Unittd State*
Environmental Protectton
Agency
Off ic« of
Solid Wast* and
Emergency Re*pon*
oEPA
DIRECTIVE NUMBER: 9285.7-13
SilLt. Implementing the Deputy Administrator's Risk
Characterization Memorandum
APPROVAL DATE: 5/25/92
EFFECTIVE DATE: 5/26/92
ORIGINATING OFFICE: 0ERR
Q FINAL
D DRAFT
LEVEL OF DRAFT
n A — Signed by AA or DAA
n B — Signed by Office Director
dj C — Review & Comment
REFERENCE (other documents):
OS WER OS WER OS WER
ME DIRECTIVE DIRECTIVE Dl
-------
-------
_,"'1 ~-:~ :::a:~s :-..::--o/".,a, ::.:, ~::.':i. .:.,;~C'! -'~-::,~~._-:~ \
,oEPA .\ ~s::,r:;::,~ :c 2C4I50
OSWER Directive Initiation Reauest 9285. 7-13
2. Orlolnator Inlomatlon
Name 0' C,::,-::ac: ':er3on .'..1a,1 C.:ce :: ~ce . :-eJ~r.cr.e :.::::e
Betti VanErmF; 5201G OERR 260-9760
J. Tille
Implementing the Deputy Administrator's Risk Characterization MemorandtIDl
4. Suml":"ar{::1 ::rp.c::'.e >:r:::~::e one! s:a:el":",e~: :r :::.r.::cse:
The purpose of this memorandtIDl is to implement in the Superfund program the
recommendations of the Deputy Administrator in his memorandum of February
26, 1992, "Guidance on Risk Characterization for Risk Managers and Risk
Assessors. /I
5. :,e 5uper~eae PreviOus :':rec:",e:sl' : ; No I ~ Yes What directIVe (numoer. mle)
1L- -
b O.,es It S~pplemen! Previous DirectlVe(s)' ---, Ilve!
.xJ No What direc:ive (number. trtle)
-
.
~r._e'tel - I I C.. For Review &. Comment ' : J - In Oevejcpm~:
I ' ~ - Sl~ned :;;y AACAA ! 1 S u 5,gMd t)'f Office Cirec!or
i I -
8. Document to be distributed to States by Headquarters?
,
I
-
; Ves
~
." ~..:
.Oare
:,,~~
: .,A:e
/
EPA Form 1315-17 (Rev. 5-87) PreviOUS editionS are ooso:e!e.
OSWER
'.IE DIRECTIVE
O~'f\'ER OSWER 0
DIRECTIVE DIRECTIVE
-
~ .
-------
United States
Environmental Protection
Agency
Office of
Solid Waste and
Emergency Response
8EPA
DIRECTIVE NUMBER: 9285.7-13
TITLE: Implementing the Deputy Administrator's Risk
Characterization Memorandum
APPROVAL DATE:
EFFECTIVE DATE:
5/26/92
5/26/92
ORIGINATING OFFICE: OERR
Q FINAL
o DRAFT
LEVEL OF DRAFT
o A - Signed by M or OM
o B - Signed by Office Director
D. C - Review & Comment
REFERENCE (other documents):
OSWER OSWER OSWER
DIRECTIVE DIRECTIVE
-------
..;~t.O ~4r~#
i ft ~
\Sf
i>4( 1JIIIt1I4f'';I
UNITED STATES ENVIRONMENTAL PROTECTION AGENCY
WASHINGTON. D.C. 20460
MAY 2 6 \99Z
OFFICE OF
SOLID WASTE AND EMERGENCY RESPONse
MEMORANDUM
SUBJECT:
OSWER Directive 9285.7-13
FROM:
TO:
Bruce Diamond, Directo
Office of Waste progra
Directors, Waste Management Division
Regions I, IV, V, and VII
Director, Emergency and Remedial Response Division
Region II
Directors, Hazardous Waste Management Division
Regions III, VI, VIII, and IX
Director, Hazardous Waste Division
Region X . .
Directors, Environmental Services Division
Regions I, VI, and VII
PurDose
The purpose of this memorandum .is to implement in the
Superfund program the recommendations of the Deputy Administrator
in his memorandum of February 26, 1992, "Guidance on Risk
Characterization for Risk Managers and Risk Assessors."
Background
')
On. February 26, 1992, in a memorandum to Assistant
Administrators and Reqional Administrators, the Deputy
Administrator issued new quidance for Aqency managers and risk
assessor. on describi~g risk assessment results in EPA reports,
presentationa, and decision packages (The "Risk Characterization
Guidance"). The Risk Characterization Guidance is designed to
ensure a full and complete analysis of risk in the decision-
making process and to promote greater consistency and
comparability in risk assessments across Aqency proqrams. It is
the culmination Of a mUlti-year project of the EPA's Risk
Assessment Council to improve the Agency',s risk assessment
process. Attached is a copy of the Risk Characterization
Guidance and its accompanying appendix. A similar copy was
provided to Regional Superfund risk assessors early in March.
@ Pnnte(J 0" gK~C'ed P3I»f
-------
,
I
I '
2
The Risk Characterization Guidance is relevant to all
Superfund risk assessments since it discusses not only risk
assessments performed for national regulations but also site-
specific risk assessments, e.g., Superfund baseline risk
assessments.
After studying the Risk Characterization Guidance and
current Superfund policy, we are pleased to report that Superfund
policy already addresses most of the points raised in the Risk
Characterization Guidance. Implementation of current policy with
minor supplementation'should bring Superfund, risk assessment and
risk management fully in line with recommendations in the Risk
Characterization Guidance. ",
Specifically, the Risk Characterization Guidance lists the
following three principles for presenting risk assessment
information in new Agency reports, presentations, and decision
packages:
1)
Risk assessment information should be clearly presented
separate from any non-scientific risk management
considerations.
2)
Key scientific data and methods and their uncertainties
should be identified in the risk characterization and a
statement of confidence should be included that
identifies all major uncertainties along with comment
on their influence on the assessment.
Information on the range of exposures derived from
exposure scenarios and on the use of multiple risk
descriptors shoul~ be presented.
Our current pOlicies address the principles set out in the
Risk Characterization Guidance in the following ways:
3)
1)
The superfund remedy evaluation and selection process
is designed to keep risk assessment and risk management
separate. The results of the baseline risk assessment
are just one of the tools used by risk managers for
making cleanup decisions. .
The Risk Assessment Council's (RAC's) guidance states
that uncertainties should be presented for each step of
the risk assessment process. The Risk Assessment
Guidanc, for SUDerfund. Part A is already in line with
this guidance. Each chapter of Part A has a section
specifically devoted to presentation of risk
assessment-related uncertainties to aid risk managers,
Further, we deal with uncertainty a~d variability in
site sampling data through our Guidance on Data
Useabilitv and use of the 95 per~ent upper confi=c~=:
2)
-------
3
limit (95 UCL) on the arithmetic mean of site sampling
data. In addition, information on the level of
uncertainty in the toxicity criteria we use in our
- assessments is provided by the Office of Research and
Development in the Integrated Risk Information System
(IRIS) and the Health Effects Assessment Summary Tables
(HEAST) they prepare for the Superfund program.
Thus we can meet the second principle by inclusion in
the Record of Decision of appropriate highlights from a
well-done baseline risk assessment completed as part of
the Remedial Investigation.
, .
3)
currently, we address two of the four risk-descriptors
listed in the RAC guidance. Our "Reasonable Maximum
Exposure" (RME) estimate is designed to be a measure of
"high-end" exposure. Our guidance also advocates
assessments of risk for sensitive subpopulations (e.g.,
childhood exposure to lead, recreational and
subsistence fishers).
We do not, however, in our current assessments typically
include an estimate of central tendency (or average) exposure or
an estimate of population risk. In Regions where average
exposures are estimated, findings are often presented in the
uncertainty section of the risk assessment. Thus, in order to be
fully in line with the third principle, the superfund program
should develop additional guidance on estimating central tendency
exposures and on addressing population risk, recognizing, however'
that due to the lack of sufficient exposure data at most
Superfund sites, it generally is not possible to estimate
population risks.
. Obiective
The Deputy Administrator recognized that program offic..
must continue to make risk management decisions during the time
in which changes in policy are made to reflect the principl.. ot
the Risk Characterization Guidance. Specifica'lly he states !hat
"we do not expect risk assessment documents that are close to
completion to be rewritten." 'Our objective is speedy, effective
updating of SUperfund policies on risk assessment an~ risk
management, while continuing to meet all remedial program
objectives.
ImD'lementation
To implement in the Superf11nd program the recommendat ~"n, :..
the Risk Characterization Guidance, the fOllowing steps shou~j c.
taken: '
o
1..0 iIlodificatioiilii ba:i.:sd Oil this Risk CharactEsri~C&t"(..I.
-------
'[
4
Guidance need to be. made for Records of Decision (RODs)
already signed or the risk assessments supporting them.
o
RODs to be signed in FY 93 should consider the Risk
-Characterization Guidance in developing and drafting
their risk management decisions. This may require some
additional limited risk assessment work, particularly
to provide an estimate of central tendency exposure.
o
For risk assessments completed or close to completion
in support of FY 92 RODs:
A risk assessment for a site for which the
Proposed Plan has been developed generally need
not be revised to reflect the principles in the
Risk Characterization Guidance;
Risk assessments in draft or under development
should generally include data to reflect the
principles in the Risk Characterization Guidance.
A supplemental discussion of average exposure as
part of the uncertainty discussion, either in the.
risk assessment or the ROD, should generally be
sufficient.
o
Risk assessment guidance on performing "central
tendency" exposure assessments should be developed:
A group of Headquarters and Regional Superfund
risk assessors is already working to provide
guidance to be in line with the third principle
and generally aid in interpreting the Risk
Characterization Guidance. The group will first
develop guidelines for ev~luating central
tendency. Any ~~rther recommeudati:--s on
addressing population risk will come later, after
the group has completed its work on central
tendency and any other descriptors.
Regardinq risk management, we will continue to use the RME
scenario, a. described in the preamble of the National
Contingency Plan, in the remedial decision in evaluating what is
necessary to achieve protection against risk t.o human health.
The central tendency estimate will be used for informational
purposes in discussing uncertainties.
Attachment
co:
Regional Branch Chiefs
HQ Division Directors
-------
.;1t.D "..,
'I ft i
.'S'
\.. ,I
?4t ~W
UNITEO STATES ENVIRONMENTAl.. PROTECTION AGENCY
WASI-4INGTCN. C C 204&0
FEB 26 1992
O~~'CI o~
""1 &0""""'1".""0-
MJD(ORAllDUM
TO:
Guidance on Risk Characterization
and Risk Asses.or.~
F. Henry Habicht I ~n ~ V
Deputy Admini.trato ~~
A..i.tant Admini.trator.
Reqional Admini.trator.
for Ri.k Manaqers
SUBJWC'f :
raC1lla
IlI'raODUC'fI08
This memorandum provide. quidance for manaqer. and assessors
on describinq risk a...ssment re.ult. in BPA report.,
presentation., and deci.ion packaqe.. The quidance addresses a
problem that affect. public perception reqardinq the reliability
of EPA'. .cientific a..e..ment. and related requlatory decisions.
EPA has talented scienti.t., and public confidence in the quality
of our scientific output will be enhanced by our vi.ible
interaction with peer scienti.t. and thorouqh pre.entation of
risk a..e..ment. and underlying .cientific data.
Specifically, althouqb a qreat deal of careful analy.i. and
scientific judgment q08. into the development of BPA ri.k
a..e..ment., .ignificant information i. often omitted a. the
re.ult. of the a..e..ment are pa..ed alonq in the deci.ion-making
proce... Often, vhen ri.k.information i. pre.ented to the
ultimate deci.ion-8&ker and to the public, the re.ult. have been
boiled dawD to a point e.timate of ri.k. Such -.hort hand-
approach.. to ri.k a..e..ment do not fully convey the ranqe of
information con.idered and u.ed in developinq the a..e..ment. In
.hort, infor8Ati.e ri.k characterization clarifies the .cientific
ba.i. for IPA deci.ion., while number. alone do not qi.e a true
picture of the a..e..ment. .
Thi. probl.. i. not BPA'. alone. Aqency contractor.,
indu.try, environmental qroup., and other participant. in the
overall requlatory proce.. u.e similar -.bort band- approache..
w. mu.t do everythinq we can to en.ure that critical
information from eacb .taqe of the ri.k a..e..ment i.
communicated from ri.k a..e..or. to their manaqer., from middle
-------
2
to upper "~agement, from EPA to the public, and from others to
EPA. The R~sk Assessment Council considered this problem over
many M9.nth. and reached several conclu.ion.: 1) We need to
present a full and complete picture of risk, including a
statement of confidence about data and methods used to develop
the assessment; 2) we need to provide a ba.i. for greater
con.i.tency and comparability in ri.k a..e..ment. across Agency
program.; and 3) profe..ional scientific judgment play. an
important role in the overall .tatement of ri.k. The Council
al.o concluded that Agency-wide guidance would be u.eful.
BACKOROUIID
Principle. empha.ized during Ri.k A..e..ment Council
discus.ion. are summarized below and detailed in the attached
'Appendix.
rull Characteri.atioD 01 Ri.k
EPA deci.ion. are ba.ed in part on ri.k a..e..ment, a
technical analysis of scientific information on exi.ting and
projected risk. to human health and the environment. A.
practiced at SPA, the risk a..e..mentproce.. depend. on many
different kinds of scientific data (..~., expo.ure, toxicity,
epidemiology), all of which are u.ed to wcharacterizeW the
expected risk to human health or the environment. Informed use
of reliable scientific data from many different .ource. i. a
central feature of the ri.k a..e..ment proce...
Highly reliable data are available for many a.pect. of an
as.essment. Sowever, .cientific uncertainty i. a fact of lif,
for the ri.k a..e....nt proce.. a. a whole. A. a re.ult, agency
manager. make deci.ion. u.ing .cientific a..e...ent. that are
les. certain than the id.al. Th. i..ue., then, become when i.
scientific confid.nce .ufficient to u.e the a..e..ment for
decision-making, and how .hould the a..e..ment be u.ed? In order
to mak. th... deci.ion., manager. need to under.tand the
strenqt~ aDd the limitation. of the a..e..ment.
'>
OD ~. point, the guidance empha.ize. that informed SPA
risk a.....or. and 8&Dager. need to be compl.tely candid about
confidence and uncertainti.. in de.cribing ri.k. and in
.xplaining regulatory deci.ion.. Specifically,. the Agency'. risk
a....sment guideline. call for full and open di.cu..ion of
uncertainties in the body of each BPA ri.k a.....ment, including
prominent di.play of critical uncertainties in the ri.k
characterization. Numerical ri.k e.timat.. .hould alway. be
accompanied by de.criptive information carefully ..lected to
ensure an objective and balanced characterization of ri.k in risk
a..e...ent report. and regulatory document..
-------
3
Scientists call for fully characterizing risk not to
question the validity of the assessment, but to fully inform
others_about critical information in the assessment. The
emphasis on "full" and "complete" characterization does not refer
to an ideal assessment in which ri.k is completely defined by
fully satisfactory scientific data. Rather, the concept of
complete ri.k characterization mean. that information that i.
needed for informed evaluation and u.e of the a..e.sment i.
carefully highlighted. Thu., even though ri.k characterization
details limitation. in an a.sessment, a balanced di.cussion of
reliable conclu.ions and related uncertainties enhances, rather
than detrac:ts, from the overall credibility of each assessment.
. This quidance is not new. Rather, it re-state., clarifies,
and expand. upon current risk a..e..ment concept. and practices,
and emphasizes aspects of the proce.. that are often incompletely
developed. It articulates principle. that have long quided
experienced risk assessors and well-informed risk managers, who
recognize that risk is best described not a. a cla..ifica~ion or
single number, but as a composite of information from many
different sources, each with varying degree. of .cientific
certainty.
eoaDarabilit. aadeoasisteacy
The Council's second finding, on the need for greater
comparability, aro.. for several rea.ons. One wa. confusion --
for example, many people did not under.tand that a ri.k estimate
of 10.' for an "average" individual should not be compared to
another 10.' risk estimate for the "mo.t expo.ed individual".
Use of such apparently .imilar e.timate. witbout furtber
explanation lead. to mi.understanding. about the relative,
siqnificance of riaks and the ~r.otectivene.. of ~i.k reduction
dctiona. Another cataly.t for change va. the SAB'. report,
Reducina R~: Settina Priorities and Strateaie. for
Environmental Protection. In order to impl...nt the SAB'.
recommendation that ve target our effort. to achieve the greatest
risk reduction, we need common mea.ure. of ri.k.
1.&'. nevly revi.ed Ixposure A..e....nt Guideline. provide
standard d88Criptor. of exposure and ri.k. O.e of the.e terma in
all Agency ri.k a..e..ment. will promote con.i.tency and
comparability. O.e of .everal de.criptor., rather than a single
.descriptor, vill enable u. to pre.ent a more complete picture of
risk that correspond. to the range of different expo.ure
conditions encountered by various population. expo.ed to mo.t
environmental chemicals.
'rofes.ioDal Jud.-eat
The call for more extensive characterization of risk has
'obvious limits. For example, the ri.k cbaracterization includes
-------
4
-
only the ...t siqnificant data and uncertainties from the
assess~D~ (those that define and explain the main risk
conclus.0ft8) .0 that decision-makers and the public are not
overwhe~med by valid but secondary information.
The degree to which confidence and uncertainty are addressed
depend. largely on the .cope of the a..e..ment and available
re.ource.. When .peeial circum.tance.
-------
5
package.. In general, such information should be pre.ented a.
carefully .elected highlights from the overall asses.ment. In
this reqard, common .en.e reqardinq information needed to fully
inform Agency deci.ion-maker. is the be.t quide for determining
the information to be highlighted in decision packages and
briefings. .
1.
Reqardinq the interface between ri.k a..e..ment and
risk management, ri.k a..e..ment information mu.t be
clearly pre.ented, .eparate from any non-.cientific
risk management conJideration.. Discu..ion of risk
manage.ent option. should follow, ba.ed on .
con.ideration of all r.1evant factor., .cientific and
non-.cientific.
2.
Regarding ri.k characterization, key .cientific
information on data and method. (S.S., u.e of animal or
human data for extrapolating fro. high to low do.es,
u.e of pharmacokinetics data) mu.t be higblighted. We
al.o expect a .tatement of confidence in the asse.sment
that identifies all major uncertainties along with
comment on tbeir influence on the a..e..ment,
con.i.tent with quidance in the attached Appendix.
Regarding expo.ure and ri.k characterization, it is
Agency policy to pre.ent information on the range of
expo.ure. derived from expo.ure .cenario. and on the
. u.e of multiple risk-de.criptor. (i.s., central
tendency, high end of individual risk, population risk,
important .ubgroup., if known) con.istent with
terminology in the attached Appendix and Agency
quideline.. .
3.
Thi. quidance applies to all Agency office.. It applies to
a..ess.ent. generated by BPA .taff and to tho.e generated by
contractor. for BPA'. u.e. I believe adherence to this Agency-
wide quidance will i8prove under.tanding of Agency ri.k
as.e..ment., lead to acre informed deci.ion., and heighten the
credibility of both a..e..ment. and deci.ion..
r~ this tim8 forward, pre.entation., report., and decision
package. fraa all Agency office. .hould characterize ri.k and
related uncertainties a. de.cribed here. Plea.e be prepared to
identify and di.cu.. with .e any program-.pecific modification.
that may be appropriate. lowever, we do not eXpect ri.k
a..e.s.ent document. that are clo.e to completion to be
rewritten. Although this i. internal quidance that applies
directly to a..e...ent. developed under BPA au.pice., I al.o
encouraqe Aqency .taff to u.. th..e principle. a. guidance in
evaluating a..e..ment. .ubmitted to BPA ~ro. other .ource., and
in di.cus.ing the.e .ub.i..ion. with me and with the
Admini.trator.
-------
6
Thi.guidance is intended for both manaqement and technical
staff. ,1.a.. distribute this document to tho.e who develop or
review .....sments and to your manaqers who use them to implement
. Aqency pr09rams. Also, I encouraqe you to di.cu.. the principles
outlined here with your .taff, particularly in briefinq. on
. particular asses.ment.. .
In addition, I expect that the Ri.k As.es.men~ Council will
endors.' new guidance on Aqency-wide approaches to ri.k
charact.rization now b.inq developed in the Ri.k A..essment Forum
for EPA'. ri.k a.se.sment quideline., and that the Aqency and the
Council will auqment that quidance a. needed.
The Admini.trator and I believe that thi. effort i. very
important. It further. our qoal. of riqor and candor in the
preparation, presentation, and use of EPA ri.k a...s.ment.. The
ta.ks outlined above may require extra effort from you, your
manaqers, and your technical staff, but they are critical to full
implementation of these principle.. We are mo.t qrateful for the
hard work of your representative. on the RAC and other staff in
pullinq this document toqether. I appreciate your cooperation in
this important area of science policy, and look forward to our
discus. ion..
AttachDient
cc:
The Administrator
Risk As.e.sment Council
-------
OUIDABC8 POR RIIX A8'8SSMI~
Section 1.
Ri.k A..e..ment-Ri.k Manaqement
Interface
Section 2.
Ri.k Characterization
Section 3.
Expo.ure and Ri.k Ce.criptors
u.S. Environmental Protection Aqency
Ri.k A.....ment Council
Nov.mber, 1991
-------
SIC~IO. 1.
RISK ASSESSMENT - RISK MAHAGEMEKT INTERrACE
Recoqnizing that for many people the term risk assessment
has wide meaning, the National Research Council's 1983 report on
risk assessment in the federal government (hereafter "HRC
report") distinguished between ri.k a..es..ent and" ri.k
management 0
Broader uses of the term [risk assessment] than ours
also embrace analy.i. of perceived ri.k.,
compari.on. of ri.k. a..ociated with different
regulatory .trategie., and occa.ionally analy.i.
of the economic and .ocial implication. of
regulatory decision. -- function. that we a..ian
to risk manaaement (empha.i. added). (1)
In 1984, EPA endorsed these di.tinction. b.tween risk assessment
and risk management for Agency u.e (2), and later relied on them
in developing risk a.sess.ent quideline. (3).
This distinction suggests that IPA participant. in the
process can be grouped into two main categorie., each with
somewhat different respon.ibilitie., ba..d on their roles with
respect to risk a..es..ent and ri.k manage.ent.
Risk A..e....~t
One group qenerate. the ri.k a.......nt by collectinq,
analyzing, and .ynthe.izing .cientific data to produce
the hazard identification, do.e-re.pon.e, and expo.ure
a..e..ment portion of the ri.k a..e..ment and to
characterize ri.k. Thi. group relies in part on Agency
ri.k a..e..ment quidelin.. to addre.. .cienc. policy
i..ue. and .cientific uncertaintie..
Generally, this qroup include. .cienti.t. and
stati.tician. in the Office of Re..arch and
Develop.ent, the Office of pe.ticide. and Toxic
Sub.tance. and other proqraa office., the Carcinogen
Risk As.e..ment Verification Indeavor (CRAVE), and the
RfD/RfC Workgroup..
2
-------
Others ~ analyses produced by the first group to
generate site- or media-specific exposure assessments
and risk characterizations for use in regulation
development. These assessors rely on existing
databa.es (..~., IRIS, ORD aealth Assessment Documents,
CRAVE and RfD/RfC Workgroup document.) to develop
regulations and evaluate alternatives.
Generally, this group include. .cientist. and analyst.
in program office., regional office., and the Office of
Research and Development.
Ri.k Manaoement
A third group inteorate. the ri.k characterization
other non-.cientific con.ideration. .pecified in
applicable statute. to make and ju.tify requlatory
decision..
with
Generally, this group include. Agency manager. and
deci.ion-maker..
Each group has different re.pon.ibilitie. for ob.erving the
distinction between risk asses.ment and rilk management.
same time, the ri.k as.e.sment proce.. involve. regular
At the
interaction between each of the group., with overlapping
responsibilities at various stage. in the overall proces..
The quidance to follow outline. principle. .pecific tor
those who generate, review, u.e, and integrate ri.k as.e..ment.
for decision-making-
3
-------
1. Ri.k a..e..or. and ri.k ..nager. .hould be .en.itive to
di.tinc~lOD. betv.en ri.k a..e..ment and ri.k aanage.ent.
The major participants in the risk assessment process have
many shared responsibilities.
Where responsibilities differ, it
is important that participants confine themselves to ta.ks in
their area. of re.pon.ibility and not inadvertently ob.cure
difference. between risk a.se..ment and ri.k management.
fhared respon.ibilitie. of a..essor. and managers include
initial decisions regarding the planning and conduct of an
assessment, discus. ions a. the assessment develop., decisions
reqardinq new data needed to complete an a..e..ment and to
address significant uncertaintie..
At critical junctures in the
assessment, such consultations .hape the nature of, and schedule
for, the assessment.
For the aenerators of the a..e..ment, di.tinquishinq between
risk assessment and ri.k management means that scientific
information i. selected, evaluated, and pre.ented without
considerinq non-.cientific factor. including how the scientific
analy.i. might influence the requlatory deci.ion.
A..e..or. are
charqed with (1) g.nerating a credible, objective, reali.tic, and
balanced analy.i., (2) pre.enting information on bazard, dose-
re.pon.., expo.ure and ri.k; and (3) explaining confidence in
each a..e....nt by clearly delineating uncertainties and
assumption. along with the impact. of th... fac.tor. (e.g.,
confidence limit., u.e of con.ervative/non-con.ervative
as.umptions) on the overall asse..ment.
They do not make
deci.ion. on the acceptability of any ri.k level for protecting
4
-------
public. health or selecting procedures for reducing risks.
For users of the assessment and for decision-makers who
integrate these assessments into requlatory decisions, the
distinction between risk ass.ssment and risk management means
refraining from influencing the ri.k de.cription through
consideration of non-scientific factor.
-- e. g. ,
the regulatory
outcome -- and from attempting to .hape the ri.k a.sessment to
avoid statutory con.traint., ..et regulatory objective., or serve
political purpo.e..
Such management con.ideration. are often
legitimate consideration. for the overall regulatory decision
(see next principle), but they have no role in e.timating or
describing risk.
Bowever, deci.ion-maker. e.tabli.h policy directions that
determine the overall nature and tone of Agency risk assessments
and, as appropriate, provide policy quidance on difficult and
controversial risk assessment is.ue..
Matters such as risk
assessment priori tie., degree of con.ervatis., and acceptability
of particular rilk levell ar~ ~elerved for d.ci.ion-make~s who
are charged with making deci.ion. regarding protection of public
health.
5
-------
2. the risk assessment product, that is, the risk
character1.ation, is only one of several kinds of inforaation
used fo~. ~gulatorr decision-aaking.
Risk characterization, the last step in r~sk assessment, is
the starting point for risk management considerations and the
foundation for regulatory decision-making, but it ~s only one of
several important components in such decisions.
Each of the
environmental laws administered by EPA calls for con8ideration of
non-scientific factors a~ various .tage. in the requlatory
process.
As authorized by different .tatute., decision-makers
. evaluate technical feasibility (e.g., treatability, detection
limits), economic, social, political, and legal factors a. part
of the analysis of whether or not to regulate and, if so, to what
extent.
Thus, regulatory decisions are usually ba.ed on a
combination of the technical analy.i. u.ed to develop the risk
assessment and information from other fields.
For this reason, risk a.sessor. and managers should
understand that the regulatory deci.ion i. usually not determined
solely by the outC988 of the risk a.se..ment.
That is, the'
analysis of the overall regulatory problem may not be the same a8
the picture pre.ented by the ri.k analysis alone.
For example, a
pesticide ri.k a..e..ment may de.cribe moderate risk to some
popula~ioD. but, if the agricultural benefit. of it. u.e are
importan~ for the nation'. food .upply,the product may be
allowed to remain on the market with certain re.triction. on U8e
to reduce po..ible expo.ure.
Similarly, a..e..ment efforts may
produce an RfD for a particular chemical, but other
6
-------
considerations may result in a regulatory level that is more or
less protective than the RfD itself.
For decision-makers, this means that societal considerations
(~.g., costs, benefit.) that, along with the risk assessment,
shape the regulatory decision should be de.cribed a. fully as the
scientific information .et forth in the ri.k characterization.
Information on data source. and analy.e., their strengths and
.
limitations, confidence in the a..e..ment, uncertainties, and
alternative analy.e. are a. important here a. they are for the
scientific component. of the regulatory decision.
Deci.ion-
makers should be able to expect, for example, the .ame level of
rigor from the economic analy.i. a. they receive from the risk
analysis.
Decision-makers are not "captive. of the number.."
On the
contrary, the quantitative and qualitative ri.k characterization
is only one of many important factor. that mu.t be considered in
reaching the final decision -- a difficult and di.tinctly
'different ta.k from risk a.se..ment per .e.
Ri.k management,
deci.ion. involve numerous a..umption. and uncertainties
regarding technology, economic. and .ocial factor., which need to
be explicitly identified for the deci.ion-maker. and the public.
7
-------
SECTION 2.
RISK CHARACTERIZATION
EPA risk assessment principles and practices draw on malny
sources_.
The environmental laws administered by EPA, the
National Research Council's 1983 report on risk assessment (1),
the Aqency's Ri.k Assessment Guideline. (3), and various proqram-
specific quidance (e.q., the Ri.k A.se..ment Guidance for
Superfund) are obvious .ource..
Twenty yenr. of EPA experience
in developinq, defendinq, and enforcinq ri.k a.....ment-based
requlation is another.
Toqether the.e various .ources stress the
importance of a clear explanation of Aqency proce..es for
evaluatinq hazard, dose-response, expo.ure, and other data that
provide the scientific foundation for characterizinq risk.
This section focuses on two requirement. for full
characterization of risk.
First, the characterization mu.t
address qualitative and quantitative feature. of the as.e..ment.
Second, it must identify any important uncertainties in the
assessment a. part of a di.cu..ion on confidence in the
assessment.
Thi. empha.i. on a full de.cription of all element. of the
assessmentdrav8 attention to the importance of the qualitative
as well.. the quantitative dimen.ion. of the a..e..ment.
1983 HlC report carefully di.tinqui.hed qualitative ri.k
The
assessment fro. quantitative as.e.sment., preferrinq ri.k
statements tha~ are not .trictly numerical.
The term. ri.k as.e..ment i. often qiven
narrower and broader meaninq. than ve
have adopted here. For .ome ob.erver.,
the term i. synonymous with auantitative
8
-------
. risk asse~sment and emphasizes reliance
~n numerical results. Our broader definition
~ncludes quantification, but also includes
qualitative expressions of risk. Quantitative
estimates of risk are not always feasible, and
they may be eschewed by aqencies for policy
reasons. (Emphasis in original) (1)
More recently, an Ad Boc Study Group (with represenatives
from EPA, BBS, and the private sector) on Ri.k Presentation
reinforced and expanded upon these principle. by specifyinq
several "attributes" for ri.k characterization.
1.
The major component. of ri.k (hazard
identification, do.e-re.pon.e, and
exposure assessment) are pre.ented in
summary statements, alonq with quantitative
estimates of risk, to give a combined
and integrated view of the evidence.
The report clearly identifies key
assumptionl, their rationale, and the
extent of scientific consen.ul; the
uncertainties thus accepted; and the
effect of reasonable alternative
assumptions on conclu.ion. and estimates.
2.
3.
The report outline. specific ongoing or
potential research projectl that would
probably clarify significantly the extent
of uDcertaiDt, in the ri.k e.timation.
. . . "( 4 )
Par~icularly critical to full characterization of risk is a
frank and open di.cu.lion of the uncertainty in the overall
assessment and in each of its component..
The uncertainty
stat.men~ i. important for s.veral realonl.
Information from different .ourc.. carrie. different
kind. of uncertainty and knowledge of the.e differences
is important when uncertainties are combined for
characterizinq rilk.
Decisions must be made on expending re.ource. to
acquire additional information to reduce the
uncertaintiel-
9
-------
~---
. .
A.c~ear.and explic~t statement of the implications and
l~~tat~ons of a r~sk assessment requires a clear and
explicit statement of ~elated uncertainties.
Uncertainty analysis qives the decision-maker a better
understandinq of the implications and limitations of
the assessments.
A discussion of uncertainty require. commeht .on such is.ues
as the quality and quantity of available data, qap. in the data
base for specific chemical., incomplete understandinq of qeneral
bioloqical phenomena, and scientific judgments or science po~icy
positions that were employed to bridqe information qaps.
In short, broad aqreement exist. on the importance of a full
picture of risk, particularly includinq a .tatement of confidence
in the assessment and that the uncertainties are within reason.
This section discusses information content and uncertainty
aspects of risk characterization, while Section 3 discusses
various descriptors used in ri,k characterization.
10
-------
1. The ri.k a..e..ment proce.. call. for characterizing
ri.k a. a coabination of qualitative information, quantitative
. informatioD, and information regarding uncertaintie..
Risk alsessment is based on a series of question. that the
assessor asks about the data and the implications of the data for
human risk.
Each question calls for analysis and interpretation
of the available studies, selection of the data that are most
scientifically reliable and mOlt relevant to the problem at hand,
and scientific conclu.ions reg~rding the que.tion pre.ented.
As
suqqe.ted below, becau.e the que.tion. and analy.e. are complex,
a complete characterization includes several different kinds of
information, carefully selected for reliability and relevance.
a.
Bazard Identification -- What do we know about the
capacity of an environmental aqent for caulinq cancer
(or other adverse effects) in laboratory animals and in
humans?
Bazard identification is a qualitative description based on
factors such as the kind and quality of data on humans or
laboratory animal., the availability of ancillary information
(e.q., structure-activity analy.i., qenetic toxicity, pharmac9-
kinetics) from other .tudies, and the weight-of-the evidence from
all of the.e data .ource..
For example, to develop this
description, the i..ue. addres.ed include:
1.
the nature, reliability, and con.i.tency of the
particular .tudie. in human. and in laboratory animals;
the available information on the machani.tic basis for
activity, and
2.
3.
experimental animal re'pon.e. and their relevance to
human outcome.. .
These is.ue. make clear that the ta.k of hazard
11
-------
identification is characterized by describing the full range of
available information and the implications of that information
for human health.
Dose-Res90nse Assessment -- What do we know about the
biological mechanisms and do.e-response relationships
underlying any effecta ob.erved in the laboratory or
epidemiology studies providing data for the assessment?
The doae-response aaaessment examines quantitative
b.
relationships between exposure (or do.e) and effect. in the
studies used to identify and define effecta of concern.
This
information is later used along with "real world" expo.ure
information (see below) to develop estimates of the likelihood of
adverse effects in populations potentially at risk.
Methods for establishing do.e-respon.e relationships often
depend on variou. a.sumptions used in lieu of a complete data
base and the method chosen can strongly influence the overall
assessment.
This relationship means that careful attention to
the choice of a high-to-low do.e extrapolation procedure is very
important.
As a re.ult, an a.se..or who i. characterizing a.
dose-respon.e relation.hip con.ider. .everal key issue.:
1.
2.
relation.hip between extrapolation model. aelected and
available information on biological mechani...;
how appropriate data .et. were selected from tho.e that
.how the range of pos.ible potencie. both in laboratory
animal. and human.;
3.
ba.i. for .electing inter.peeie. do.e .caling factors
to account for scaling do... from experimental animals
to human.; and
corre.pondence between the expeeted route(.) of
expo.ure and the expo.ure route(.) utilized in the
hazard .tudie., as well a. the interrelation.hips of
potential effects from different expo.ure routes.
4.
12
-------
EPA'. Integrated Risk Information System (IRIS) is a primary
source of this information.
IRIS includes data summaries
-
representing Agency con.ensus on specific chemicals, based on a
careful review of the scientific issues listed above.
For
specific risk a..e.sments ba.ed on data in IRIS and on other
sources, risk as.essor. .hould carefully review the information
presented, emphasizing confidence in the databaee and
uftcertainties (see subsection d below).
The IRIS .tatement of
confidence should be included a. part of the ri.k
characterization for hazard and do.e-respon.e information.
c.
EXDosure Assessment -- What do we know about the path.,
pattern., and maqnitudes of human expo.ure and numbers
of persons likely to be expo.ed?
The exposure a..essment examine. a wide range of expo8ure
parameters pertaining to the "real world" environmental scenarlOS
of people who may be expo.ed to the agent under study.
The data
considered for the expo.ure a..e.sment range from monitoring
studies of chemical concentration. in environmental media, tood,
and other material. to information on activity patterns of
different population .ubgroup..
AD a..es.or who characterizes
exposure .hould addre.. .everal i..ue..
1.
The ba.i. for the value. and input parameter. U8ed tor
each expo.ure scenario. If ba.ed on data, inforaatl0n
on the quality, purpo.e, and repre.entativene.. of the
databa.e i. needed. If ba.ed on a..umption., the
source and qeneral logic used to develop the a.8u.ptlon
(e.9., monitoring, modeling, analoqy, profe..ional
judqment) .hould be de.cribed.
13
-------
2.
3.
The major factor or factors (e.g., concentration body
uptake, duration/frequency of exposure) thought to
acc~unt for the greatest uncertainty in the exposure
.st~ate, due either to sensitivity or lack of data.
The link of the exposure information to the risk
descriptors discussed in Section 3 of this Appendix.
This issue includes the conservatism or non-
conservatism of the scenarios, as indicated by the
choice of descriptors.
In summary, confidence in the information used to
characterize risk is variable, with the result that risk
characterization requires a statement r~q.rdinq the assessor's
confidence in each aspect of the assessment.
d.
Risk Characterization -- What do other assessors,
decision-makers, and the public need to know about the
primary conclusions and assumptions, and about the
balance between confidence and uncertainty in the
assessment?
In the risk characterization, conclusions about hazard and
dose response are inteqrated with those from the exposure
assessment.
In addition, confidence about these conclusions,
includinq information about the uncertainties associated with the
As summarized below, the
final risk summary, is hiqhliqhted.
characterization inteqrates all of the precedinq information to
communicate the overall meaninq of, and confidence in, the
hazard, exposure, and risk conclusions.
Generally, risk a..essmentscarry two cateqorie. of
uncertainty, and each merits con.ideration.
Mea.urement
.uncertainty refer. to the usual variance that accompanies
scientific measurements (such as the ranqe around an expo8ure
estimate) and reflects the accumulated variances around the
individual measured values used to develop the estimate.
A
14
-------
different kind of uncertainty stems from data gaps -- that is,
information needed to complete the data base for the assessment.
Often, the data gap is broad, such as the absence of information
on the effects of exposure to a chemical on humans or on the
biological mechanism of action of an agent.
The degree to which confidence and uncertainty in each of
these areas is addressed depend. largely on the scope of the
I
I
I
I'
,
,
assessment and the resources available.
For example, the Agency
does not expect an assessment to evaluate and as.e.. every
conceivable exposure scenario for every pos.ible pollutant, to
examine all susceptible populations potentially at risk, or to
characterize every possible environmental scenario to determine
the cause and effect relationships between expo.ure to pollutants
and adverse health effects.
Rather, the uncertainty analysis
should reflect the type and complexity of the risk assessment,
with the level of effort for analy.i. and discussion of
uncertainty corre.ponding to the level of effort for the
assessment.
50me .ources of confidence and of uncertainty are
described below.
Often ri.k a..e..ors and manager. .implify di.cussion of
risk i..ue. by .peaking only of the numerical components of an
asses.ment.
That i., they refer to the weight-of-evidence, unit
risk, the ri.k-.pecific dose or the q1. for cancer ri.k, and the
'RfD/RfC for health effects other than cancer, to the exclusion of
other information bearing on the ri.k ca.e.
Bowever, since every
asses.ment carrie. uncertaintie., a simplified numerical
15
-------
presentation of risk is always incomplete and often misleading.
For this reason, the NRC (1) and EPA risk assessment guidelines
(2) call for "characterizing" risk to include qualitative
information, a related numerical risk estimate and a discussion
of uncertainties, limitation., and as.umption..
Qualitative information on methodoloqy, alternative
interpr~tations, and working as.umption. i. an important
component of risk characterization.
For example, specifying that
animal studies rather than human studies were 'used in an
assessment tells others that the ri.k estimate is based on
assumptions about human respon.e to a pa~::ular chemical rather
than human data.
Information that human exposure estimates are
based on the subjects' presence in the vicinity of a chemical
accident rather than tissue measurement. define. known and
unknown aspects of the exposure component of. the study.
Qualitative descriptions of this kind provide crucial
information that augments understanding of numerical risk
estimates.
Uncertainties .uch a. the.e are expected in
scientific studies and in any risk a..essment based on these
studies.
Such uncertainties do not reduce the validity of the
assessment.
Rather, they are highlighted along with other
important ri.k a.....ment conclu.ion. to inform other. fully on
the results of the assessment.
16
-------
,----
~. Well-balanced ri.k characterization pr..ent. info~tion
for oth.r ri.k a.....or., EPA deci.ion-maker., and the public
regarding the .trength. and li.itation. of the a.......nt.
The risk assessment process calls for identifying and
highlighting significant risk conclusions and related
uncertainties partly to aS8ure full communication among risk
assessors and partly to assure that decision-makers are fully
informed.
Issues are identified by acknowledging noteworthy
qualitative and quantitative factors that make a difference in
the overall assessment of hazard and risk, and hence in the
ultimate regulatory decision.
The key word i. "noteworthy":
information that
significantly influences the analy.is is retained -- that is,
noted -- in all future presentations of the risk assessment and
in the related decision.
Uncertainties and assumptions that
strongly influence confidence in the risk estimate require
special attention.
As discussed earlier, two major sourc.s of uncertainty are
ve-d.ability in the factors up~n '4hich .stimat". are based and the
existence of fundamental data gaps.
This distinction is relevant
for some aspects of the risk characterization.
Por example, the
central t.ndency and high end individual exposure estimates are
intended to capture the variability in exposure, lifestyles, and
other factors that lead to . distribution of risk acro.s a
population.
ley consideration. underlying the.e risk estimate.
should be fully described.
In contrast, scientific assumDtions
are used to bridge knowledge gaps .uch .s the us. of scaling or
17
-------
extrapolation factors and the use of a particular upper
confidence limit around a dose-response estimate.
Such
assumptions need to be discussed separately, along with the
implications of using alternative assumptions.
For users of the assessment and others who rely on the
assessment, numerical estimates should never be separated from
the descriptive informatinn that is integral to risk
char~cterization.
All d~uments and presentations should,include
both; in short reports, this information is abbreviated but never
omitted.
For decision-makers, a complete characterization (key
descriptive elements alonq with numerical estimates) should be
retained in all discussions and papers relatinq to an assessment
used in decision-makinq.
Fully visible information assures ~~at
important features of the assessment are immediately availab:e at
each level of decision-making for evaluatinq whether risk. are
acceptable or unrea.onable.
In short, difference. in assumpt~~n.
;~ unc-.rtaintie., coupled wi~~ non-scientific cJnsider-tlon.
called for in various environmental statute., can clearly l.ad ~o
different risk manaqement deei.ion. in ea.e. with os~en.ibly
identical quantitative ri.k.~
determine the deci.ion.
i.e.,
the -number" alone do.. ~ot
Con.ideration of alternative approaehe. involve. ex~nl~q
selected plausible option. for addr.s.inq a qiven uncertalnty.
1
The key word. are -selected" and -plau.ible~"
listing al~
options, regardle.. of their merit. would be .uperfluoul.
18
-------
Generators of the assessment should outline the strengths and
weaknes.e. of each alternative approach and as appropriate,
estimate. of central tendency and variability (e.g., mean,
percentiles, range, variance.)
Describing the option cho.en involve. several statements.
1.
2.
A rationale for the choice.
Effects of option selected on the assessment.
Comparison with other plau.ible options.
3.
4.
Potential impact. of new re.earch (on-going,
potential near-te~and/or long-term studies).
For users of the assessment, giving attention to uncertainties in
all decisions and discussions involving the a.sessment, and
preserving the statement of confidence in all presentations is
important.
For decision-makers, understanding the effect of the
uncertainties on the overall asse..ment and explaining the
influence of the uncertainties on the requlatory
decision.
19
-------
SBCTIO. 3.
EXPOSURZ ASSESSMENT AND RISK DESCRIPTORS
The results of a risk assessment are usually communicated to
the risk manager in the risk characterization portion of the
assessment.
This communication i. often accomplished through
risk descrictors which convey information and answer questions
about risk, each descriptor providing different information and
insights.
Exposure assessment play. a key role in developing
these risk descriptors, since each descriptur is based in part on
the exposure distribution within the population of interest.
The
Risk Assessment Council (RAC) has been di.cus.ing the use of risk
descriptors from time to time over the pa.t two years.
The recent RAC efforts have laid the foundation for the
discussion to follow.
Fir.t, as a r..ult of a discussion paper
on the comparability of risk assessment. acro.. the Agency
programs! theRAC discussed how the program presentations of risk
led to ambiguity when. risk assessments were compared across
programs.
Because different asses.ments presented different
descriptors of ri.k without alway. making clear what was being
described, the RAC discus.ed the advisability of u.ing separate
descriptor. for population risk, individual ri.k, and
identif~c.tion of ..n.itive or highly expo..d population
seqment..
Th. RAC al.o discus..d the n..d for con.i.tency across
programs and the advisability of requiring risk a..essments to
provide rough~y comparable information to risk managers and the
public through the u.. of a consistent s.t of risk descriptors.
20
-------
The following guidance outlines the different descriptors in
a convenient order that should not be construed .s a hierarchy of
importance.
These descriptors should be used to describe risk in
a variety of ways for a given assessment, consistent with the
assessment's purpose, the data available, and the information the
risk manager needs.
Use of a range of descriptors instead of a
single descr:.ptor enables Agency programs to present a picture of
risk that corresponds to the range of different exposure
,conditions encountered for mo.t environmental chemicals.
This
analysis, in turn, allows =~sk managers to identify populations
at greater and lesser risk and to shape regulatory solutions
accordingly.
EPA risk assessment. will be expected to addre.. or provide
descriptions of (1) individual ri.k to include the central
tendency and high end portions of the risk distribution,
(2) important subgroups of the population such as highly exposed
or highly susceptible group. or individual., if known, and
( 3) popu.j,ation ri.k.
A.sessor. may also u.e addi~_onal
descriptor. of ri.k a. n.eded when the.e add to the clarity of
the presentation.
With the exception of a..e..ment. where
particular de.criptor. clearly do not apply, .088 form of these
three type. of de.criptors .hould be routinely developed and
presented for EPA ri.k a.se.sment..
Furthermore, presenters of
risk assessment information should be prepared to routinely
answer question. by ri.k managers concerning the.e de.criptors.
21
-------
rt is essential that presenters not only communicate the
results of the assessment by addressing each of the descriptors
where appropriate, but they a180 communicate their confidence
that these results portray a reasonable picture of the actual or
projected exposures.
This task will u.ually be accomplished by
highlighting the key assumption. and parameters that have the
greatest impact on the results, the basis or rationale for
choosing these assumptions/parameters, and the consequences of
choosing other assumptions.
In order for the risk assessor to successfully develop and
present the various risk descriptor., the exposure assessment
must provide exposure and dose information in a form that can be
combined with exposure-response or dose-response relationships to
estimate risk.
Althouqh there will be differences amonq
individuals within a population a. to absorption, intake rates,
susceptibility, and other variables such that a high exposure
does not necessarily result in a hiqh do.e or risk, a moderate or
h'1r.ly Dositive correlation ~or7 expo.ure, do.e, and risk is
assumed in the followinq discussion.
Since the generation of all
descriptors i. Dot appropriate in all ri.k a..e.amenta and the
type of d8ecriptor tran.latea fairly directly into the type of
analysis that the expo.ure assesaor mu.t perform, the exposure
assessor need. to b8 aware of the ultimAteqoal. of the
assessment.
The followinq section. discu.. what type of
information ia nece..ary.
22
-------
1. Information about individual .xpo.ur. and ri.k i.
important to c088unicatiD9 the r..ult. of a ri.k a"...m.Dt.
Individual risk descriptors are intended to address
questions dealing with risks borne by individuals within a
population.
These questions can take the form of:
Who are the people at the highest ri.k?
What risk level. are they .ubjected to?
What are they doing, where do they live, etc., that
might be putting them at this higher risk?
What is the ave=~ge risk for individuals in the
population of interest?
:-:-.':' " "- : - 0-
~ :::. I' :: ~ ~ ::- -= _:5 ~ ~: 3:':"~. ~.: ~ :.. ::-.. is,
conceptually,
above the 90th percentile of the actual (either measured or
estimated) distribution.
This conceptual range is not meant to
precisely define the limits of this descriptor, but should be
used by the assessor a. a target range for characterizing '.h~gh
end risk".
Bounding estimate. and worst ca.e scenario.1 shou:d
not be termed high end risk e.timate..
The high end ri.k de.criptor i. a plau.ible
estimate of the individual ri.k for tho.e
person. at the upper end of the ri.k
distribution. The intent of this de.criptor
i. to convey an e.timate of ri.k in the
upper range of the distribution, but to
avoid e.timate. which are b.yond the
)
1 Bi9h end e.timate. focus on e.timate. of the expo.ur. or
dose in the actual population.. -Boundin9 ..timate.,- on tn.
other hand, purpo.ely ov~estimate the expo.ure or do.e in an
actual population for the purpo.e of developin9 a .tat.ment tnat
the risk i. "not greater than....- A -~or.t ca.. .cenario.
refers to a combination of event. and condition. .uch that. ta~.n
together, produce. the highe.t conceivable ri.k. Althouqh 1t ~.
po.sible that .uch an expo.ure, do.., or sen.itivity combl~at1~n
might occur in a given population of intere.t, the probabl~~ty :t
a~ individual receivin9 this combination of event. an~. cond~~~:~~
is u.ually .mall, and often .0 .mall that .uch a comb1n.t~:n w...
not occur in a particular, actual population.
23
-------
r
true distribution. Conceptually, high
end risk means risks above about the
9?th percentile of the population
d1stribution, but not higher than the
individual in the population who has
the highest risk.
This descriptor is intended to estimate the risks that are
expected to occur in small but definable -high end~ segments of
the subject population.
The individual. with these risks may be
members of a special population segment or individuals in the
general population who are highly exposed because of the inherent
stochastic nature of the factors which give rise to exposure.
Where no particular difference in sensitivity can be identified
within the population, the high end risk will be related to the
high end exposure or dose.
In those few cases where the complete data on the population
distributions of exposures and dose. are available, high end
exposure or dose estimates can be represented by reporting
exposures or doses at selected percentiles of the distributions,
such as the 90th, 95th, or 98th percentile.
Sigh end exp08U4es
.
or doses, as appropriate, can then be u.ed to calculate high end
risk estimate..
In the majority of ca... where the complete distributions
are not ...ilable, .everal method. help e.timate a high end
exposure or do.e.
If sufficient information about the
.variability in life.tyle. and other factor. are available to
simulate the distribution through the u.e of appropriate
modeling,. S.g., Monte Carlo simulation, the ..timate from the
simulated distribution may b. used.
A. in the method above, the
risk manager should b. told where in the high end range the
24
-------
estimate i. beinq made by statinq the percentile or the number of
persons above this estimate.
The assessor and risk manaqer
-
should be aware, however, that unless a qreat deal is known about
exposures and doses at the hiqh end of the distribution, these
estimates will involve considerable uncertainty which the
exposure a.se.sor will need to describe.
If only limited information on the distribut~~on of the
exposure or dose factors is available, the assessor .hould
approach estimatinq the hiqh end by identifyinq the most
sensitive parameters and u.inq maximum or near-maximum values for
one or a few of these variables, leavinq others at their mean
values2.
In doinq this, the exposure a..e.sor needs to avoid
combinations of parameter value. that are inconsistent, e.q., low
body weiqht used in combination with hiqh intake rate., and must
keep in mind the ultimate objective of beinq within the
distribution of actual expected exposure. and dose., and not
beyond it.
I~
lmost no data are av..._l4ble on the ranqes £0::' t.ne
various parameters, it will be difficult to estimate exposures or
doses in the hiqh end with much confidence, and to develop the
hiqh end ri.k e.timate.
One method that ha. been used in these
cas.s i. to start with a boundinq .~timate and "back off" the
limits u.ed until the combination of parameter values is, in the
2 Haximizinq all variables will in virtually all case.
re.ult in an ..timate that is above the actual values seen in the
~opulation. When the principal parameters of the dose equation
. (e.q., concentration, intake rate, duration) are broken out into
subcomponents, it may be necessary to use maximum value. for more
than two of these subcomponent param.ters, dependinq on a
sensitivity analysi..
2S
-------
judqment of the assessor, clearly within the distribution of
expected exposure, and still lies within the upper 10' of persons
exposed.
Obviously, this method results in a large uncertainty
and requires explanation.
The risk descriptor addressing central
tendency may be either the arithmetic
mean risk (Averaqe Estimate) or the
median risk (Median Estimate), either
of which should be clearly la::.'ed.
Where both the arithmetic mean 1nd
the medi~n are available but th~y
differ substantially, it is helpful
to present both.
T~e Average Estimate, used to approximate the arithmetic
mean, can be derived by usinq averaqe values for all the exposure
factors.
It does not necessarily represent a particular
individual on the distribution.
The Averaqe Estimate is not very
meaninqful when exposure across a population varies by several
orders of magnitude or when the populatio~ has been truncated,
e.g., at some prescribed distance from a point source.
Becaus. of the skewness of typical exposure profiles, the
arit~~:tic ~ean is 'not nec6.sarily a qood .nQic~. ~r of ~he
midpoint (median, 50th percentile) of a distribution.
A Median
Estimate, e.q., qeometric'mean, is usually a valuable descriptor
for thi. type of distribution, since half the population will b.
above and half below this value.
26
-------
2'. IDfoC1Ution about population expo.ure lead. to another
i.port&D~ waf to de.cribe ri.k.
population risk refers to an assessment of the extent of
harm for the population a. a whole.
In theory, it can be
calculated by summing the individual ri.k. for all individuals
within the .ubject population.
Thi. ta.k, of cour.e, require. a
great deal more information than i. normally, if ever, available.
Some questions addressed by de.criptor. of population risk
include:
Bow many cases of a particular health effect might be
probabilistically e.timated in thi. population for a
specific time period?
For noncarcinogens, what
within a specified range
exceedance of the RfD (a
concentration), or other
portion of the population are
of .088 benchmark level, e.g.,
do.e), the RfC (a
health concern level?
For carcinogen., how many per.on. are above a certain
risk level such a. 10-' or a series of ri.k levels such
a. 10-5, 10-., etc?
Answering the.e question. require. .0.. knowledge of the
exposure frequency"di.tribution in the population.
In
particular, addre..ing the .econd and third que.tion. may require
graphing the ri.k di.tribution.
The.e que.tion. can lead to two
different de.cr~ptor. of population ri.k.
The fir.t de.criptor i. the probabili.tic
number of health effect ca.e. e.timated
in the population of intere.t oVer a
specified time period.
Thi. de.criptor can be obtained either byea) .umming the
individual ri.k. over all the individual. in the population when
such information i. available, or (b) through the u.e of a risk
model such a. carcinogenic model. or procedure. which a..ume a
21
-------
linear non-threshold response to exposure.
If risk varies
linearly with exposure, knowinq the mean risk and the population
size can lead to an estimate of the extent of harm for the
population as a whole, excluding sensitive subgroups for which a
different dose-response curve need. to be u.ed.
Obviously, the more information one ha., the more certain
the estu.ate of this risk descriptor, but inherent uncertainties
in risk assessment methodoloqy place limitations on the accuracy
of the estimate.
With the current state of the science, explicit
steps should be taken to assure that this descriptor is not
confused with an actuarial prediction 'of ca.es in the population
(which is a statistical prediction based on a great deal of
empirical data).
Although estimating population risk by calculatinq a mea~
individua,l risk and multiplying by the population size i.
sometimes appropriate for carcinogen a..essments using lin.ar.
non-threshold model.J, this is not appropriate for non-
carcinogenic effect. or for other type. of cancer models.
F~r
non-linear cancer model., an e.timate of population risk sust be
calculated by sumain9 individual ri.k..
Por non-cancer effects,
we generally have not developed the ri.k a..e..ment techniq~.s to
the point of knowin9 how to add ri.k probabilitie., so . 8econd
,descriptor, below, i. more appropriate.
Another de.criptor of population ri.k
is an e.timAte of the percentage of
the population, or the number of
person., above a specified level of
3 Certain important caution. apply. The.e cautions ar. acr.
explicitly spelled out in the Agency' Guideline. for Expo8~r.
Assessment, tentatively scheduled to )1 publi.hed in lat. :~f~
28
-------
"risk or within a specified range of
some benchmark level, e.g., exceedance
of the RfD or the RfC, LOAEL, or other
specific level of interest.
This descriptor must be obtained through measuring or simulating
the population distribution.
I .
29
-------
3. IAforaatioD about the di.tributioD of .Xpo.ure and ri.k
for diffe~Dt 8ubQrOuD. of the population are i.portant
CO.pOD.D~8 of a ri.k a.......nt.
A risk manaqer miqht also ask questions about the
distribution of the risk burden amonq various seqments of the
subject population such as the followinq:
Bow do exposure and risk impact various subqroups?
What is the populstion risk ot a particular sub9roup?
Questions about the distribution. of expo.ure and risk amonq such
population seqments require additional risk descriptors.
Biqhly exposed subqroups can be
identified, and where possibl., characterized
and the maqnitude of risk quantified.
This de.criptor is useful when there
is (or is expected to be) a subqroup
experiencinq siqnificantly ditfer.nt
exposures or doses from that of the
larqer population.
These subpopulations may be identified by aqe, sex, life-
style, economic factors, or other demoqraphic variables. For
example, toddlers who play in contaminated soil and certain hiqh
fish consumers represent subpopulations that may have qreater
e....tl0.\l: .. -: to certain aqent...
liqhly sU8ceptible subqroup. can al.o
be identified, and if po..ible,
characterized and the maqnitude of
risk quantified. This descriptor i.
u.eful when the sensitivity or
8u8ceptibility to the effect tor
.pecific subqroups is (0= i.
expected to be) siqniticantly
different from that of the larqer
population. In order to calculate
risk for these subqroups, it will
sometimes be necessary to use a
.ditterent do.e-respon.. relation.hip.
30
-------
For ex.-pl., upon exposure to a chemical, pregnant women, elderly
people, children, and people with certain illnesses may each be
more sen.itive than the population a. a whole.
Generally, selection of the population seqment. i. a matter
of either a priori intere.t in the .ubgroup, in which ca.e the
ri.k as.e..or and ri.k manager can jointly agree on which
subgroup. to highlight, or a matter of di.covery of a .en.itive
or highly exposed subgroup during the a.....ment proce...
In
either ca.e, once identified, the .ubgroup can be treated a. a
population in it.elf, and characterized the .... way a. the
larger population using the de.criptor. for population and
individual ri.k.
31
-------
.
4. Situation-apecific information adda perapective on
poaaible future event. or regulatory optiona.
These poatulated questions are normally designed to answer
"what if" questions, which are either directed at low probability
but possibly high consequence event. or are intended to examine
candidate risk management optiona.
. following forma
Such queationa might take the
What if a peaticide applicator applies
this pe.ticide without u.ing protective
equipment?
What if this site become. re.idential
in the future?
What risk level will occur if we .et
the standard at 100 ppb?
The assumption. made in an.wering the.e po.tulated questiona
should not be confu.ed with the as.umption. made in developing a
baseline estimate of exposure or with the adjustment. in
parameter values made in performing a aen.itivity analysis.
The
answers to the.. po.tulated que.tion. do not give information
about how likely the c~mbination of value. might be in the actual
population or about how many (if &DY) per.on. might be aubjected
to the calculated expo.ure or ri.k in the real world.
A calculation of risk baaed on .pecific
hypothetical or actual combination.
of factor. po.tulated within the
expo.ure a..e..ment can al.o be .
u.eful a. a ri.k de.criptor. It
i. often valuable to a.k and an8W8r
specific que.tion. of the -what if-
nature to add perapective to the
risk a..e..ment.
The only information the an.wer. to the.e que.tion. convey
.is that if condition. A, S, and Care a..umed, then the resulting
e~poaure or riak will be X, Y, or I, reapectively.
The values
32
-------
for x, Y, and Z are u.ually fairly straightforward to calculate
and can be expressed &S point estimat.. or rang...
-
Each asses.ment may have none, one, or several of the.. types of
descriptors.
The answers 40 Dot directly give information about
how likely that combination of value. might be in the actual
population, so there are some limit. to the applicability of
these descript~r..
"
. ,
33
-------
.-- --- ~-~_.__._--~._--_. -.
2.
RefereDce.
1.
National Research Council. Risk Assessment in the Federal
Government: Managing the Process. 1983
U.s. EPA. Risk Assessment and Manaqement:
Decision Making. 1984.
Framework for
3.
U.S. !PA. Ri.k A.ses.ment Guideline..
33992-34054. September 24, 1986.
51 rederal Register
4.
Presentation of Risk A.ses.ment of Carcinogen.; Ad Hoc St~dy
Group on Risk Assessment Pre.entation. American Industri~.l
Health Council. 1989.
34
------- |