 2001a; Hertzberg et al., 1999) can be specific to one target organ, or if the similarity is
6 justified on mechanistic grounds, can be applied to all toxic endpoints for the chemical group
7 being assessed. The mixture risk is determined from the dose-response curve for the index
8 chemical and so will give different risk estimates for each endpoint. The resulting mixture risks,
9 then, are presented separately for each toxic endpoint or each target organ.
10
11 Most multichemical exposures involve dissimilar chemicals, such as metals and pesticides,
12 and so are likely to contribute to joint toxicity by other than dose-additive means. In many cases,
13 the component toxicities influence each other (i.e., are not independent) and so must be
14 considered simultaneously. The traditional approach to toxicologic interdependence has been the
15 determination of synergism and antagonism for categories of pairwise interactions (i.e., those
16 involving just two chemicals). The present EPA mixture guidance (USEPA, 200la) uses such
17 categories in a interaction-based Hazard Index. This modified Hazard Index incorporates the
18 weight of evidence for pairwise interactions into a formula that adjusts each chemical's
19 contribution to toxicity by all the possible toxic interactions with the other chemicals in the
20 mixture. While immediately useful for regulatory decisions, especially for mixtures of only a few
21 chemicals, such approaches are of limited use and questionable accuracy when addressing more
22 complex mixtures. Current research efforts are seeking to identify toxicologic principles of joint
23 action that are applicable to mixtures of many chemicals (Portier, 2001; Yang, 2001; Hertzberg
24 and Teuschler, 2001).
25
26 Another method for assessing the combined risk of a mixture is to use data obtained from
27 testing the mixture itself, rather than building up the mixture risk from data for the component
28 chemicals. Testing of whole mixtures is expensive because environmental mixtures do not stay in
29 constant total dose or composition, forcing testing of many variations of the same mixtures. One
30 relatively inexpensive test method, called the comparative potency approach, involves in vitro or
31 short-term in vivo experiments that are then numerically scaled or extrapolated to public health
32 risk12 (Albert, et al., 1983; Lewtas, 1985, 1988; Gandolfi, et al., 1995). Whereas dose addition
33 combines risks of lexicologically similar chemicals, comparative potency models the risks for
34 groups of lexicologically similar mixtures, an approach that requires considerable scientific
35 judgment.
36
37
This presupposes the availability of an "index chemical", for which both the simple toxicity test data (e.g., skin
painting assay, enzyme activity) and the more comprehensive test data (e.g., 2-year cancer bioassay from which a potency
estimate has been derived) are available.
H-57
-------
DRAFT- Risk Assessment Forum Review Draft - August 2, 2001 - Do Not Quote or Cite
1
2 4.1.2. Other Impacts or Effects
3
4 Just as the effects from chemical stressors discussed in the previous section need to be
5 sorted into similar effects before being combined, the effects from non-chemical stressors also
6 need to be sorted into similar effect groups. There are a wide variety of effects from biological
7 stressors, for example, and these can be grouped into a number of categories by the types of
8 hazard they pose. Biological stressors, like their chemical counterparts, can interact and change
9 the overall risk in non-additive ways13. Obviously, there is an additional difference between
10 chemicals and biological stressors when evaluating exposure. Chemicals may degrade or
11 accumulate in the environment or in tissue, but possible growth and transmittal of biological
12 vectors adds another dimension to the challenge of evaluating exposure.
13
14 As cumulative risk assessment requires a broad focus shaped by aspects of the specific
15 problem, other impacts besides chemical-based and biological-based effects may need to be
16 considered and evaluated. As an example, current physical and mental health status and past
17 exposure histories may be a cumulative risk stressor. Economic considerations such as economic
18 status, community property values, source of income, level of income, and standard of living may
19 be stressors in that they affect susceptibility and exposure of subpopulations to certain other
20 stressors. Risks resulting from chemical or biological stressors may be significantly affected by
21 "vulnerability factors" such as lack of health care or genetic predisposition to some diseases and
22 effects. Community traditions and beliefs may affect activity patterns and behaviors and therefore
23 affect exposure to stressors as well as the risk management options deemed acceptable.
24 Depending on the scope of the assessment, so-called lifestyle factors such as smoking habits,
25 nutritional habits and others may be important components of overall risk. Finally, there may be
26 some additional (but hopefully, lesser) risks associated with acceptable remedial options, since
27 adverse effects can be associated with construction and implementation of a remedy or risk
28 reduction option.
29
30 In trying to assess all of these different types of stressors, it is helpful to determine what
31 types of effects the stressors produce, and then to try to group stressors by like effects. In an ideal
32 situation - one quite remote from today's state of the science, to be sure - one would also know
33 the mechanism or mode of action by which the stressor causes the effect, allowing more refined
34 grouping by mechanism/mode of action.
35
36 4.1.2.1. Stress-Induced Risks.
37
A person weakened by one disease may be devastated by a second disease infection which, if the person were
healthy, would be fought off easily. This is typical of AIDS victims, for example.
H-58
-------
DRAFT- Risk Assessment Forum Review Draft - August 2, 2001 - Do Not Quote or Cite
1 The Agency for Toxic Substances and Disease Registry (ATSDR) held an expert panel
2 workshop in 1995 on the subject of psychological responses to hazardous substances (ATSDR,
3 1995). In this report, the panel noted that there is "a significant lack of information" about how
4 often communities near hazardous waste sites or spills suffer chronic stress reactions, but that
5 psychological stress causes both psychological changes that can be measured by self-reports and
6 objective tests, as well as physical changes such as increased blood pressure, heart rate, and
7 biochemical parameters such as changes in stress hormones. Assessing the levels of stress, and
8 their potential contribution to risk, is difficult for a variety of reasons. The report notes that
9 "unlike the damage and injuries caused by a natural disaster, many toxic substances are invisible to
10 the senses.... In the face of no external cues and uncertain circumstances, each person affected by
11 a hazardous exposure develops their own beliefs about the nature of the resultant harm. These
12 beliefs are based on the facts available to them, pre-existing opinions, cultural factors, sensory
13 cues, and the beliefs of leaders and others in the community. On the other hand, scientists tend to
14 rely on objective data produced by specialized testing that is subject to statistical analysis....
15 Unlike a natural disaster, which hits and has a low point after which recovery can begin, the
16 response to a hazardous waste site can take 12 to 20 years."
17
18 Although the ATSDR report indicates that stress related to hazardous chemicals in the
19 community can show measurable physical effects, they stopped short of saying that long-term
20 health effects from this stress can be converted to risk estimates at this time. One of the questions
21 the panel was asked to address was, "Given what is known regarding the psychology of stress, are
22 there interactions between chronic stress and exposure to neurotoxicants that could shift the dose-
23 response curve for neurotoxins?" The panel concluded:
24
25 A methodology does not exist that would allow for discrimination between stress or
26 neurotoxicant-mediated effects in community-based studies.... Experimental animal data
27 exist to suggest that stress levels can modulate a toxic response; however, the question of
28 specificity remains. Given that stress can induce or unmask a latent effect of a toxicant,
29 there is the possibility that chronic stress could alter basal levels of neurofunctioning and
30 shift the threshold for neurotoxi city. Indeed, one may find a shift in the dose response to a
31 neurotoxi cant; however, a specific effect of the neurotoxicant needs to be examined in
32 greater detail than the generalized non-specific endpoints. Detecting such a shift would
33 require the knowledge of toxicant-specific biological mechanisms of actions, which most
34 often are not known. (ATSDR, 1995, page 30)
35
36 The ATSDR report made many suggestions for research to fill data gaps in this area, and
37 scientists may make significant progress in this area in the coming years.
38
39 4.1.2.2. Quality-of-Life Risks.
40
41 Another group of stressors and effects whose evaluation may require a different approach
H-59
-------
DRAFT- Risk Assessment Forum Review Draft - August 2, 2001 - Do Not Quote or Cite
1 from the traditional NRC risk paradigm are the quality-of-life issues. To evaluate the effects from
2 these types of stressors, a more deliberative approach is needed than is used in, say, cancer risk
3 analysis. EPA's Guidebook to Comparing Risks and Setting Environmental Priorities (EPA,
4 1993b) suggests a six-step process in Quality-of-Life Analysis:
5
6 1. Identify impacts and determine the values of the community.
7
8 2. Identify and define evaluative criteria.
9
10 3. Collect and analyze data on impacts.
11
12 4. Characterize impacts for all problem areas.
13
14 5. Present findings and rank problem areas for quality-of-life impacts.
15
16 6. Analyze future environmental conditions and risk management considerations.
17
18 Quality-of-Life impacts are determined by analyzing a set of criteria developed for each
19 community, depending on what they value. Stressors are those things that threaten to degrade the
20 quality-of-life criteria for that community. An example of a set of quality-of-life criteria, and their
21 descriptions, is given below. These criteria were developed by the State of Vermont's Agency of
22 Natural Resources (State of Vermont, 1991):
23
24 Impacts on Aesthetics: Reduced visibility, noise, odors, dust and other unpleasant sensations,
25 and visual impact from degradation of natural or agricultural landscapes.
26
27 Economic Weil-Being: Higher out-of-pocket expenses to fix, replace, or buy items or services
28 (e.g., higher waste disposal fees, cost of replacing a well, higher housing costs), lower income or
29 higher taxes paid because of environmental problems, and health-care costs and lost productivity
30 caused by environmental problems.
31
32 Fairness: Unequal distribution of costs and benefits (e.g., costs and benefits may be economic,
33 health, aesthetic).
34
35 Future Generations: Shifting the costs (e.g., economic, health risks, environmental damage) of
36 today's activities to people not yet able to vote or not vet born.
37
38 Peace of Mind: Feeling threatened by possible hazards in air or drinking water, or potentially
39 risky structures of facilities (e.g., waste sites, power lines, nuclear plants), and heightened stress
40 caused by urbanization, traffic, etc.
41
H-60
-------
DRAFT- Risk Assessment Forum Review Draft - August 2, 2001 - Do Not Quote or Cite
1 Recreation: Loss of access to recreational lands (public and private), and degraded quality of
2 recreation experience (e.g., spoiled wilderness, fished-out streams).
3
4 Sense of Community: Rapid growth in population or number of structures, or development that
5 changes the appearance and feel of a town; loss of mutual respect, cooperation, ability, or
6 willingness to solve problems together; individual liberty exercised at the expense of the
7 individual; the loss of Vermont's landscape and the connection between the people and the land.
8
9 Vermont's experience in evaluating these criteria was described as a qualitative description
10 of risk:
11
12 Because most of these seven criteria are intangible, they are extremely difficult to measure
13 or quantify. The Quality-of-Life Work Group described how each problem area affects
14 each criterion and how widespread or intense the effects are. Although these non-
15 quantitative descriptions of risk often lack precision and scientific objectivity, they focus
16 attention on specific critical issues and thus are useful tools for comparing the problems
17 systematically and consistently. (State of Vermont, 1991)
18
19 Quality-of-life issues can encompass much more than the criteria used here as an example.
20 Some cumulative risk assessments may include quality-of-life criteria as measures of effects, in
21 addition to human health effects or ecological effects. How these very different types of risks may
22 be included in a cumulative assessment is discussed in the following section.
23
24 4.1.3. Combining Different Types of Risk
25
26 An important aspect of the concept of multiple-agent cumulative risk is that it represents
27 the combined risks from the multiple agents or stressors acting together. This means that a
28 stressor by stressor listing of risks does not constitute a cumulative assessment unless this listing
29 can be interpreted in a way that provides an integrated characterization of the overall risk.
30 Therefore, an important cumulative risk assessment activity is determining how (if at all possible)
31 to combine disparate measures of risk and present them in an integrated manner. This is not to say
32 that all cumulative risk assessments must use a single, common metric to describe overall risk, but
33 that the combined effects of the stressors acting together should be discussed and characterized.
34
35 The assessment of a single stressor often results in the identification and, possibly, the
36 quantifying of a variety of hazards and risks. For example, a single stressor may be associated
37 with adverse human health effects that result from exceeding a threshold exposure during a brief
38 period of time. These "risks" are often represented by using Margins of Exposure (MOEs) as
39 surrogates (i.e., the margin that exists between environmental exposures and the highest dose
40 believed to be without adverse effects) (USEPA, 1996c). This same stressor may also be
41 associated with adverse health effects that result from longer term or lifetime exposures. These
H-61
-------
DRAFT- Risk Assessment Forum Review Draft - August 2, 2001 - Do Not Quote or Cite
1 exposures may be presented as the percent of a reference dose (%RfD) or other chronic dose
2 believed to be without adverse effects. Finally, if the same stressor is associated with cancer, risks
3 may be presented as a probability of developing cancer.
4
5 The goal of a cumulative risk assessment is to portray disparate risks in a manner that will
6 inform the decision-making process. The general approach to multichemical assessment has been
7 to present separate risk estimates for each toxic endpoint of concern. This approach can be
8 expanded to also include non-chemical stressors in a cumulative risk assessment. Even so, one is
9 left with a complex matrix of hazards and risks for various stressors.
10
11 One, but certainly not the only, approach to simplifying this problem is to collapse this "n-
12 dimensional matrix" of hazards and risks into a few or even a single measure (Murray, 1994).
13 However, this requires converting the various measures of risk to a common metric or otherwise
14 translating them into a common scale or index. Some methods for combining disparate measures
15 of risk are briefly described below.
16
17 4.1.3.1. Converting Adverse Effects to a Common Metric
18
19 As discussed at the beginning of Chapter 3, there are several different theoretical
20 approaches to cumulative risk assessment. Some of these require synthesizing a risk estimate (or
21 risk indication) by "adding up" risks for different parts of the risk picture. Actual mathematical
22 addition, of course, requires a common metric. Finding a common metric for dissimilar risks
23 (cancer vs. non-cancer, human vs. ecological, etc.) is not strictly an analytic process, since some
24 judgments must be made as to how to link two or more separate scales of risks. These judgments
25 often involve subjective values, and because of this, it is a deliberative process.
26
27 As an example of combining different effects into a common metric and the consequent
28 judgment needed to achieve a common metric, the EPA Office of Pollution Prevention and Toxics
29 in 1999 released its CD-ROM called "Risk-Screening Environmental Indicators Model, Version
30 1.0" (USEPA, 1999h)14. In this model, emissions for both carcinogens and non-carcinogens are
31 weighted by a toxicity factor so that they can be combined in a risk-based screening "score" for a
32 particular geographic area. The scale for this weight for carcinogens is related to the unit risk
33 factor, and the weight for the non-carcinogens is based on the RfD. According to the authors, it is
34 possible to relate these two scales by making a judgment as to how they relate. They note that in
35 their case, "when combining cancer and noncancer endpoints, it is assumed that exposure at the
36 RfD is equivalent to a 2.5 x 10"4 cancer risk" (Bouwes and Hassur, 1998).
37
38 Obviously, as Bouwes and Hassur acknowledge, equating an Hazard Quotient value of 1.0
As of this writing, EPA has RSEI version 2.0 in beta test. Details are at www.epa.gov/oppt/env_ind'beta_test.htm.
H-62
-------
DRAFT- Risk Assessment Forum Review Draft - August 2, 2001 - Do Not Quote or Cite
1 (exposure at the RfD) with a cancer risk of 2.5 x 10"4 is a judgment that is outside the strictly
2 analytic part of an assessment; the equating of the two points in the respective scales represents a
3 value judgment and as such can be debated. This particular part of the assessment is deliberative
4 in nature. In most cases, construction of a single scale for different types of endpoints will involve
5 comparative risk, a field where different types of risks or endpoints are ranked, compared, or
6 converted to a scale based on the judgments and values of the persons doing the assessments
7 (USEPA, 1993b, 1998f, 1999f). When converting such diverse endpoints as human health,
8 ecological, and quality of life, comparative risk is almost always involved, and this makes
9 combining of diverse risks a deliberative rather than an analytic process.
10
11 There have been some attempts to quantify diverse risks in a common metric without
12 resorting to the values needed as input for comparative risk. It has been suggested that "time is
13 the unit of measure for the burden of disease"; whether the disease results in disability or
14 premature mortality (Murray, 1994). Based on this premise, economic analyses of the costs and
15 benefits of disease intervention strategies have used Quality Adjusted Life Years (QALYs) and
16 Disability Adjusted Life Years (DALYs) as the metrics for the adverse effects of disease. These
17 metrics are intended to reflect the years of life spent in disease states and the years of life lost due
18 to premature mortality resulting from disease as a surrogate measure or risk from a variety of
19 different types of effect.
20
21 But even if this conversion of effects into QALYs or DALYs were successful, for diseases
22 that result in periods of morbidity and disability (but not death), weighting factors (based on
23 judgments) are used to equate time spent in various disease states with years lost to mortality. In
24 this way, dissimilar adverse effects can be combined to provide a single measure of disease
25 burden. However, it should be noted that aggregation of effects in this manner obscures the
26 meaning of the final measure. QALYs and DALYs do not represent an actual shortening of the
27 lifespan but are indicators of the overall degradation of well-being that results from various
28 disease states. Therefore, QALYs and DALYs may be best suited for ranking and comparative
29 analyses.
30
31
32 Experience with applying such measures as QALYs and DALYs to environmental risk
33 problems is extremely limited. Some very early methods development work has been initiated
34 which explores the use of QALYs for combining microbial and disinfection by-product risks
35 (USEPA, 1998f). However, some concerns have been raised about the adequacy of such
36 measures, especially when integrated with economic information for decision making USEPA,
37 2000d). Further methods development work is needed to improve the utility of QALYs and
38 DALYs for environmental risk assessments; especially with respect to the incorporation of
39 uncertainty (USEPA, 1999f).
40
41 Categorical regression may provide another tool for combining disparate effects using a
H-63
-------
DRAFT- Risk Assessment Forum Review Draft - August 2, 2001 - Do Not Quote or Cite
1 common metric. In this approach, adverse effects are assigned to severity categories (again, a
2 judgment making the process deliberative) and the ordered categories are regressed against
3 increasing dose (Teuschler et al., 1999). The results of the regression analysis may provide an
4 RfD that can reflect a variety of effects. Furthermore, the probability of experiencing effects
5 associated with a particular severity category at doses above the RfD can be determined. To date,
6 categorical regression has been applied to data for individual chemicals and has been used to
7 compare chemicals with similar effects (see Dourson et al., 1997 and Teuschler et al., 1999). The
8 use of categorical regression as a tool for combining disparate effects will require considerable
9 methods development research.
10
11 4.1.3.2. Translating Adverse Effects into an Index
12
13 Although methods such as described in the previous section have been used in screening,
14 ranking, and priority-setting exercises, EPA currently uses no health risk assessment procedures
15 for regulatory analyses that combine dissimilar toxic endpoints. The Superfund program uses a
16 screening tool (USEPA, 1989a) that combines all risks using a Hazard Index formula. Any
17 situations where acceptable risk cannot be assumed are further assessed by separating the toxic
18 endpoints. EPA has used "decision indices" based on dissimilar measures, and while they do not
19 produce risk estimates, the indices still prove useful. The approach involves developing a
20 composite score - or index - from measures of various risk dimensions (e.g., public deaths,
21 occupational deaths, and morbidity).
22
23 Fischhoff et al. (1984) provide an example of this approach as applied to the evaluation of
24 energy technologies. In this case, disparate risks are assigned a score from a fixed scale (e.g.,
25 from 0, representing no risk, to 100, representing the worst risk for that dimension). The scores
26 are then weighted to reflect value judgments about the importance of the various risk dimensions
27 and the composite score is calculated by summing the individual weighted scores. Again, the
28 aggregation of dissimilar adverse effects obscures the meaning of the final score making it more
29 appropriate for ranking and comparative analyses.
30
31 Various environmental risk indices have been developed and applied to ranking and
32 comparative analyses. Often, these indices employ surrogate measures for risk rather than using
33 actual calculations of the probability of adverse effects. One such index is the Hazard Ranking
34 System (HRS) [47 Fed Reg. 31219, dated July 16, 1982, and amended 55 Fed. Reg. 51532,
35 dated December 14, 1990], used to place uncontrolled waste sites on the National Priorities List
36 (NPL) for Superfund. This index is based on the likelihood of off-site movement of waste, the
37 toxicity of the waste, and the people and sensitive environments that may be affected. It also uses
38 corrosivity, toxicity, fire hazard and other factors, all scored and combined into one numerical
39 indicator of overall hazard potential. Such an approach for a composite index has been suggested
40 for communication of cumulative risk (Hertzberg, 2000).
41
H-64
-------
DRAFT- Risk Assessment Forum Review Draft - August 2, 2001 - Do Not Quote or Cite
1 Recently, EPA has been working on several index-based approaches to dealing with
2 cumulative risk issues. EPA Region III and the Office of Research and Development have been
3 jointly working to develop a Potential Risk Indexing System (USEPA, 1993c, 1995c, 1997b).
4 This index also uses a vulnerability index, and gauges the overall well-being of a locale and
5 various subpopulations. Again, the volume and toxicity of released stressors serve as surrogate
6 measures of risk in developing this index.
7
8 EPA's Region VI has developed a system called the Cumulative Risk Index Analysis
9 (CRIA), primarily for NEPA-type assessments (Osowski, et al., 2001). The CRIA contains some
10 90 criteria to evaluate the health of an area and its ecosystem/human populations. Each criterion,
11 which leads to an indexing of 1-5, has been through the deliberative process, peer review, and is
12 well documented.
13
14 Combining diverse effects and risk using either common metrics or indices each have pros
15 and cons. A weakness of the index approach is that information is "lost," and the meaning of the
16 final score can be obscured, by aggregating dissimilar information through index scores. One
17 strength, however, is common to both approaches. Both techniques have the ability to
18 incorporate social values in an explicit and quantitative manner in the risk assessment. For
19 example, in the derivation of DALYs, weights can be used to reflect the different social roles
20 people play as they age (Murray, 1994). In the composite scores developed by Fischhoff (1984),
21 public concern was incorporated as an adverse effect. This is an important feature for methods
22 that will be applied to cumulative risk assessments, especially for communities. Given that
23 cumulative assessments have a community/population focus, the ability to incorporate social
24 values in an overall assessment of well-being will be critical.
25
26 4.1.3.3. Other Approaches
27
28 Another way that cumulative risk may be expressed is as margins of exposure (MOEs).
29 Margins of exposure, defined as the no adverse effect level (NOAEL)15 divided by estimated
30 exposure, give a sense of how close estimated exposures in a situation might be to levels that
31 could cause harm.. Much like a hazard index, they provide perspective, but without providing a
32 statement of the probability of effects occurring if the exposure is greater than the NOAEL.
33 MOEs can be used as an indication of possible risk, and can be mathematically combined across
34 routes of exposure. The advantage of using MOEs for expressing risk is that one can preserve the
35 route-specific nature of the different exposures and then add them to generate a total MOE. The
36 inverse of the different pathway MOEs are added together and then the inverse of that sum is
37 taken as the total MOE (USEPA, 2000i)
Other points of departure, such as the benchmark response (see USEPA, 1996c, 2000h), may sometimes be used
instead of the NOAEL.
H-65
-------
DRAFT- Risk Assessment Forum Review Draft - August 2, 2001 - Do Not Quote or Cite
1 Collapsing the various measures of risk into a single entity (whether a common metric or
2 index) may not be appropriate in every case. The inability to construct or inappropriateness of
3 constructing a single numeric does not necessarily preclude the preparation of a cumulative risk
4 assessment. As long as the disparate measures of risk can be presented in a manner that conveys
5 a sense of the combined well-being of a community or population, the goals of cumulative risk
6 assessment can be achieved. Geographic Information Systems (GIS) and related mapping
7 techniques (e.g., Environmental Defense, 2001) appear to hold some promise as tools for
8 presenting integrated information concerning cumulative risks without mathematically combining
9 disparate measures. As with the common metric and index approaches, however, considerable
10 methods development work remains to be completed.
11
12 4.1.3.4. General Issues with Combining Risks
13
14 As described above, each approach to portraying the results of a cumulative risk
15 assessment has benefits and disadvantages. While common metrics and indices can incorporate
16 social values in an explicit and quantitative manner, the meaning of the final measure can be
17 obscured by aggregation of dissimilar effects. The abstract meaning of the final measure could
18 lead to difficulties when communicating the results of the cumulative risk assessment to the
19 public. Graphical and mapping techniques do not necessarily overcome such problems with
20 communication. While these techniques may avoid some of the problems associated with the
21 mathematical aggregation of dissimilar effects, it can be difficult to accurately describe the
22 information a graphic is intended to convey.
23
24 The ideal with regard to cumulative health risk assessment may be when we can make
25 projections about the potential for a particular complex exposure to cause particular effects to
26 different physiological systems, and integrate these projections into a qualitative characterization
27 of potential overall impact to human health.
28
29 Because we have relatively little experience in combining different types of risk, a key
30 issue is the need for methods development in this area. The approaches described above indicate a
31 beginning. Additional exploratory work is needed, however, to further develop existing methods
32 and to find additional methods that are flexible, can incorporate social values, are easy to
33 communicate, and provide an integrated portrayal of the overall well-being of a community and its
34 various subpopulations.
35
36
37 4.2. Risk Description
38
39 The ultimate useable product in the risk assessment process is the risk characterization, in
40 which the information from all the steps is integrated and an overall conclusion about risk is
41 synthesized that is complete, informative, and useful for decision-makers. The nature of the risk
H-66
-------
DRAFT- Risk Assessment Forum Review Draft - August 2, 2001 - Do Not Quote or Cite
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
characterization will depend on the information
available, the regulatory application of the risk
information, and the resources (including time)
available. It is important to identify and discuss all
major issues associated with determining the nature
and extent of the risk. Further, the EPA
Administrator's March 1995 Policy for Risk
Characterization (U.S. EPA, 1995a) specifies that
a risk characterization "be prepared in a manner
that is clear, transparent, reasonable, and consistent
with other risk characterizations of similar scope
prepared across programs in the Agency." In
short, estimates of health risk are to be presented in
the context of uncertainties and limitations in the
data and methodology.
The 1995 Guidance for Risk
Characterization (USEPA, 1995b) lists several
guiding principles for defining risk characterization
in the context of risk assessment (see text box),
both with respect to information content and
uncertainty aspects and with respect to descriptions
of risk. EPA has recently published a handbook on
risk characterization (USEPA, 2000c).
Risk assessments are intended to address or
provide descriptions of risk to one or more of the
following: (1) individuals (including highly
susceptible individuals) exposed at average levels
and those in the high-end portions of the risk
distribution; (2) the exposed population as a whole; and (3) important subgroups of the
population such as highly susceptible groups or individuals (e.g., children), if known. Risk
predictions for sensitive subpopulations are a subset of population risks. Sensitive subpopulations
consist of a specific set of individuals who are particularly susceptible to adverse health effects
because of physiological (e.g., age, gender, pre-existing conditions), socioeconomic (e.g.,
nutrition), or demographic variables, or significantly greater levels of exposure (USEPA, 1992a).
Subpopulations can be defined using age, race, gender, and other factors. If enough information
is available, a quantitative risk estimate for a subpopulation can be developed. If not, then any
qualitative information about subpopulations gathered during hazard identification should be
summarized as part of the risk characterization.
RISK CHARACTERIZATION GUIDING
PRINCIPLES
Regarding information content and uncertainty aspects:
• The risk characterization integrates the information from
the exposure and dose-response assessments, using a
combination of qualitative information, quantitative
information, and information regarding uncertainties.
• The risk characterization includes a discussion of
uncertainty and variability.
• Well-balanced risk characterizations present risk
conclusions and information regarding the strengths and
limitations of the assessment for other risk assessors, EPA
decision-makers, and the public.
Regarding risk descriptors:
• Information about the distribution of individual exposures
is important to communicating the results of a risk
assessment.
• Information about population exposure leads to another
important way to describe risk.
• Information about the distribution of exposure and risk for
different subgroups of the population are important
components of a risk assessment.
• Situation-specific information adds perspective on
possible future events or regulatory options.
• An evaluation of the uncertainty in the risk descriptors is
an important component of the uncertainty discussion in
the assessment.
Source: USEPA, 1995b.
H-67
-------
DRAFT- Risk Assessment Forum Review Draft - August 2, 2001 - Do Not Quote or Cite
1
2
3
4
5
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
4.3. Uncertainty Analysis
In their 1990 book Uncertainty: A Guide to Dealing with Uncertainty in Quantitative
Risk and Policy Analysis, Morgan and Henrion (1990) note that historically, the most common
approach to uncertainty in policy analysis (including in risk assessment) has been to ignore it. In a
section titled, "Why Consider Uncertainty?", they advance three primary reasons, all of which are
especially relevant to an analytic-deliberative process such as cumulative risk assessment. They
suggest that it is important to worry about uncertainty:
• when one is performing an analysis in which people's attitude toward risk is likely to be
important, for example, when people display significant risk aversion;
• when one is performing an analysis in which uncertain information from different sources
must be combined. The precision of each source should help determine its weighting in the
combination; and
• when a decision must be made about whether to expend resources to acquire additional
information. In general, the greater the uncertainty, the greater the expected value of
additional information.
Although all of Morgan and
Henri on's "ten commandments" are
commendable, and several have been
discussed elsewhere in this Framework,
we should look more closely at numbers
6-8 in the box at right for some insight
into uncertainty analysis. There are
many resources available which talk in
detail about how to perform uncertainty
analysis (e.g., USEPA, 1997c, Morgan
and Henrion, 1990). While we believe
that detailed instruction on how to
perform uncertainty analysis to be
beyond the scope of this Framework,
we believe that a discussion of some
general principles are in order.
Morgan & Henrion's "Ten Commandments" for Good
Policy Analysis
1. Do your homework with literature, experts, and users.
2. Let the problem drive the analysis.
3. Make the analysis as simple as possible, but no simpler.
4. Identify all significant assumptions.
5. Be explicit about decision criteria and policy strategies.
6. Be explicit about uncertainties.
7. Perform systematic sensitivity and uncertainty analysis.
8. Iteratively refine the problem statement and the analysis.
9. Document clearly and completely.
10. Expose the work to peer review.
Source: Morgan and Henrion, 1990.
Cumulative risk assessment will usually be used in a decision-making process to help
inform the decision-maker(s). For this reason, it is important that the decision makers be made
explicitly aware of any assumptions that may significantly affect the conclusions of the analysis
H-68
-------
DRAFT- Risk Assessment Forum Review Draft - August 2, 2001 - Do Not Quote or Cite
1 (item #6 in the box above). Morgan and Henrion suggest that these assumptions include:
2
3 • the main policy concerns, issues, or decisions that prompted the assessment to be done;
4 • the evaluation criteria to be used to define issues of concern or options;
5 • the scope and boundaries of the assessment, and ways in which alternate selections might
6 influence the conclusions reached;
7 • soft or intangible issues that are ignored or inadequately dealt with in the quantitative
8 analysis (e.g., intrinsic value of wilderness, equity of distribution of risks and benefits);
9 • approximations introduced by the level of aggregation or by level of detail in models;
10 • value judgments and tradeoffs; and
11 • the objective function used, including methods of combining ratings on multiple criteria
12 (or combining risk scales), [adapted from Morgan and Henri on, 1990]
13
14 Identifying significant assumptions can often highlight "soft" uncertainties that are not
15 easily quantified, and are therefore often left out of a quantitative uncertainty analysis.
16 Nevertheless, these "soft" assumptions can many times contribute more to the overall uncertainty
17 of the assessment than the factors more easily quantified.
18
19 In item #7 in Morgan and Henrion's "ten commandments," they list three types of
20 uncertainty that analysts should explicitly include:
21
22 • uncertainty about technical, scientific, economic, and political quantities (e.g., quantities
23 like rate constants often lend themselves to quantitative uncertainty estimates relatively
24 easily);
25 • uncertainly about the appropriate functional form of technical, scientific, economic, and
26 political models (e.g., are the models used, such as dose-response models, biologically
27 sound?);
28 • disagreements among experts about the values of quantities or the functional form of
29 models (e.g., different health scientists using different forms of dose-response models).
30
31 In Item #8 in the box on the previous page, Morgan and Henrion suggest that an assessor
32 needs to find out which assumptions and uncertainties may significantly alter the conclusions, and
33 that process can be done using sensitivity and uncertainty analysis. Techniques for these include:
34
35 • deterministic, one-at-a-time analysis of each factor, holding all others constant at nominal
36 values;
37 • deterministic joint analysis, changing the values of more than one factor at a time;
38 • parametric analysis, moving one or a few inputs across reasonably selected ranges to
39 observe the shape of the response; and
40 • probabilistic analysis, using correlation, rank correlation, regression, or other means to
41 examine how much of the uncertainty in the conclusions is attributable to which inputs.
H-69
-------
DRAFT- Risk Assessment Forum Review Draft - August 2, 2001 - Do Not Quote or Cite
1
2 Finally, Morgan and Henri on answer the question of why we should consider uncertainty
3 analysis with the following point. "Policy analysts have a professional and ethical responsibility to
4 present not just "answers" but also a clear and explicit statement of the implications and
5 limitations of their work. Attempts to fully characterize and deal with important associated
6 uncertainties help them to execute this responsibility better." (Morgan and Henrion, 1990)
7
8 4.4. The Information Provided by Cumulative Risk Assessment
9
10 It is important to clarify how cumulative risk assessment and this Framework document
11 relate to community assessments and community decision making. Certainly, the Agency's Risk
12 Characterization Handbook (USEPA, 2000c) emphasizes that whatever information is imparted,
13 it be transparent, clear, consistent, and reasonable. In simple terms, what can a cumulative risk
14 assessment tell us, and what can't it tell us?
15
16 4.4.1. Making Sense of Multiple Stressor Effects
17
18 The information provided by cumulative risk assessment is only a portion of the
19 information that communities and governments need to make informed decisions about risks. It
20 should not be the only consideration in decisions. There are almost always additional factors to
21 those considered in the assessment that affect health in a community (e.g., crime, drugs, health
22 care access, vehicle safety, climate, infectious disease, diet...). Community decision-making will
23 also take into account risks to the environment, and consideration about historical and cultural
24 values, as well as questions of fairness and distribution of risks. The methodology is not currently
25 well established to take all of these factors (stressors) into account in cumulative risk assessments.
26
27 Additionally, benefits that may be associated with chemical or other stressor exposures -
28 benefits such as jobs and useful products or services - may be important contexts for decisions on
29 the risks considered in cumulative risk assessments.
30
31 The Framework document is not an attempt to lay out protocols to address all the risks or
32 considerations that are needed to adequately inform community decisions. Rather, it is focused
33 on describing various aspects of cumulative risk, whether or not the methods or data currently
34 exist to adequately analyze or evaluate those aspects of the assessment. The Framework
35 document devotes considerable time to a discussion of improving the methods for a single part of
36 the broader picture - characterizing health risks associated with exposures to multiple chemicals
37 via multiple routes. Because of the limitations of the current state of the science, cumulative risk
38 assessments in the near future will not be able to adequately answer all questions posed by
39 stakeholders or interested parties. This does not mean, however, that they can't be useful in
40 providing insights to some of the questions asked; in fact, cumulative risk assessment may be the
41 best tool available to address certain questions dealing with multiple stressor impacts.
H-70
-------
DRAFT- Risk Assessment Forum Review Draft - August 2, 2001 - Do Not Quote or Cite
1
2 4.4.2. Cumulative Risk Assessments in a Public Health Context
3
4 The public, in a variety of forms, continually draws attention to health statistics, asking for
5 clarification of the relationship between environmental pollution (and risk assessments concerning
6 it) and public health. It is important to clarify 1) that to draw relationships between
7 environmental pollutant exposures and disease incidence, a body of epidemiological study is
8 necessary, and 2) trying to "work backwards" from health statistics to risk factors requires full
9 knowledge of the risk factors associated with the relevant disease(s).
10
11 Health statistics, including death rates and incidence of various diseases, illustrate the
12 impact of a variety of risk factors (e.g., smoking as well as environmental pollutants) and risk
13 reduction factors (e.g., exercise and good nutrition, as well as pollution control measures).
14 Indeed, population health statistics are reflective of all risk and risk reduction factors in a
15 population's history-to-date. Even the best cumulative risk assessment given today's state of the
16 science would fall short of being able to include an evaluation of the magnitude and interactions of
17 all stressors and effects. At best, the risk estimates of a cumulative risk assessment will reflect
18 some of the risks which may be reflected in community health statistics. With rare exceptions16,
19 cumulative risk assessment estimates would not be expected to match exactly with community
20 health statistics, even for specific health endpoints such as specific cancers.
21
22 4.4.3. How Scope and Purpose of the Assessment Affect Results
23
24 Historically, the Agency's risk assessments were usually aimed at assessing the risks of
25 environmental pollutants to public health or the environment, for the purposes of prioritizing risk
26 management activities or triggering regulatory action. Although there was a wide variety of
27 specific pollutants - chemical, biological, radiological, noise - these were evaluated separately and
28 each in the context of being protective of public health or the environment. Given the need for
29 public health protective decisions, traditional risk assessment tools usually yield "upper confidence
30 level" and not "best estimates" of cancer risk, and are not designed to predict risk of noncancer
31 disease. Additionally, the many environmental pollutants comprise only some of the categories of
32 risks to public health. When public health risks are viewed from a population-based perspective,
33 many of the traditional risk assessments, while being quite adequate for answering the questions
34 for which they were commissioned, leave large gaps in understanding place-based (community)
35 public health issues. The Agency is doing more place-based assessments (both human health and
36 ecological) than in the past, but it will be some time before place-based assessments become
It is conceivable that high risks to rare specific effects could be comparable between a risk assessment and
community health statistics given current state of the art. To be sure this is not fortuitous, a substantial effort to match risk
assessment scenarios with actual histories or exposures would have to be made.
H-71
-------
DRAFT- Risk Assessment Forum Review Draft - August 2, 2001 - Do Not Quote or Cite
1 commonplace. Even with more cumulative risk assessments being done as time goes by, initial
2 efforts may also be largely driven by specific risk management needs and not driven by exactly the
3 same questions that a community would ask when inquiring about local health concerns. For this
4 reason, users of cumulative risk assessments are advised to carefully study the scope and purpose
5 of the assessment at hand, and determine whether it is suitable (or partly suitable) to answer
6 questions outside its stated scope and purpose.
7
8 Finally, much of the activities and data needed for cumulative risk assessment overlap with
9 the jurisdiction of other public health agencies, and academia. The most successful cumulative risk
10 assessments of the future are likely to be those where cooperation among organizations (Federal,
11 State, private, environmental, academic, etc.) leads to use of the best data and tools for the
12 various parts of the assessment.
13
14
15 4.5. Using the Results of the Assessment
16
17 Once the results of an assessment are in hand, the assessment participants will usually
18 focus primarily on the use of those results. The intended use of the assessment was considered at
19 the beginning, in the Problem Formulation Phase, both to plan the assessment work and to set the
20 framework for what possible actions might be taken at this point.. A detailed discussion of the use
21 of the results of a cumulative risk assessment is beyond the scope of this document, but in
22 deciding on a course of action, other considerations will need to be taken into account along with
23 the results of the cumulative assessment.
24
25 As discussed in the Introduction, the results of the assessment should speak directly to the
26 question or questions addressed in the purpose for doing the assessment. Results from cumulative
27 risk assessments can also serve a variety of other purposes, however. Results may also be used to
28 meet regulatory mandates, to identify targets for enforcement actions, or to shape policy and
29 regulation. They may be used for general educational purposes not directly related to an
30 immediate decision on a course of action. Assessment results can also be used to set priorities for
31 voluntary or regulatory action, or to mobilize community efforts to address concerns.
32
33 If the goals of a cumulative risk analysis are to estimate the total risk from multi-chemical
34 and multi pathway exposure to individuals living within a geographical area of concern, then an
35 important objective is to identify the major risk contributors in order to understand the sources,
36 pathways, and stressors which contribute most to that overall risk. The results of a cumulative
37 risk assessment provide an additional tool for the risk manager, one that permits a more complete
38 accounting and more explicit analysis to target follow-up risk mitigation strategies toward those
39 stressors which most contribute to the population's risk.
40
41 If action to mitigate or prevent risk is the goal of the stakeholders, then options for action
H-72
-------
DRAFT- Risk Assessment Forum Review Draft - August 2, 2001 - Do Not Quote or Cite
1 discussed in the planning of the assessment can be re-evaluated in light of the results of the
2 assessment. Some of the issues after re-evaluating the action alternatives might include: "Is
3 regulatory authority available to address concerns or are voluntary actions better suited to address
4 the risks?" or "Can the concerns be addressed by the stakeholders involved in the assessment or
5 are the options for mitigation and prevention beyond the scope of their control?" In the latter
6 case, for example, siting issues are usually decided locally and may be within the authority of the
7 participants of a local assessment. In contrast, risk from mobile sources or acid rain are likely to
8 require action beyond the scope of a single local community. In that case, taking action will
9 require working with other communities and is likely to take more time. Discussion of the
10 options available for addressing results of a risk assessment will help to keep expectations in line
11 with possibilities.
12
13 Finally, it is important to keep in mind that the results of the risk assessment will be only
14 one of the factors that will need to be considered in making a decision on action to address the
15 risk. Risk information can make an important and valued contribution to the decision-making
16 process, but risk information, by itself, can not and should not determine the decision. Factors
17 such as the availability of resources for change, fairness and other community values, politics,
18 business and employment considerations, quality of life issues, concern for future generations,
19 etc., will also influence any decision made. In the siting example mentioned above, the assessment
20 may determine that the new facility does not significantly increase risk to the community and a
21 decision not to site the facility might still be made on the basis of a quality of life issue unrelated
22 to risk. Or, in contrast, a community may decide that the economic and employment benefits
23 outweigh the risks associated with the siting. Other risk factors not considered in the assessment
24 may also enter into the decision-making process. This can include both the environmental risks
25 not covered in the cumulative risk assessment as well as the non-environmental risks that may
26 affect a community. With limited resources, a community may use all available risk information to
27 most effectively target its resources.
28
29
H-73
-------
DRAFT- Risk Assessment Forum Review Draft - August 2, 2001 - Do Not Quote or Cite
1 5. GLOSSARY
2
3 [to be added]
4
5 6. REFERENCES
6
7 ACGIH, 1998. Threshold Limit Values for Chemical Substances and Physical Agents.
8 American Conference of Government Industrial Hygienists, Cincinnati, OH.
9
10 AIChE, 1992. Guidelines for Hazard Evaluation Procedures, 2nd Edition with Worked
11 Examples. Center for Chemical Process Safety, American Institute of Chemical Engineers, New
12 York, NY.
13
14 AIChE, 1996. Guidelines for Use of Vapor Cloud Dispersion Models, 2nd Ed. Center for
15 Chemical Process Safety, American Institute of Chemical Engineers, New York, NY.
16
17 AIHA, 2000. Emergency Response Planning Guidelines Series. Emergency Response Planning
18 Guidelines Committee, American Industrial Hygiene Association, Fairfax, VA.
19
20 Albert, Roy E., Joellen Lewtas, Stephen C. Nesnow, Todd W. Thorslund, and Elizabeth L.
21 Anderson, 1983. Comparative potency method for cancer risk assessment application to diesel
22 particulate emissions. Risk Analysis 3:101-117
23
24 American Heart Association, 2000. Stroke Risk Factors. Internet: www.americanheart.org.
25
26 APCA, 1999. "Cumulative & Aggregate Risk Evaluation System: CARES. Conceptual Model."
27 American Crop Protection Association. Washington, DC. Internet: www.alphacares.org/
28
29 ATSDR, 1995. Report of the Expert Panel Workshop on the Psychological Responses to
30 Hazardous Substances. Agency for Toxic Substances and Disease Registry, U.S. Department of
31 Health and Human Services. Atlanta, Georgia. Internet: www.atsdr.cdc.gov/HEC/PRHS/
32
33 Barnes, Donald G. and Michael L. Dourson, 1988. Reference dose (RfD): description and use in
34 health risk assessments. Regulatory Toxicology and Pharmacology 8:471-486.
35
36 Berenblum, Isaac, and Philippe Shubik, 1947. A new quantitative approach to the study of stages
37 of chemical carcinogenesis in the mouse's skin. British Journal of Cancer 1:384-391
38
39
40
41
H-74
-------
DRAFT- Risk Assessment Forum Review Draft - August 2, 2001 - Do Not Quote or Cite
1 Bouwes, Nicolaas W. and Steven M. Hassur, 1998. "OPPT's Risk-Screening Environmental
2 Indicators: Toxic Weights for Toxic Release Inventory (TRI) Chemicals and Chemical
3 Categories." Office of Pollution Prevention and Toxic Substances, Office of Prevention,
4 Pesticides, and Toxic Substances, U.S. Environmental Protection Agency. Washington, DC. April
5 28, 1998
6
7 Bullard, Robert D., 1990. Dumping in Dixie: Race, Class, and Environmental Quality.
8 Westview Press, Boulder, CO. ISBN: 0-8133-7954-7
9
10 Carpy, Serge A, Werner Kobel, and John Doe, 2000. Health risk of low-dose pesticides mixtures:
11 A review of the 1985-1998 literature on combination toxicology and health risk assessment.
12 Journal of Toxicology and Environmental Health Part B, 3:1-25
13
14 CEQ, 1997. "Considering Cumulative Effects Under the National Environmental Policy Act."
15 Council on Environmental Quality, Executive Office of the President, Washington, DC.
16
17 Clemen, Robert T., 1996. Making Hard Decisions: An Introduction to Decision Analysis. 2nd Ed.
18 Duxbury Press, Wadsworth Publishing Co., Belmont, CA. ISBN 0-534-26034-9
19
20 Cohen, Bernard L., 1991. Catalog of risks extended and updated. Health Physics 61:317-335.
21
22 DOT, 1998. "High-Speed Ground Transportation Noise and Vibration Impact Assessment." Final
23 Draft. Office of Railroad Development, Federal Railroad Administration, U.S. Department of
24 Transportation. Washington, DC. Report No. 293630-1 Internet:
25 http://projectl. parsons. com/ptgnechsr/noise_manual. htm
26
27 Dourson, Michael L., Linda K. Teuschler, Patrick R. Durkin, and William M. Stiteler, 1997.
28 Categorical regression of toxicity data: A case study using aldicarb. Regulatory Toxicology and
29 Pharmacology 25:121-129.
30
31 Environmental Defense, 2001. "Scorecard." Internet: http://www.scorecard.org
32
33 European Commission, 2000. "The European Multi-Hazard Risk Assessment Project
34 (T.E.M.R. A.P.)" Directorate General XII for Science Research and Development, Environment
35 and Climate - 1994/1998 - Climatology and Natural Hazard, European Commission. Brussels,
36 Belgium. Internet: http://phypc9.geo.ulg.ac.be/nouveau/temrap/MainWebPage.htm
37
38
39
40
41
H-75
-------
DRAFT- Risk Assessment Forum Review Draft - August 2, 2001 - Do Not Quote or Cite
1 FIFRA SAP, 2000. "Cumulative Risk Assessment Methodology Issues of Pesticide Substances
2 that Have a Common Mechanism of Toxicity." Report from Session II of the FIFRA Scientific
3 Advisory Panel Meeting of December 9, 1999 (Report dated February 4, 2000). FIFRA Scientific
4 Advisory Panel, Office of Science Coordination and Policy, Office of Prevention, Pesticides and
5 Toxic Substances, U.S. Environmental Protection Agency. Washington, DC. SAP Report 99-
6 06B
7
8 Fischhoff, Baruch, Watson, S., and Hope, C., 1984. Defining risk. Policy Sciences 17:123-139
9
10 Foran, Jeffrey A. and Susan A. Ferenc (Eds.), 1999. Multiple Stressors in Ecological Risk and
11 Impact Assessment. Society of Environmental Toxicology and Chemistry, SET AC Press,
12 Pensacola, Florida. ISBN 1-880611-32-5
13
14 Gandolfi, A.J., I.K. Brendel, R.L. Fisher, and J-P Michaud, 1995. Use of tissue slices in chemical
15 mixtures toxicology and interspecies investigations. Toxicology 105:285-290
16
17 GAO, 1983. "Siting of Hazardous Waste Landfills and Their Correlation with Racial and
18 Economic Status of Surrounding Communities." U.S. General Accounting Office. Washington,
19 DC. GAO/RCED 83-168
20
21 Haber, F., 1924. Zur Geschichte des Gaskrieges [On the History of Gas Warfare], in: Funf
22 Vortrage aus den Jahren 1920-1923 (Five Lectures from the years 1920-1923), Springer, Berlin,
23 76-92.
24
25 Hampshire Research Institute, 1999. "Review of an Aggregate Exposure Assessment Tool."
26 Presentation to the U.S. EPA FIFRA Scientific Advisory Panel, September, 1999. Hampton
27 Research Institute, Alexandria, VA. Internet:
28 www.epa.gov/scipoly/sap/1999/september/aggbkgd.pdf
29
30 Hampshire Research Institute, 2000. "Overview of the Fundamentals of Version 1.0 of LifeLine
31 Software for Modeling Aggregate and Cumulative Exposures to Pesticides." Hampshire Research
32 Institute, Alexandria, VA. Internet:
33 www.epa.gov/scipoly/sap/2000/september/fmal_fundamentals.pdf
34
35 Hertzberg, Richard C. "Communicating cumulative risk: sound bites from chaos." Luncheon
36 address, presented at the conference, Toxicology and Risk Assessment Approaches for the 21st
37 Century, April 10-13, 2000, Kings Island, OH.
38
39
40
41
H-76
-------
DRAFT- Risk Assessment Forum Review Draft - August 2, 2001 - Do Not Quote or Cite
1 Hertzberg, Richard C., Glenn Rice, and Linda K. Teuschler, 1999. "Methods for health risk
2 assessment of combustion mixtures." In: Hazardous Waste Incineration: Evaluating the Human
3 Health and Environmental Risks. S. Roberts, C. Teaf and J. Bean, eds. CRC Press LLC, Boca
4 Raton. Pages 105-148
5
6 Hertzberg, Richard C. and Linda K. Teuschler, 2001. "Ideas for Evaluating Quantitative Formulas
7 for Dose-Response Assessment of Chemical Mixtures," presented at the Application of
8 Technology to Chemical Mixture Research Conference, January 9-11, 2001, Fort Collins, CO.
9
10 HUD, 1991. "The Noise Guidebook." Office of Community Planning and Development, U.S.
11 Department of Housing and Urban Development. Washington, DC. HUD-953-CPD(l)
12
13 ILSI, 1999. A Framework for Cumulative Risk Assessment. International Life Sciences Institute,
14 Washington, DC. ISBN 1-57881-055-8
15
16 Johnson, Ted, Gary Mihlan, Jacky LaPointe, Kris Fletcher, Jim Capel, Arlene Rosenbaum,
17 Jonathan Cohen, and Pat Stiefer, 2000. "Estimation of Carbon Monoxide Exposures and
18 Associated Carboxyhemoglobin Levels for Residents of Denver and Los Angeles using
19 pNEM/CO (version 2.1)." Draft report prepared for Office of Air Quality Planning and Standards,
20 Office of Air and Radiation, U.S. Environmental Protection Agency. Research Triangle Park, NC.
21 Contract No. 68-D6-0064.
22
23 Kroschwitz, Jacqueline I. and Mary Howe-Grant (Eds.), 1994. Kirk-Othmer Encyclopedia of
24 Chemical Technology. 4th Edn. John Wiley and Sons, New York. ISBN 0-47152-677-0
25
26 Lantz, Paula M, James S. House, James M. Lepkowski, David R. Williams, Richard P. Mero and
27 Jieming Chen, 1998. Socioeconomic factors, health behaviors, and mortality: Results from a
28 nationally representative prospective study of US adults. Jour. Amer. Medical Assoc. 279:1703-
29 1709.
30
31 Lee, Charles, 1987. Toxic Waste and Race in the United States: A National Report on the Racial
32 and Socio-Economic Characteristics of Communities with Hazardous Waste Sites. United
33 Church of Christ Commission for Racial Justice. New York, NY.
34
35 Lewtas, Joellen, 1985. Development of a comparative potency method for cancer risk assessment
36 of complex mixtures using short-term in vivo and in vitro bioassays. Toxicol. Ind. Health 1:193-
37 203
38
39 Lewtas, Joellen, 1988. Genotoxicity of complex mixtures: Strategies for the identification and
40 comparative assessment of airborne mutagens and carcinogens from combustion sources.
41 Fundamentals of Applied Toxicology 10:571-589
H-77
-------
DRAFT- Risk Assessment Forum Review Draft - August 2, 2001 - Do Not Quote or Cite
1 Lynch, John W., George A. Kaplan, Elsie R. Pamuk, Richard D. Cohen, Katherine E. Heck,
2 Jennifer L. Balfour, and Irene H. Yen, 1998. Income inequality and mortality in metropolitan
3 areas of the United States. Am. J. Public Health 88:1074-1080
4
5 Morgan, M. Granger, and Max Henri on, 1990. Uncertainty: A Guide to Dealing with
6 Uncertainty in Quantitative Risk and Policy Analysis. Cambridge University Press, New York,
7 NY. ISBN 0-521-36542-2
8
9 Mottram, J.C., 1944. A developing factor in experimental blastogenesis. J. Pathol. Bacteriol.
10 56:181-187
11
12 Murray, C., 1994. Quantifying the burden of disease: the technical basis for disability-adjusted
13 life years. Bull. World Health Org. 72 (3):429-445
14
15 NAP A, 1995. Setting Priorities, Getting Results: A New Direction for EPA. National Academy of
16 Public Administration. Washington, DC. LCCN: 95-68048
17
18 NRC, 1983. Risk Assessment in the Federal Government: Managing the Process. Committee on
19 the Institutional Means for Assessments of Risk to Public Health, Commission on Life Sciences,
20 National Research Council. National Academy Press, Washington, DC. ISBN 0-309-03349-7
21
22 NRC, 1993. Pesticides in the Diets of Infants and Children. National Academy Press,
23 Washington DC.
24
25 NRC, 1994. Science and Judgment in Risk Assessment. Committee on Risk Assessment of
26 Hazardous Air Pollutants, Board on Environmental Sciences and Technology, Commission on
27 Life Sciences, National Research Council. National Academy Press, Washington, DC. ISBN 0-
28 309-04894-X
29
30 NRC, 1996. Understanding Risk: Informing Decisions in a Democratic Society. Committee on
31 Risk Characterization, Commission on Behavioral and Social Sciences and Education, National
32 Research Council. National Academy Press, Washington, DC. ISBN 0-309-05396-X
33
34 Osowski, Sharon L., Joseph D. Swick, Jr., Gerald R. Carney, Hector B. Pena, Jeffrey E.
35 Danielson, and David A. Parrish, 2001. A watershed-based cumulative risk impact analysis:
36 Environmental vulnerability and impact criteria. Environmental Monitoring and Assessment
37 66:159-185
38
39 Perry, Robert H., Don W. Green, and James O. Maloney, Eds., 1997. Perry's Chemical
40 Engineers' Handbook. 7th Edn. McGraw Hill Professional Publishing, New York. ISBN 0-
41 07049-841-5
H-78
-------
DRAFT- Risk Assessment Forum Review Draft - August 2, 2001 - Do Not Quote or Cite
1 Portier, Christopher J., 2001. "Virtual Human Concept and its Application to Chemical Mixture
2 Research," presented at Application of Technology to Chemical Mixture Research Conference,
3 January 9-11, 2001, Fort Collins, CO.
4
5 Presidential/Congressional Commission on Risk Assessment and Risk Management, 1997. Risk
6 Assessment and Risk Management in Regulatory Decision-Making. Washington, DC.
7
8 Random House, 1966. The Random House Dictionary of the English Language: The Unabridged
9 Edition. Random House, New York.
10
11 Rouse, P. and J.G. Kidd, 1941. Conditional neoplasms and sub-threshold neoplastic states: A
12 study of the tar tumors of rabbits. J. Exp. Med. 73:3 69-3 90
13
14 Sexton, Ken, David E. Kleffman & Michael A. Callahan, 1995. An introduction to the National
15 Human Exposure Assessment Survey (NHEXAS) and related phase I field studies. J. Expos.
16 Anal. Environ. Epidem. 5, 229-232. Related papers are in the same issue of J. Expos. Anal.
17 Environ. Epidem. 5, 233-444.
18
19 State of Vermont, 1991. "Environment 1991: Risks to Vermont and Vermonters." Agency of
20 Natural Resources, State of Vermont. Waterbury, VT.
21
22 Teuschler, Linda K., Michael L. Dourson, William M. Stiteler, Peter McClure, and Heather Tully,
23 1999. Health risks above the reference dose for multiple chemicals. Regulatory Toxicology and
24 Pharmacology 30: S19- S26.
25
26 USEPA, 1986a. "The Risk Assessment Guidelines of 1986." Risk Assessment Forum, Office of
27 Research and Development, U.S. Environmental Protection Agency. Washington, DC.
28 EPA/600/8-87/045
29
30 USEPA, 1986b. "Guidelines for the Health Risk Assessment of Chemical Mixtures." Risk
31 Assessment Forum, Office of Research and Development, Environmental Protection Agency.
32 Washington, DC. EPA/630/R-98/002
33
34 USEPA, 1986c. "Guidelines for Mutagenicity Risk Assessment." Risk Assessment Forum, Office
35 of Research and Development, Environmental Protection Agency. Washington, DC. EPA/630/R-
36 98/003
37
38 USEPA, 1986d. "Guidelines for Carcinogen Risk Assessment." Risk Assessment Forum, Office of
39 Research and Development, Environmental Protection Agency. Washington, DC. EPA/630/R-
40 98/001
41
H-79
-------
DRAFT- Risk Assessment Forum Review Draft - August 2, 2001 - Do Not Quote or Cite
1 USEPA, 1987. "The Total Exposure Assessment Methodology (TEAM) Study." Office of Acid
2 Deposition, Environmental Monitoring and Quality Assurance, Office of Research and
3 Development, U.S. Environmental Protection Agency. Washington, DC. EPA/600/6-87/002
4
5 USEPA, 1989a. "Risk Assessment Guidance for Superfund, Volume 1: Human Health Evaluation
6 Manual. Office of Emergency and Remedial Response, Office of Solid Waste and Emergency
7 Response, U.S. Environmental Protection Agency. Washington, DC. EPA/540/1-89/002
8
9 USEPA, 1989b. "Interim Procedures for Estimating Risks Associated with Exposure to
10 Chlorinated dibenzo-p-dioxins and -dibenzofurans (CDDs and CDFs) and 1989 update. Risk
11 Assessment Forum, Office of Research and Development, U.S. Environmental Protection Agency.
12 Washington, DC. EPA 625/3-89/016
13
14 USEPA, 1990a. "Technical Support Document on Health Assessment of Chemical Mixtures."
15 Office of Research and Development, U.S. Environmental Protection Agency. Washington, DC.
16 EPA/600/8-90/064
17
18 USEPA, 1990b. "Reducing Risk: Setting Priorities and Strategies for Environmental Protection."
19 Science Advisory Board, U.S. Environmental Protection Agency. Washington, DC. SAB-EC-90-
20 021
21
22 USEPA, 1991a. "Locational Data Policy." U.S. Environmental Protection Agency. Washington,
23 DC. IRM Policy Manual 2100-CGH2, dated April 8, 1991
24
25 USEPA, 1991b. "Guidelines for Developmental Toxicity Risk Assessment." Risk Assessment
26 Forum, Office of Research and Development, U.S. Environmental Protection Agency.
27 Washington, DC. EPA/600/FR-91/001
28
29 USEPA, 1992a. "Safeguarding the Future: Credible Science, Credible Decisions. A Report of the
30 Expert Panel on the Role of Science at EPA." Washington, DC. EPA/600/9-91/050
31
32 USEPA, 1992b. "Framework for Ecological Risk Assessment." Risk Assessment Forum, Office
33 of Research and Development, U.S. Environmental Protection Agency. Washington, DC.
34 EPA/630/R-92/001
35
36 USEPA, 1992c. "Guidelines for Exposure Assessment." Risk Assessment Forum, Office of
37 Research and Development, U.S. Environmental Protection Agency. Washington, DC.
38 EPA/600/Z-92/001
39
40
41 USEPA, 1993a. "Guidance on the Application of Refined Dispersion Models for
H-80
-------
DRAFT- Risk Assessment Forum Review Draft - August 2, 2001 - Do Not Quote or Cite
1 Hazardous/Toxic Air Release." Office of Air Quality Planning and Standards, Office of Air and
2 Radiation, U.S. Environmental Protection Agency, Research Triangle Park, NC. EPA 454/R-93-
3 002
4
5 USEPA, 1993b. "A Guidebook to Comparing Risks and Setting Environmental Priorities." Office
6 of Policy, Planning and Evaluation, U.S. Environmental Protection Agency. Washington, DC.
7 EPA 230-B-93-003
8
9 USEPA, 1993c. "Chemical Indexing System for the Toxic Chemical Release Inventory, Part I:
10 Chronic Index." Air, Radiation and Toxics Division, EPA Region III, U.S. Environmental
11 Protection Agency. Philadelphia, PA. EPA/903/R-93/002
12
13 USEPA, 1994. "Methods for Derivation of Inhalation Reference Concentrations and Application
14 of Inhalation Dosimetry." Office of Health and Environmental Assessment, Office of Research
15 and Development, U.S. Environmental Protection Agency, Research Triangle Park, NC.
16 EPA/600/8-90/066F
17
18 USEPA, 1995a. "Policy for Risk Characterization." Memorandum from U.S. Environmental
19 Protection Agency Administrator Carol M. Browner, dated March 21, 1995. Washington, DC.
20
21 USEPA, 1995b. "Guidance for Risk Characterization." Policy paper dated February, 1995.
22 Science Policy Council, U.S. Environmental Protection Agency. Washington, DC.
23
24 USEPA, 1995c. "Chemical Indexing System for the Toxic Chemical Release Inventory, Part I:
25 Chronic Index; Addendum." Air, Radiation and Toxics Division, EPA Region III, U.S.
26 Environmental Protection Agency. Philadelphia, PA. EPA/903/R-93/002a (August, 1995)
27
28 USEPA, 1995d. "Compilation of Air Pollutant Emission Factors, Volume I: Stationary Point and
29 Area Sources." Fifth Edition. Office of Air Quality Planning and Standards, Office of Air and
30 Radiation, U.S. Environmental Protection Agency. Research Triangle Park, NC. EPA AP-42
31
32 USEPA, 1996a. "RAGS Reform Stakeholder Forums: Synopsis of Participants' Comments. San
33 Francisco, California, October 30-November, 1, 1996, and Washington, DC, November 6-
34 November 8, 1996." Office of Emergency and Remedial Response, U.S. Environmental
35 Protection Agency, Washington, DC.
36
37 USEPA, 1996b. "Guidelines for Reproductive Toxicity Risk Assessment." Risk Assessment
38 Forum, Office of Research and Development, U.S. Environmental Protection Agency.
39 Washington, DC. EPA/63O/R-96/009
40
41 USEPA, 1996c. "Proposed Guidelines for Carcinogen Risk Assessment." Risk Assessment
H-81
-------
DRAFT- Risk Assessment Forum Review Draft - August 2, 2001 - Do Not Quote or Cite
1 Forum, Office of Research and Development, U.S. Environmental Protection Agency.
2 Washington, DC. EPA/600/P-92/003C
3
4 USEPA, 1996d. "Compilation of Air Pollutant Emission Factors, Volume I: Stationary Point and
5 Area Sources." Fifth Edition, Supplements A & B. Office of Air Quality Planning and Standards,
6 Office of Air and Radiation, U.S. Environmental Protection Agency. Research Triangle Park, NC.
7 EPA AP-42
8
9 USEPA, 1997a. "Guidance on Cumulative Risk Assessment, Part 1. Planning and Scoping."
10 Science Policy Council, U.S. Environmental Protection Agency, Washington, DC. Attachment to
11 memo dated July 3, 1997 from the Administrator, Carol Browner, and Deputy Administrator,
12 Fred Hansen, titled "Cumulative Risk Assessment Guidance-Phase I Planning and Scoping."
13
14 USEPA, 1997b. "Chemical Indexing System, Part II: Vulnerability Index." Waste and Chemicals
15 Management Division, EPA Region III, U.S. Environmental Protection Agency. Philadelphia, PA.
16 EPA/903/R-97/021
17
18 USEPA, 1997c. "Guiding Principles for Monte Carlo Analysis." Risk Assessment Forum, Office
19 of Research and Development, Washington, DC. EPA/630/R-97/001
20
21 USEPA, 1997d. "Compilation of Air Pollutant Emission Factors, Volume I: Stationary Point and
22 Area Sources." Fifth Edition, Supplement C. Office of Air Quality Planning and Standards, Office
23 of Air and Radiation, U.S. Environmental Protection Agency. Research Triangle Park, NC. EPA
24 AP-42
25
26 USEPA, 1998a. "Risk Assessment Guidance for Superfund (RAGS) Stakeholder Forum:
27 Synopsis of Participants' Comments. Atlanta, Georgia, March 2-4, 1998." Office of Emergency
28 and Remedial Response, U.S. Environmental Protection Agency, Washington, DC.
29
30 USEPA, 1998b. "Guidelines for Ecological Risk Assessment." Risk Assessment Forum, Office of
31 Research and Development, Washington, DC. EPA/630/R-95/002F
32
33 USEPA, 1998c. "General Guidance for Risk Management Programs (40 CFR Part 68) " Chemical
34 Emergency Preparedness and Prevention Office, Office of Solid Waste and Emergency Response,
35 U.S. Environmental Protection Agency, Washington, DC. EPA 550-B-98-003
36
37 USEPA, 1998d. "Methods for Exposure-Response Analysis for Acute Inhalation Exposure to
38 Chemicals: Development of the Acute Reference Exposure." Review Draft. Office of Research
39 and Development, U.S. Environmental Protection Agency. Washington, DC. EPA/600/R-98/051
40
41 USEPA, 1998e. "Guidelines for Neurotoxicity Risk Assessment." Risk Assessment Forum, Office
H-82
-------
DRAFT- Risk Assessment Forum Review Draft - August 2, 2001 - Do Not Quote or Cite
1 of Research and Development, U.S. Environmental Protection Agency. Washington, DC.
2 EPA/630/R-95/001F
3
4 USEPA, 1998f. "Comparative Risk Framework: Methodology and Case Study." SAB Review
5 Draft dated November 9, 1998. National Center for Environmental Assessment, Office of
6 Research and Development, U.S. Environmental Protection Agency. Cincinnati, OH.
7
8 USEPA, 1998g. "Handbook for Air Toxics Emission Inventory Development, Volume I:
9 Stationary Sources." Office of Air Quality Planning and Standards, Office of Air and Radiation,
10 U.S. Environmental Protection Agency. Research Triangle Park, NC. EPA-454/B-98-002
11
12 USEPA, 1998h. "Human Health Risk Assessment Protocol for Hazardous Waste Combustion
13 Facilities." Peer Review Draft. Office of Solid Waste and Emergency Response, U.S.
14 Environmental Protection Agency. Washington, DC. EPA530-D-98-001 Internet:
15 www.epa.gov/epaoswer/hazwaste/combust/riskvol.htm
16
17 USEPA, 1998L "Report of the Common Sense Initiative Council's Stakeholder Involvement
18 Work Group." Common Sense Initiative Council, U.S. Environmental Protection Agency.
19 Washington, DC.
20
21 USEPA, 1998J. "An SAB Report: Review of Disproportionate Impact Methodologies." Science
22 Advisory Board, U.S. Environmental Protection Agency. Washington, DC. EPA-SAB-IHEC-99-
23 007
24
25 USEPA, 1999a. "Guideline on Air Quality Models." Office of Air Quality Planning and
26 Standards, Office of Air and Radiation, U.S. Environmental Protection Agency. Research
27 Triangle Park, NC. 40CFR Ch. I (7-1-99 Edition), Appendix W to Part 51, Pages 390-481
28
29 USEPA, 1999b. "EPA's Framework for Community-Based Environmental Protection." Office of
30 Policy/Office of Reinvention, U.S. Environmental Protection Agency, Washington, DC. EPA 237-
31 K-00-001
32
33 USEPA, 1999c. "Risk Assessment Guidance for Superfund: Volume 1 - Human Health
34 Evaluation Manual. Supplement to Part A: Community Involvement in Superfund Risk
35 Assessments." Office of Solid Waste and Emergency Response, United States Environmental
36 Protection Agency. Washington, DC. EPA 540-R-98-042/PB99-963303
37
38 USEPA, 1999d. "Risk Management Program Guidance for Offsite Consequence Analysis."
39 Chemical Emergency Preparedness and Prevention Office, Office of Solid Waste and Emergency
40 Response, U.S. Environmental Protection Agency, Washington, DC. EPA 550-B-99-009
41
H-83
-------
DRAFT- Risk Assessment Forum Review Draft - August 2, 2001 - Do Not Quote or Cite
1 USEPA, 1999e. "Review of Revised Sections of the Proposed Guidelines for Carcinogen Risk
2 Assessment." Science Advisory Board, U.S. Environmental Protection Agency. Washington, DC.
3 EPA-SAB-EC-99-015
4
5 USEPA, 1999f "An SAB Report on the National Center for Environmental Assessment's
6 Comparative Risk Framework Methodology." Science Advisory Board, U.S. Environmental
7 Protection Agency. Washington, DC. EPA-SAB-DWC-99-016
8
9 USEPA, 1999g. "Residual Risk Report to Congress." Office of Air Quality Planning and
10 Standards, Office of Air and Radiation, U.S. Environmental Protection Agency. Research
11 Triangle Park, NC. EPA-453/R-99-001
12
13 USEPA, 1999h. "Risk-Screening Environmental Indicators: 1988-1997 TRI Data 'Air-Only'
14 Model." Office of Pollution Prevention and Toxic Substances, Office of Prevention, Pesticides,
15 and Toxic Substances, U.S. Environmental Protection Agency. Washington, DC. CD-ROM
16 Version 1.0, dated July 6, 1999. (Version 2.0 is now in beta-testing.)
17
18 USEPA, 1999L "Handbook for Criteria Pollutant Inventory Development: A Beginner's Guide
19 for Point and Area Sources." Office of Air Quality Planning and Standards, Office of Air and
20 Radiation, U.S. Environmental Protection Agency. Research Triangle Park, NC. EPA-454/R-99-
21 037
22
23 USEPA, 1999J. "Total Risk Integrated Methodology, TREVI.Expo.: Technical Support
24 Document." External Review Draft. Office of Air Quality Planning and Standards, Office of Air
25 and Radiation, U.S. Environmental Protection Agency. Research Triangle Park, NC. EPA-453/D-
26 99-001
27
28 USEPA, 2000a. "Toward Integrated Environmental Decision-Making." Science Advisory Board,
29 U.S. Environmental Protection Agency. Washington, DC. EPA-SAB-EC-00-011
30
31 USEPA, 2000b. "Science Policy Council Handbook: Peer Review." 2nd Edition. Science Policy
32 Council, U.S. Environmental Protection Agency. Washington, DC. EPA 100-B-OO-OOl
33
34 USEPA, 2000c. "Science Policy Council Handbook: Risk Characterization." Science Policy
35 Council, U.S. Environmental Protection Agency. Washington, DC. EPA 100-B-00-002
36
37 USEPA, 2000d. "Handbook for Non-Cancer Health Effects Valuation." Non-Cancer Health
38 Effects Valuation Subcommittee of the EPA Social Science Discussion Group, Science Policy
39 Council, U.S. Environmental Protection Agency. Washington, DC. Dated November, 2000.
40
41
H-84
-------
DRAFT- Risk Assessment Forum Review Draft - August 2, 2001 - Do Not Quote or Cite
1 USEPA, 2000e. "Guide to Field Storage of Biosolids, Appendix A: Odor Characterization,
2 Assessment and Sampling." Office of Wastewater Management, Office of Water, U.S.
3 Environmental Protection Agency. Washington, DC. EPA/832-B-00-007 Internet:
4 www.epa.gov/owm/bio/fsguide/
5
6 USEPA, 2000f. "AP-42: Compilation of Air Pollutant Emission Factors, Volume II: Mobile
7 Sources." Office of Transportation and Air Quality, Office of Air and Radiation. Washington, DC.
8 EPA AP-42, Volume II Internet: www.epa.gov/otaq/ap42.htm
9
10 USEPA, 2000g. "Catalog of Hazardous and Solid Waste Publications." 13th Edition. Office of
11 Solid Waste and Emergency Response, U.S. Environmental Protection Agency. Washington, DC.
12 EPA530-B-00-001 Internet: www.epa.gov/epaoswer/osw/catalog.htm
13
14 USEPA, 2000h. "Benchmark Dose Technical Guidance Document" Draft report. Risk
15 Assessment Forum, Office of Research and Development, U.S. Environmental Protection Agency.
16 Washington, DC. EPA/630/R-00/001
17
18 USEPA, 2000i. "Proposed Guidance on Cumulative Risk Assessment of Pesticide Chemicals that
19 Have a Common Mechanism of Toxicity." Public Comment Draft. Office of Pesticide Programs,
20 Office of Prevention, Pesticides, and Toxic Substances, U.S. Environmental Protection Agency.
21 Washington, DC. Internet: www.epa.gov/fedrgstr/EPA-PEST/2000/June/Day-30/6049.pdf
22
23 USEPA, 2001a. "Supplementary Guidance for Conducting Health Risk Assessment of Chemical
24 Mixtures." Risk Assessment Forum, Office of Research and Development, U.S. Environmental
25 Protection Agency. Washington, DC. EPA/630/R-00/002
26
27 USEPA, 2001b. "Stakeholder Involvement & Public Participation at the U.S. EPA: Lessons
28 Learned, Barriers, & Innovative Approaches." Office of Policy, Economics and Innovation, U.S.
29 Environmental Protection Agency. Washington, DC. EPA-100-R-00-040.
30
31 USEPA, 200Ic. "Regional Air Impact Modeling Initiative: Pilot Study - Initial Phase." Draft.
32 Office of Solid Waste, Office of Solid Waste and Emergency Response, U.S. Environmental
33 Protection Agency. Washington, DC. EPA-906-R-01-001
34
35 USEPA, 200Id. Personal communication. Debbie Sisco,. Biological and Economic Analysis
36 Division, Office of Pesticide Programs, Office of Prevention, Pesticides, and Toxic Substances,
37 U.S. Environmental Protection Agency. Washington, DC. July 31, 2001
38
39
40
41 USEPA, 2001e. Personal communication. TSCA Assistance Office Hotline, Office of Prevention,
H-85
-------
DRAFT- Risk Assessment Forum Review Draft - August 2, 2001 - Do Not Quote or Cite
1 Pesticides, and Toxic Substances, U.S. Environmental Protection Agency. Washington, DC. July
2 31,2001
3
4 Yang, Raymond S.H., 2001. "Application of Computer Modeling to Simple or Complex
5 Mixtures," presented at Application of Technology to Chemical Mixture Research Conference,
6 January 9-11, 2001, Fort Collins, CO.
7
8 Zartarian, Valerie G., Haluk Ozkaynak, Janet M. Burke, Maria J. Zufall, Marc L. Rigas, and
9 Edwin J. Furtaw, Jr, 2000. "A Modeling Framework for Estimating Children's Residential
10 Exposure and Dose to Chlorpyrifos via Dermal Residue Contact and Non-Dietary Ingestion."
11 Environmental Health Perspectives 108:505-514
12
13
14
15
H-86
-------
DRAFT- Risk Assessment Forum Review Draft - August 2, 2001 - Do Not Quote or Cite
1 APPENDIX A: A RESOURCE LIST FOR METHODS RELEVANT TO EXPOSURE
2 ASSESSMENT
3
4 [Note to reviewer: Appendix A is incomplete and is actively being developed]
5
6 The following is a brief discussion of where to find some of the methods for assessing
7 exposures to specific sources and stressors. This is not meant to be an exhaustive list, but is
8 provided to assist the assessor in finding recognized methods for dealing with certain parts of an
9 assessment. This list is a starting place for assessors, not a comprehensive guide to risk
10 assessment. It is not envisioned that all cumulative risk assessments will need methods for
11 assessing all of these sources, stressors, and pathways. Furthermore, the specific methods
12 mentioned below may not be adequate for some cumulative risk assessment situations. Finally,
13 new methods are constantly being developed as the state of science progresses; it is the
14 responsibility of the assessor to determine the currency and applicability of methods used for a
15 given assessment.
16
17 A.I. Resources Relevant to Chemical Exposures
18
19 General guidelines: [to be completed]
20
21 Air-related sources and activities: The methods for evaluating air-related exposures
22 generally start with compiling an emissions inventory, then air modeling augmented by monitoring
23 data. EPA's Clearinghouse for Inventories and Emission Factors (CHIEF) website
24 (www.epa.gov/ttn/chief/) is an excellent starting place that has many of the relevant documents on
25 methods and data for constructing emissions inventories available for download. These include
26 Handbook for Criteria Pollutant Inventory Development: A Beginner's Guide for Point and
27 Area Sources (USEPA, 1999i), Handbook for Air Toxics Emission Inventory Development,
28 Volume I: Stationary Sources (USEPA, 1998g),and Compilation of Air Pollutant Emission
29 Factors (for both stationary and mobile sources) (USEPA, 1995d, 1996d, 1997d, 2000f), as well
30 as many other documents and software. Likewise, the Support Center for Regulatory Air Models
31 (SCRAM) site (www.epa.gov/ttn/scram/) provides extensive information on the models discussed
32 in Guideline on Air Quality Models (USEPA, 1999a), including downloadable software and users
33 guides for many of the models. The Ambient Monitoring Technology Information Center
34 (AMTIC) site (www.epa.gov/ttn/amtic/) contains information on monitoring programs,
35 monitoring methods, and other monitoring-related information. The umbrella website for all three
36 of the above is the Technology Transfer Network (www.epa. gov/ttn/). which also has other
37 useful information and links in addition to the above.
38
39 Water-related sources and activities: [to be completed]
40
41 Sources to land, and waste-related activities: The EPA Office of Solid Waste and
H-87
-------
DRAFT- Risk Assessment Forum Review Draft - August 2, 2001 - Do Not Quote or Cite
1 Emergency Response has published an extensive catalog summarizing their publications (USEPA,
2 2000g). They have also published a "peer review draft" document called Human Health Risk
3 Assessment Protocol for Hazardous Waste Combustion Facilities (USEPA, 1998h) which deals
4 with how to assess risks from hazardous waste incinerators. These reports are available on-line.
5
6 Chemical accidents, transportation-related spills: In a population-focused assessment
7 such as a community-based cumulative risk assessment, the threat or risk from chemical accidents
8 may be an important factor in the assessment. Spills or other transportation-related accidental
9 releases of materials could cause very severe short term pollution episodes and could contribute
10 to longer term pollution. In addition, the increased likelihood of vehicular accidents could directly
11 affect local residents. Appendix B describes the kinds of analyses conducted to determine the
12 degree of human exposure (both to workers and the general public) associated with accidental
13 releases of chemicals, with appropriate references.
14
15
16 A.2. Resources Relevant to Exposures to Non-Chemical Stressors
17
18 Biological stressors: [to be added]
19
20 Radiological stressors: [to be added]
21
22 Noise, vibration, and congestion: Increases in noise levels, e.g., from truck and/or rail
23 traffic, could result in increased stress to local residents, as could the additional traffic congestion.
24 Increased vibrations from additional truck or rail traffic could also increase or accelerate damage
25 to local roads and other structures such as residences (foundation cracks), water and sewer lines,
26 etc.. These types of damage could result in additional costs and stress to the local population.
27 The U.S. Department of Housing and Urban Development has issued The Noise Guidebook
28 (HUD, 1991), which implements the existing noise regulations [24 CFR 51-B] and includes the
29 HUD Noise Assessment Guidelines. (The Guidebook is available in hard copy only.) The Federal
30 Railroad Administration has developed a manual called High-Speed Ground Transportation Noise
31 and Vibration Impact Assessment (DOT, 1998) which provides the theory, equations, and
32 applications of noise and vibration analysis for high-speed railroads. Much of the theory and
33 information is also applicable to other noise and vibration problems. Appendix A of the DOT
34 Guide is a general discussion of noise concepts, with references. The Guide is available on-line.
35
36 Odor: EPA's Office of Wastewater Management has issued a report called Guide to Field
1>1 Storage ofBiosolids (USEPA, 2000e) which contains an appendix on "Odor Characterization,
38 Assessment, and Sampling." Odor assessment is an analytic-deliberative process, involving both
39 science-based analytical methods and more subjective analysis. The appendix of the Guide
40 discusses sensory characterization of odors (character, intensity, pervasiveness, quantity), some
41 practical options for assessing odors in a community, and the chemistry of odors (including range
H-88
-------
DRAFT- Risk Assessment Forum Review Draft - August 2, 2001 - Do Not Quote or Cite
1 of odor thresholds). It also discusses odor sample collection and analysis, and has several dozen
2 references for further information. This report is available on-line.
3
4 Other non-chemical stressors: [to be added]
5
6
7 APPENDIX B: ASSESSING ACCIDENTAL CHEMICAL RELEASE EXPOSURE
8
9 There are several steps in assessing an accidental chemical release exposure. The typical
10 analytical steps in an overall accidental chemical release risk assessment are process analysis,
11 likelihood or frequency of accidents, source term modeling, dispersion or consequence modeling,
12 and the exposure assessment.. A brief description of each step is provided below. Each of these
13 steps can be evaluated quantitatively or qualitatively.
14
15 The process analysis is a formal, systematic analysis of the process where a chemical is
16 handled to determine the probabilities and consequences of acute, catastrophic failures of
17 engineered systems leading to an accidental release of the chemical. This analysis is often called a
18 Process Hazards Analysis (PHA). For example, if a process temperature control fails, allowing
19 pressure to build in a reactor system, an emergency pressure relief valve may open, venting
20 chemical to the atmosphere. More severe hypothetical scenarios are often also evaluated, such as
21 the failure of a storage tank, leading to a massive spill. Several formal PHA evaluation techniques
22 are available including "What-If," "Failure Mode and Effect Analysis," "Event-Tree", or "Fault-
23 Tree." (USEPA 1998c, AIChE, 1992)
24
25 The likelihood or frequency of accidents step is an evaluation of each of the scenarios
26 uncovered in the process analysis step for likelihood or frequency of occurrence. For example,
27 equipment failure rate data or accident histories can be used to judge how frequently certain
28 accidental releases might occur.
29
30 Source term modeling, which estimates the amount or rate of release in case of accident, is
31 performed once the failure scenarios are determined. A wide variety of published calculation
32 methods or models are available (USEPA 1998c, USEPA 1999d) to determine the source terms
33 for an accidental chemical release.
34
35 Dispersion or consequence modeling is performed once the source terms (rate and
36 duration of the release) are known. In this step, the consequences associated with those predicted
37 releases can be evaluated. If the chemical released is a gas and toxic by inhalation, the
38 consequence assessment would involve an analysis of the downwind dispersion, or the distance
39 the chemical will travel downwind to a particular toxic concentration. If the chemical is
40 flammable, the consequences of an explosion or fire might be analyzed. A wide variety of
41 dispersion and consequence modeling tools, ranging from simple screening models to
H-89
-------
DRAFT- Risk Assessment Forum Review Draft - August 2, 2001 - Do Not Quote or Cite
1 sophisticated and complex computer applications, are available for this step (USEPA 1999d,
2 AIChE 1996, USEPA 1993a). In addition to the source terms generated above, several other
3 data elements are needed, such as physical/chemical properties (e.g., whether the vapor cloud is
4 heavier than air or water reactive), meteorological conditions (e.g., wind speed and direction,
5 temperature, humidity), and terrain surrounding the facility (e.g., buildings or valleys that may
6 channel or disperse a vapor cloud). Physical/chemical properties can be found in chemical
7 reference texts such as Kirk-Othmer 's Encyclopedia of Chemical Technology (Kroschwitz and
8 Howe-Grant, 1994), Perry's Chemical Engineers' Handbook (Perry, et al., 1997), on Material
9 Safety Data Sheets (MSDS)17, or in the Guidance for Offsite Consequence Analysis (USEPA
10 1999d). Meteorological conditions are often collected on-site or at local airports; information
11 about terrain can be collected from topological maps or by visual inspection. Guidance on all
12 these parameters is available in USEPA 1999d.
13
14 The final step in a chemical accident exposure analysis is the exposure assessment. The
15 exposure assessment is related to, and builds from, the dispersion or consequence modeling step.
16 The dispersion or consequence modeling depends on a health endpoint and the exposure level
17 related to that endpoint. The endpoint reflects the health effect of concern; e.g., if lethality
18 resulting from an acute toxic exposure is the concern, then the endpoint would be the airborne
19 concentration necessary to cause acute lethality. Besides lethality, concentrations for certain
20 health effects (e.g., odor thresholds, eye irritation) are available for several common toxic
21 substances (NIOSH 1997, ACGffl 1998, AfflA 2000). These established concentrations,
22 however, may be based on toxicity studies that are weak, derived by consensus, or may not be the
23 most representative of actual exposure effects. Further, little is known about the chronic or long-
24 term effects associated with an acute, non-lethal accidental exposure, so most assessments of the
25 risk from chemical accidents focus on acute or short-term effects. Work is currently underway to
26 develop more appropriate emergency exposure concentrations for a number of common toxic
27 substances.
28
29 Other factors that play a role in this type of assessment, and in the dispersion analysis step
30 above, include an evaluation of the duration of exposure, the average or range of chemical
31 concentration in the cloud, at what portion of a cloud an individual might be exposed, the
32 likelihood that people are present, and whether they are indoors and indirectly exposed. Often,
33 determination of the actual human exposure dose in a vapor cloud is a complicated exercise;
34 typical consequence analyses only identify the distance downwind for the plume to reach a
35 particular concentration without consideration of actual human exposure, since it is the short-
36 term, threshold type effects that are being evaluated.
37
38 The results of all of these steps can be combined to generate a number of measures of risk
There are many searchable MSDS data bases on-line that can be located with most search engines.
H-90
-------
DRAFT- Risk Assessment Forum Review Draft - August 2, 2001 - Do Not Quote or Cite
1 associated with accidental releases. For example, individual risk profiles can be generated to
2 measure the acute risk as a function of distance and direction from a chemical source. Or a
3 societal risk can be generated to determine the cumulative probability or frequency of events that
4 cause fatalities, injuries, or exposures over time.
5
6
7 APPENDIX C. DATA QUALITY ISSUES IN MONITORING AND OTHER
8 EXPOSURE-RELATED DATA
9
10 There are a number of separate and important issues associated with input data quality
11 when doing a cumulative assessment. Three of these issues are: 1) data quality needed for the
12 assessment, based on how the data will be used; 2) the relative quality of available data from
13 various sources; and 3) combining data of different quality in a single assessment.
14
15 The Data Quality Needed for the Assessment. The level of data quality necessary for a
16 individual assessment is an issue that cannot be overlooked. The level of certainty needed for the
17 decision to be made relates directly to using appropriate data and analytical techniques for
18 assessments. For a cumulative risk assessment, this means that the type of assessment - and
19 therefore the level of certainty required - should be determined before beginning the assessment.
20
21 From the planning and problem formulation phase of the assessment, the type of
22 assessment and depth of the assessment (i.e., screening, in-depth, etc.) should be known.
23 Depending on the type and depth of assessment, the nature of the analytical tools used, and the
24 quality and breath of the input data needed, can be quite different.
25
26 The Quality of Existing Data. Often when doing an assessment, it is difficult or impossible
27 to collect new data due to time, financial, or other constraints, and the assessor must depend upon
28 existing data bases for analysis purposes. Appendix D gives some considerations about the
29 quality of the data found in frequently used data bases.
30
31 Combining Data of Different Quality. The assessor will encounter, and most likely use,
32 data of differing quality when doing cumulative risk assessments. This raises the concern that the
33 value and benefit of high quality data might be lost if combined with lower quality data. The
34 Office of Pesticide Programs asked its FIFRA Scientific Advisory Panel (SAP) in 1999 how this
35 issue should be addressed. The SAP (FIFRA SAP, 2000) recommended the following approach
36 for cumulative risk assessment for pesticides:
37
38 • Clearly document the quality of the data and input parameters used in the risk analysis.
39 Quality thresholds could be established for data use. Monitoring data, properly accounted
40 for measurement errors, are preferred over screening level inputs.
41
H-91
-------
DRAFT- Risk Assessment Forum Review Draft - August 2, 2001 - Do Not Quote or Cite
1 • Focus on individual-based analysis to ensure capturing the high exposure and sensitive
2 individuals and account for cross-media transfer and "para-occupational" exposures.
3
4 • Cumulative risk analysis should retain the resolution of geographic, temporal, and
5 demographic variations while maintaining optimal data usage with respect to the
6 increasing uncertainties associated with lowering of sample size.
7
8 • Systematically conduct quantitative sensitivity and uncertainty analysis for both the
9 exposure and the toxicity.
10
11 • Uncertainties in data can be reduced or better characterized by (1) comparing sets of
12 similar data collected from different years and locations, (2) comparing results from
13 screening level analyses with more refined analyses from data-rich cases for selected
14 chemicals and pathways, and (3) maintaining the association between the pathways.
15
16 • Develop the process for reassessment as new quality data become available.
17
18
19
20 APPENDIX D: SOME THOUGHTS ON QUALITY OF DATA IN VARIOUS WIDELY-
21 USED DATA BASES
22
23 The following paragraphs contain some considerations when using data in commonly-used
24 databases. Over the past two decades, data in environmentally-related databases has improved,
25 but it is far from perfect. Some of the issues with databases discussed below may improve, even
26 in the short term. The paragraphs below are meant to cause an assessor to think about specific
27 aspects of the data being used for an assessment, and weigh the uncertainty involved in using
28 those data.
29
30 First, there are important limitations with respect to characterization of hazardous releases
31 to the environment. Point source release data may be based on actual measurements (e.g., Permit
32 Compliance System [PCS] data) or estimates (e.g., Toxics Release Inventory [TRI] data) that can
33 be inaccurate by a factor of 2, 5, 10 or even more. The availability and quality of permit data will
34 vary geographically. Data on a significant percentage of permitted discharges may be unavailable
35 or of poor quality due to insufficient monitoring of releases. Finding accurate non-point source
36 release information of hazardous chemicals is especially problematic. The combination of these
37 problems, along with the possible existence of non-permitted discharges, can make the
38 quantitative assessment of risk a difficult task. In addition, many of the risks to the environment
39 may not be tracked in databases (USEPA 1990b). These include motor vehicle emissions, non-
40 TRI point sources, area sources (such as gas stations or dry cleaners), consumer product use,
41 pesticide use, and others.
H-92
-------
DRAFT- Risk Assessment Forum Review Draft - August 2, 2001 - Do Not Quote or Cite
1 Second, for many types of analyses, the sources of pollution or the adverse environmental
2 conditions that are to be evaluated must be assigned to specific geographic locations. A number
3 of alternatives can be used to locate sources. The locations of monitored facilities and
4 remediation sites are collected by the EPA, by state and local environmental agencies, and their
5 contractors, and are available through many systems. The use of self-reported location data by
6 regulated operators has been a common method used to acquire geographic coordinates for sites,
7 facilities, areas, and regions of environmental concern, but analysis of the location data in EPA's
8 data bases has shown sufficient inaccuracy to require the issuance of a Locational Data Policy
9 (LDP) (US EPA, 199la). The LDP mandates preferred location data collection methodologies, as
10 well as defines accuracy and verification procedures, and the reporting of location data for
11 regulated entities. Superfund sites and other facilities that encompass large areas typically do not
12 have a single point of toxic material release (e.g., a single smokestack), and their recorded
13 location coordinates may represent the administrative location of the facility (front gate, property
14 centroid, or other office locations not even at the same site), not at the point where the pollution
15 is occurring. Most uses oflocational data for assessments employing geographically-based
16 models will require verification of reported data. At a local level, this is less difficult than when
17 doing regional or national assessments, but it can still be time consuming (and difficult if complex
18 facilities choose not to cooperate).
19
20 Third, most existing non-GIS-based program systems cannot easily accommodate
21 irregularly shaped area features, nor offer a complete set of documentation on the accuracy of the
22 data already collected. The diversity of EPA's programmatic database systems in terms of their
23 design and implementation makes it technically difficult and expensive to integrate location (and
24 associated attribute) data across program (multi-media) lines. Also, much of the location data are
25 collected independently by federal, state and local agencies, and according to different criteria and
26 methods, and can be held in either hard-copy or electronic forms, or both, in a variety of
27 locations.
28
29 Fourth, when using TRI data for cumulative risk assessments, it is important to recognize
30 that the TRI data base only contains data on larger facilities (both in terms of number of
31 employees and amount of materials involved), on a limited number of chemicals, and on specific
32 manufacturing sectors. Additional sources of release data may therefore be required for a more
33 complete assessment of risk.
34
35 Fifth, when using AIRS Facility Subsystem (AFS) release data for risk assessment, the
36 assessor must be aware that most AFS facilities prepare emissions inventories only once every five
37 years. It is therefore possible that the emissions data recorded in the AFS are somewhat out of
38 date. Also noteworthy is that release information is generally available for only five criteria air
39 pollutants: SO2, NO2, CO, O3, PM-10, and Pb. Release estimates can be made for many other
40 toxic chemicals using a model available from the Office of Air and Radiation (USEPA 1999a).
41 Use of this model, like other models, provides additional information to the analyst, but also
H-93
-------
DRAFT- Risk Assessment Forum Review Draft - August 2, 2001 - Do Not Quote or Cite
1 introduces greater uncertainty to the analysis being performed.
2
3 Finally, the Permit Compliance System (PCS) distinguishes between major discharges and
4 minor discharges based on potential threat to health and to the environment, but most often that
5 differentiation was made solely on the basis of relative volume of water discharged and not on the
6 amount or nature of the toxic chemicals contained in the discharge. Only discharge information
7 from major facilities are required to be entered into the PCS, and so minor facilities are under-
8 represented. Some PCS records only indicate the corporate address rather than give information
9 on the actual location of the toxic material release point, and some only show the location of the
10 principal facility and not, when they exist, secondary facilities.
11
12
13 APPENDIX E: SOME THOUGHTS ON BACKGROUND EXPOSURES
14
15 When looking at aggregate exposures or cumulative risks of citizens, so-called
16 "background exposures" to specific chemicals are no less "real" exposures than the pollution
17 usually studied for regulatory purposes. Whereas in historical single-chemical assessments done
18 for limiting pollution, background sources of the chemical were often irrelevant to the questions
19 being asked of the assessment (or ignored as having negligible effect on risk), background sources
20 are rarely irrelevant with cumulative risk assessments18.
21
22 Background concentrations can be categorized as either naturally-occurring, that is,
23 chemicals which are naturally present in the environment before it was influenced by humans, or
24 anthropogenic, that is, present in the environment due to historical human-made sources.
25 Naturally-occurring background chemicals may be either localized or ubiquitous. Anthropogenic
26 background sources can be either localized from a point source, or generalized from unidentified
27 sources or non-point sources.
28
29 Assessments of morbidity incidence and death rates, market basket surveys, and pesticide
30 residue surveys also provide information which can be reflective of background chemical
31 concentrations as well as overt pollution. Background issues extend across all media, beyond
1 8
The word "background" is often used in to describe exposures to chemicals or other stressors that derive from
sources other than the sources being assessed. For example, in the Agency's assessment of residual risk associated with
hazardous air pollutant emissions from particular categories of sources that remains after the implementation of technology-
based controls, "background" is defined as all hazardous air pollutant exposures (via inhalation or other routes) not associated
with the source(s) being assessed. At a Superfund site, "background contamination" refers to contamination that is not related
to the site release of chemicals, as defined by Comprehensive, Environmental Response, Compensation and Liability Act
(CERCLA).[P.L. 96-510, December 11, 1980, as amended by P.L. 98-802, August 23, 1983, and P.L. 99-499, October 17,
1986] Such focusing or segregation in a risk assessment can be useful to decisions involving pollution sources covered by
particular statutory authorities, but it is typical of a chemically-focused assessment rather than a population-focused assessment
such as a cumulative risk assessment.
H-94
-------
DRAFT- Risk Assessment Forum Review Draft - August 2, 2001 - Do Not Quote or Cite
1 regulated sources, and beyond direct exposure. Many chemicals are naturally present in the
2 environment (e.g., soils, water, vegetation and other biota) and are consequently part of dietary,
3 dermal and inhalation exposures. In some cases, naturally-occurring substances may occur at
4 levels that exceed health-based or risk-based regulatory standards (e.g., drinking water standards),
5 or other levels established to protect human health and the environment. Since cumulative risk
6 assessments are population based, exposures due to naturally-occurring background
7 concentrations should usually be of importance.
8
9 There are several important issues related to natural or anthropogenic background
10 concentrations in cumulative risk assessment. First, if the risks posed by "background"
11 concentrations of certain chemicals are significant (and some may approach or exceed health
12 reference levels), their exclusion from the cumulative risk estimates and characterization may
13 seriously distort the portion of the total estimated risk thought to be posed to the population by a
14 specific evaluated source. A second issue is the problem of whether background chemical
15 exposures can be clearly distinguished from specific source-related chemicals, and how to quantify
16 these exposures. It may be important in a cumulative risk assessment to estimate background
17 exposures separately from specific source-related exposures, so that the risk assessor can provide
18 the community with a more complete picture of both total and known source-related risks. This
19 also provides a clearer, more complete picture for making risk management decisions. Finally,
20 there may be problems in identifying representative areas for designating as "background" for
21 comparison.
22
23 Finally, background exposures for a community or population may also include both
24 voluntary and involuntary exposures, and subsequent risks. Involuntary exposures are associated
25 with the naturally-occurring or anthropogenic background concentrations described above.
26 Voluntary exposures, such as are associated with lifestyle decisions, are exposures due to
27 activities such as smoking, consuming char-grilled meats with PAHs, or other choice-based
28 exposures, and may also sometimes be defined in the assessment as "background" exposures if
29 they are not assessed directly in the cumulative risk assessment.
30
31
32 APPENDIX F: RESEARCH AND DEVELOPMENT NEEDS
33
34 The Framework for Cumulative Risk Assessment is intended to serve as initial guidance,
35 providing a basic structure for the issues and defining key terms and concepts. In some cases, the
36 concepts introduced in the Framework require the application of knowledge and methods that are
37 not currently available. The following is a discussion of the needed areas of research and
38 methods development, highlighted within the Framework document, that may be most important
39 to an evaluation of cumulative risks. This is not intended to be a comprehensive listing of
40 cumulative risk assessment research needs.
41
H-95
-------
DRAFT- Risk Assessment Forum Review Draft - August 2, 2001 - Do Not Quote or Cite
1 Understanding the Timing of Exposure and its Relationship to Effects
2
3 A key concept in the definition of cumulative risk is that it represents an accumulation of
4 risk over time. However, unlike the traditional approach to risk assessment where exposure
5 events are summed and averaged over a period of time, cumulative risk assessment will involve
6 developing an understanding of how the sequence and timing of exposures influence the ultimate
7 risk of effects. For example, for multiple stressors, it is important to understand how prior
8 exposures to one or several stressors influence the risks from subsequent exposures to the same
9 or different stressors. In addition, it is important to understand the implications of these
10 exposures occurring during critical periods of an individual's life (e.g., important periods of
11 development or periods of disease). Several exposure models are under development which
12 recognize the need to understand the timing of various exposure events (e.g., Calendex, APEX,
13 Lifeline, SHEDS, and CARES).
14
15 In addition to gaining a better understanding of the sequence and timing of routine
16 exposures and their relationship to effects, it is important to understand how acute, non-lethal
17 exposures from accidents contribute to chronic or long-term effects.
18
19 Understanding the Composition and Toxicity of Mixtures
20
21 Chemical mixtures can change or degrade over time and space making the assessment of
22 exposure a particular challenge. For cumulative risk assessment, the composition of the mixture
23 at the point of contact with the receptor needs to be well characterized. Both measurement
24 techniques (at the receptor) and predictive models are applicable in this characterization.
25
26 EPA's Guidance for the Health Risk Assessment of Chemical Mixtures (USEP A, 2001 a)
27 presents approaches for combining the toxicities of multiple chemical stressors. These approaches
28 necessarily involve a number of simplifying assumptions when the mixtures are complex.
29 Although the current methods provide a valuable resource for assessing cumulative risks, future
30 cumulative risk assessment will need a more complete understanding of the interactions among
31 chemicals in complex mixtures. Some current research efforts are seeking to identify toxicologic
32 principles of joint action that are applicable to mixtures involving many chemicals.
33
34 Applying the Risk Factor Approach to Environmental Health Risks
35
36 The risk factor approach has been used in the medical profession to predict the chances of
37 individuals developing various diseases. It has proved to be a useful approach not only in
38 assessing certain cumulative risks, but also in communicating with patients. In this approach,
39 characteristics of a population (e.g., age, ethnicity, personal habits, genetic polymorphisms, prior
40 diseases, etc.) are correlated with the incidence of disease. For some diseases (e.g., breast cancer,
41 coronary artery disease, stroke) these correlations are well established. However, there are
H-96
-------
DRAFT- Risk Assessment Forum Review Draft - August 2, 2001 - Do Not Quote or Cite
1 substantial data gaps in terms of the role played by exposures to environmental stressors in the
2 development of human disease, and correlations of environmental exposures with disease
3 outcomes are generally not available.
4
5 Using Biomarkers and Biomonitoring
6
1 The use of biomarkers of exposure or effect holds a great deal of promise for cumulative
8 risk assessment. This approach can provide a method to assess stressors in groups. Currently,
9 however, this approach is not practicable when considering a large number of diverse stressors,
10 since appropriate biomarkers for many types of stressors have not yet been developed.
11
12 Considering Hazards Presented by Non-Chemical Stressors
13
14 Cumulative risk assessment could encompass the interactions of chemical stressors with
15 biological stressors, physical stressors, ecological stressors, radiological stressors, socioeconomic
16 stressors and lifestyle conditions. In trying to assess all these different types of stressors, it is
17 helpful to determine what types of effects the stressors produce, and then to try to group stressors
18 by like effects. Ideally, one would like to know the mechanism or mode of action by which
19 various stressors cause effects to allow a more refined grouping. Currently, however, there are
20 few methods to understand how these disparate stressors interact to result in risk.
21
22 Considering Psychological Stress as Part of Cumulative Risk
23
24 Psychological stress causes both psychological and physiological changes that can be
25 measured. Assessing levels of stress and their potential contribution to risk, however, is difficult
26 for a variety of reasons. The Agency for Toxic Substances and Disease Registry (ATSDR) began
27 the process of identifying research needs in this area through an expert panel workshop held in
28 1995. There is need for followup research in this area.
29
3 0 Considering All Aspects of Vulnerability
31
32 The issue of the vulnerability of a population can be thought of as having four
33 components: susceptibility/sensitivity of individuals, differential exposures, differential
34 preparedness to withstand the insult, and differential ability to recover from effects. Traditional
35 risk assessment may consider one or more of these categories but rarely are all considered. The
36 overall consideration of all four categories may be more important in cumulative risk assessment
37 than in traditional one-chemical assessments. A cumulative risk assessment, for example, may
38 need to consider potential combinations of high exposure and high vulnerability across stressors.
39 Methods development work is needed in this area.
40
41 Methods for Combining Different Types of Risk
H-97
-------
DRAFT- Risk Assessment Forum Review Draft - August 2, 2001 - Do Not Quote or Cite
1 Another key concept in the definition of cumulative risk assessment is that it represents the
2 integration of effects from stressors acting together. This implies that, in some cases, it may be
3 necessary to combine disparate measures of risk (i.e., different types of effects) to simplify the
4 expression of cumulative risks. There have been some attempts to collapse complex arrays of risk
5 into a few or even a single measure. These approaches have involved the use of common metrics
6 (e.g., Quality Adjusted Life Years, Disability Adjusted Life Years, Loss of Life Expectancy, etc.),
7 indices (e.g., Hazard Ranking System, etc.), and the categorization of effects (e.g., as for
8 categorical regression). Alternatively, Geographic Information Systems (GIS) and mapping
9 techniques can be used to graphically portray integrated information on risks without
10 mathematically combining disparate measures. Much methods development work remains to be
11 completed in each of these areas.
H-98
------- |