UNITED STATES ENVIRONMENTAL PROTECTION AGENCY
WASHINGTON, D.C, 20460
April 8, 1986 SSB~EES:-86-018
Honorable Lee M* Thomas
Administrator THe
U.S. Environmental Protection
Agency
401 M Street SW
Washington, DC 20460
Dear Mr. Thomas:
During recent meetings of the Environental Health Conmittee of EPA's
Science Advisory Board, the members have discussed, both formally and
informally, possible ways in which EPA and the Committee might enhance the
efficiency with which we carry out our joint responsibilities in preparing
and reviewing risk assessments, respectively. Seme changes in the Committee's
methods for planning scientific reviews need to be made in the near future
because the volume of requests for reviews is growing rapidly.
The following issues represent the Committee's current concerns and
viewpoints*
Multiple Risk Assessment Documents
Different EPA program offices frequently issue their own technical source
document for the same substance.^ The Committee understands that this
practice arose because of different regulatory schedules and requirements.
The Conmittee is concerned, however, that this practice may be taken as
evidence of insufficient planning ,and coordination within the Agency.
Issuing multiple source documents for a single substance, and seeking
separate scientific reviews for each document, is not an efficient use of
EPA or SAB resources, and contention arises when minor inconsistencies are
noted by external parties.
1 For more dicussion of this issue for a specific substance, see the Committee's
report of April 26, 1985, on the Agency's draft Health Assessment Document for
Dioxins. Several reports on dioxins were prepared within the Office of Research
and Development (ORD) for different program offices. The Conmittee is aware of
more than one document prepared by OED and the Office of Toxic Substances for
asbestos, butadiene and an addendum for methylene chloride. The Committee has
also noted, during its recent reviews of Office of Drinking Water (QDW) Health
Advisories, that Health Assessment Documents and Drinking Water Criteria
Documents (both prepared by ORD) exist for the following substances: chlorc~
benzene, dichlorobenzene, epichlorohydrin, hexaehlorobensene, mercury, nickel,
tetrachlorodibenzo-pj-dioxin and toluene. Finally, ORD and ODW (or QDW contractors)
developed Drinking Water Criteria Documents for the following compounds: arsenic,
cadmium, carbon tetrachloride, chlorobenzene, chromium, raethylene chloride,
trichloroethylene, methyl chloroform, tetrachloroethylene and vinyidene chloride.
-------
-2-
What is unclear is the degree to which multiple documents imply the exis-
tence of differing scientific conclusions. In addition, such documents are
differently structured and do not clarify how they satisfy an Agency need not
net by an existing document.
The Committee recommends that the Agency utilize a core document approach.
By "core" document, the Committee means a critical evaluation of the available
health and exposure data, such that the needs of all program offices would be
met. The document could be supplemented by individual program offices. Several
characteristics of this approach include? joint planning by EPA programs to
identify their individual and collective technical assessment needs for future
documents; and use of one core document as the technical basis for program specific
regulatory, activities. To the extent that individual program offices rely on media-
specific data such as exposure information, these should be regarded as supple-
mentary to the core document and should be carefully labelled and explained,
Integration oŁ Data
Agency offices do not apply consistent decision rules in preparing assessment
documents. Many of the documents submitted for SAB review, for example, address
the hazard evaluation or dose-response functions, while others represent a risk
characterization (including exposure data). Ihe Committee recommends that the
Agency prepare documents which integrate hazard data with exposure- data (to the
extent such data are available) in order that the sources of and implications for
risk estimates are made clear.* Even a preliminary exposure assessment will
clarify the way in which the Agency plans to use such data in risk assessment*3
Hazard data also change, and the Agency has procedures in sane areas which
allow additions to a document, based on receipt of a new mechanism, bioassay or
epidendological information without rewriting the entire evaluation. Similarly,
some offices have used a detailed cover memorandum to expand interpretation of
scientific data in technical documents.*
While the desire of program offices to retain flexibility to meet evolving
and scroetimes rapid technical changes is reasonable, appropriate caveats can be
placed in core documents to satisfy future, program-specific needs. Tfie peer-
reviewed, published monitoring and modelling information can be described and
integrated with hazard data to estimate risks, these estimates need not be
prepared in great detail for each environmental route of exposure. However, if
a risk assessment serves as the technical basis of a regulatory decision, and if
Agency decision makers desire a Science Advisory Beard review, the Committee
^ See the SAB's Resolution on Exposure Assessment dated October 25, 1984*
3 The Office of Air Quality Planning and Standards (CBQPS) has submitted memoranda
to the Committee that summarize current exposure information on hazardous air pollu-
tants. For the most part, this brief summary data has been adequate to enable
Committee members to better understand the significance of health effects from
individual substances. However, for an evaluation of a more extensive exposure
analysis, see the Committee's October 1, 1985 report of its review of the Office
of Toxic Substances' assessment of risks from formaldehyde.
4 As examples, see the ODW memoranda entitled "Quantification of lexicological
Effects" that further interpret the information from OSD Health Assessment
Documents for substances such as ethylene dichloride.
-------
—3™
should evaluate as complete a risk assessment as is possible for the Agency to
develop,
Scope of Reviews
Standard Agency practice for core documents should be to review the available
literature on a substance in a comprehensive fashion. Hie Committee has been
pleased with the critical and concise evaluation in sane of the assessments
submitted for review, but assessments that look at only a subset of the existing
data for a substance can inply to the public that EPA has ignored or discarded
data in conflict with a prior position. Ihis practice, if not fully explained,
can erode the Agency's scientific credibility. Where unacceptable data are
found, then EPA, should say so* If less comprehensive documents need to be
prepared, their purpose should be explicitly stated, along with an explanation
for the more limited focus.
CutoffDates for Literature Reviews
The Committee often has requested that Agency staff provide in the preface
to a technical document, a date at which they believe the literature citations are
complete.^ The Committee's position is that EPA cannot always update the litera-
ture while simultaneously writing and reviewing a document. Therefore, studies
appearing after a cut-off date may (but need not) be acknowledged, and the
new data should be reviewed only at the discretion of Agency staff. A document
with a cutoff date will be defensible against charges that it is incomplete,
whereas the lack of a cutoff date may confuse the public, particularly when
a long delay occurs between the literature review and final document publication,
Modem Terminology
The Agency's scientific documents will further benefit from the use of
current terminology, particularly when referring to issues such as hazard and
risk. For exantple, the term "unit risk estimate" found in some EPA documents
more commonly is called a "potency" by the toxicological community,
Inconsistency of Nomenclature
Agency documents often carry different labels, depending on the program
office issuing the document. For example, quite similar documents are labelled
as health assessments, criteria documents or risk assessments* More precise and
consistent titling would help communicate to the public more clearly the aims of
such documents and provide the Committee with firmer guidelines for their review.
Seme inconsistency may be unavoidable because of statuatory language. For this
and other reasons, internal EPA guidance on the preparation of documents is
likely to enhance staff and public understanding.
Assignment of Priorities
During the past several years, the Environmental Health Ccnndttee has received
multiple requests from the Agency for reviews of essentially the same scientific
3 See, for example, the Ccronutteen's report of January 4, 1985 on OSD's Health
Assessment Documait for 1,2-dichloroethane ((ethylene dichloride).
-------
data. Given its limited resources, the Coraaittee needs further EPA assistance
in assigning priorities to those draft documents for which review is desired.
The Committee has less capability to make such management decisions. There are
several criteria by which the Agency can establish a high priority for review,
such as widespread potential health effects, legal requirerasnts, a nee<2 of
several program offices for the docutnenti and important or unique scientific
issues .
The above list of concerns and views is raised for your attention, it
represents views shaped by the past several years of experience in reviewing
the Agency's scientific documents but is not meant to be comprehensive or
prescript ive. Ws recognize that the issue of ijmproved planning and coordination
of risk assessments is currently receiving a high priority by senior Agency
managers. We hope this letter will be taken in the spirit of supporting this
effort and of improving our own scientific review methods. As these issues
are long-standing i we would now welcome a written response frcm Agency staff,
which the SAB will review,
Sincerely yours,
Richard A. Grieseraer
Chair, Environmental Health Gonroittee
Norton Nelson
Chair„ Executive Committee
------- |