United States Science Advisory EPA-SAB-RAC-ADV-99-002
Environmental Board (1400) December 1999
Protection Agency Washington, DC www.epa.gov/sab
&EPA AN SAB ADVISORY: REVIEW
OF HEALTH RISKS FROM
LOW-LEVEL ENVIRONMENTAL
EXPOSURES TO
RADIONUCLIDES
REVIEW OF THE OFFICE OF RADIATION
AND INDOOR AIR'S FEDERAL
GUIDANCE REPORT 13 - PART 1,
INTERIM VERSION (FGR-13)
-------
December 23, 1998
EPA-SAB-RAC-ADV-99-002
Honorable Carol M. Browner
Administrator
U.S. Environmental Protection Agency
401 M. Street, SW
Washington, DC 20460
Re: An SAB Advisory on the Health Risks from Low-Level Exposure to
Radionuclides, Federal Guidance Report No. 13 - Part 1, Interim Version (FOR 13
- Part 1)
Dear Ms. Browner:
The accompanying report was developed by the FGR 13 Review Subcommittee of the
Radiation Advisory Committee (RAC) of the Science Advisory Board (SAB) in response to a
request from the Office of Radiation and Indoor Air (ORIA) to review the technical aspects of the
interim version of Federal Guidance Report 13 - Part 1, dated January, 1998. The Subcommittee
met on May 6-7, 1998, and held two subsequent conference calls, to review the interim document
with respect to the following Charge:
a) Is the methodology employed for calculating health risks from radionuclide intakes
and external exposures acceptable?
b) In light of scientific information, have the major uncertainties been identified and
put into proper perspective?
c) Is the proposed method for extending the list of radionuclides to include all those
tabulated in Federal Guidance Reports 11 and 12 reasonable?
FGR 13 - Part 1 provides tabulations of unit risk estimates, or "risk coefficients", for
cancer morbidity and mortality attributable to exposure to any of approximately 100 radionuclides
via various environmental media. Radiation doses are calculated as an intermediate step in
estimating risks, but are not tabulated in the report. The risk coefficients apply to populations
that approximate the age, gender, and mortality experience characterized by the current U.S.
population. The report is intended by the Agency to promote consistency in assessments of the
risks to health from radiation and to help ensure that radiological risk assessments are based on
sound scientific information.
-------
In our enclosed report, we address these issues and offer a few suggestions in addition to
our response to the Charge. The report focuses on specific technical issues that the
Subcommittee believes should be addressed prior to publication of the final version of FGR 13 -
Part 1 and makes suggestions for future consideration by the Agency. The following findings and
recommendations are viewed as particularly important with respect to each Charge element:
a) Methodology
1) The methodology employed for calculating health risks from radionuclide
intakes and external exposures was found to be acceptable in the context
of: 1) current Agency science policy choices, such as the applicability of a
linear, no-threshold dose-response model for cancer risks at low doses of
radiation; and 2) its intended use to support broad federal radiation
protection programs such as environmental impact statements or generic
rulemaking. However, the Subcommittee is concerned about reliance on
unpublished reports for some of the detailed techniques employed, absence
of absorbed dose rate information, and other points discussed below.
2) The lack of published reports for some of the techniques employed is a
weakness of this work, for both the internal and external exposure
calculations. The Subcommittee recommends that evidence of verification
and quality assurance procedures be included in an Appendix to the final
version of FGR 13 -Parti.
3) The Subcommittee is concerned about the absence of radiation dose
information in FGR 13 - Part 1. This makes it difficult for users to update
the risk coefficients as new information on risk or dosimetry becomes
available, or to employ risk models other than the linear-no threshold
model used by the Agency. The Subcommittee recommends that dose
information be published in electronic form.
4) Some potentially valuable information has not been addressed in FGR 13.
For example, a published dosimetry model for the esophagus and trachea is
not used for the internal dose calculations or discussed in the document.
Risks calculated in FGR 13 - Part 1 are based on mortality (death) data, but
information on cancer incidence (morbidity) is generally more accurate.
-------
b) Uncertainty
1) Although the document does well in identifying many of the major
uncertainties and describing them qualitatively, it does not provide the
reader with a good sense of the overall uncertainty entailed when a specific
risk coefficient is used to predict the risks of low-level exposure to
radiation. The Subcommittee recommends at a minimum that: 1) the
individual radionuclides of highest concern for environmental exposures be
placed into categories associated with low, moderate, high, or very high
uncertainty; 2) identification of these uncertainty categories be contained
within the tables of risk estimates; 3) general guidance about the limits of
application be provided for those radionuclides that cannot be placed into
an uncertainty category, and; 4) the final version of FGR 13 - Part 1
include an appendix that describes the steps that may be taken to improve
uncertainty estimates for the risk coefficients.
2) For future revisions of the Agency's reports, such as those in the FGR 13
series, the Subcommittee recommends that quantification of uncertainty be
considered at the outset of the project.
3) Although the lack of scientific consensus on the estimation of lifetime
cancer risks from low-level radiation is acknowledged in FGR 13 - Part 1,
the Subcommittee recommends adding a concise, balanced discussion that
lays out the arguments for and against the possibility of a threshold or other
non-linear behavior at low dose and low dose rate.
c) Expansion
1) The Subcommittee believes that the Agency's proposed method for
extending the list of radionuclides to include those tabulated in FGRs 11
and 12 is reasonable, but recommends that risk coefficients also be
presented for the inhalation of radon and its short-lived decay products.
The Subcommittee would prefer that these risk coefficients be included in
the final version of Part 1 if feasible.
The Subcommittee also identified and addressed several issues not raised in the Charge.
These issues (and where appropriate) recommendations are listed below:
a) The Subcommittee observes that use of the FGR 13 - Part 1 risk coefficients could
contribute to debate or ambiguity as to whether a regulated site meets regulatory
limits, especially in cases where a regulation is stated in terms of dose limits but the
supporting technical documentation might use risk calculations.
-------
b) The Subcommittee recommends that risk coefficients for external exposure,
applicable specifically to the Uranium Mill Tailings Radiation Control Act
(UMTRCA) standard, be calculated and tabulated in the final version of FGR 13 -
Part 1 or in a separate document.
c) The Subcommittee recommends that the title of the document be changed to
"Estimated Health Risks from Low-Level Environmental Exposure to
Radionuclides, Federal Guidance Report 13 - Part 1: Cancer" (emphasis added).
This change conveys the important points that the tabulated risk values are
estimates and that cancer is the only health effect treated in the document.
We commend the Agency on its leadership and efforts in using up-to-date scientific
methods and data to develop the health risk estimates in FGR 13 - Part 1. We believe that our
recommendations, except those noted in the report for future consideration, can be incorporated
prior to publication of the final version and will strengthen the report and its credibility. We
strongly support the Agency's stated intent to publish supporting information in an electronic
form to accompany release of the final version of FGR 13 - Part 1, and we recommend that it
include the data, models, and dose information used in formulating the risk coefficients.
The RAC and its Federal Guidance Report Review Subcommittee (FGRRS) appreciate
the opportunity to provide this report to you and we hope that it will be helpful. We look forward
to your response to this report in general and to the specific comments and recommendations in
this letter in particular.
Sincerely,
/signed/ /signed/
Dr. Joan M. Daisey, Chair Dr. Stephen L. Brown, Chair
Science Advisory Board Radiation Advisory Committee
Science Advisory Board
/signed/
Dr. Thomas F. Gesell, Chair
Federal Guidance Report Review Subcommittee
Radiation Advisory Committee
-------
NOTICE
This report has been written as part of the activities of the Science Advisory Board (SAB),
a public advisory group providing extramural scientific information and advice to the
Administrator and other officials of the Environmental Protection Agency (EPA). The Board is
structured to provide balanced, expert assessment of scientific matters related to problems facing
the Agency. This report has not been reviewed for approval by the Agency and, hence, the
contents of this report do not necessarily represent the views and policies of the EPA nor of other
agencies in the Executive Branch of the Federal government. In addition, the mention of trade
names or commercial products does not constitute a recommendation for use.
-------
ABSTRACT
On May 6-7, 1998, the Federal Guidance Report Review Subcommittee
(FGRRS) reviewed technical aspects of the draft document, "Health Risks from Low-Level
Environmental Exposure to Radionuclides," Federal Guidance Report 13 - Part 1 - Interim
Version (FGR 13 - Part 1). This document provides tabulations of unit risk coefficients for
cancer morbidity and mortality attributable to exposure to approximately 100 radionuclides
through various environmental media, in a population approximated by the age, gender, and
mortality experienced in the United States.
The Subcommittee found the report to be well organized and well written and to have
used up-to-date scientific methods and data to determine the health risk estimates. Although most
of the important limitations of the risk estimates are noted in FGR 13 - Part 1, they are not
sufficiently prominent in the current draft, given the potential for misuse or misinterpretation of
the estimates. In particular, the magnitudes of the uncertainties in the computed numbers are
difficult to ascertain. Other concerns included partial reliance on unpublished methodologies, lack
of dose information, insufficient discussion of alternatives to the linear, no-threshold risk model,
and several other technical issues. The Subcommittee found that the Agency's plan to calculate
risk coefficients for an extended list of radionuclides was appropriate, except that radon and its
decay products should also be included. The Subcommittee strongly supports the Agency's
stated intent to publish supporting information in electronic form to accompany release of the final
version of FGR 13 - Part 1, and recommends that it include the data, models, and dose values
used in formulating the risk coefficients.
KEYWORDS: radionuclides; cancer; risk assessment
-------
U.S. ENVIRONMENTAL PROTECTION AGENCY
SCIENCE ADVISORY BOARD
FEDERAL GUIDANCE REPORT REVIEW SUBCOMMITTEE
of the RADIATION ADVISORY COMMITTEE
ROSTER
Chair
Dr. Thomas F. Gesell, Idaho State University, Pocatello, ID
Members
Dr. Stephen L. Brown, Risks of Radiation & Chemical Compounds, Oakland, CA
Dr. Raymond A. Guilmette, Lovelace Respiratory Research Institute, Albuquerque, NM
Dr. F. Owen Hoffman, SENES Oak Ridge, Inc., Oak Ridge, TN
Dr. Janet Johnson, Shepherd Miller, Inc., Ft. Collins, CO
Dr. Ellen Mangione, Colorado Department of Public Health and Environment, Denver, CO.
Dr. Paul J. Merges, New York State Department of Environmental Conservation, Albany, NY
Dr. John W. Poston, Sr., Texas A&M University, College Station, TX
Dr. Genevieve S. Roessler, Radiation Consultant, Elysian, MN
Dr. James E. Watson, Jr., University of North Carolina at Chapel Hill, Chapel Hill, NC
Dr. Arthur Upton, Robert Wood Johnson Medical School, Piscataway, NJ
Science Advisory Board Staff
Dr. K. Jack Kooyoomjian, Designated Federal Officer, Science Advisory Board (1400), U.S.
EPA, 401 M Street, S.W.,Washington, DC 20460
Mr. Samuel Rondberg, Designated Federal Officer, Science Advisory Board (1400), U.S. EPA,
401 M Street, S.W.,Washington, DC 204601
Ms. Diana L. Pozun, Management Assistant, Science Advisory Board (1400), U.S. EPA, 401 M
Street, S.W.,Washington, DC 20460
TABLE OF CONTENTS
Provided editorial support for this report, but did not participate in the review.
Ill
-------
1. EXECUTIVE SUMMARY 1
2. INTRODUCTION 7
2.1 Background 7
2.2 Charge 8
3. DETAILED FINDINGS 9
3.1 Assessing Methodology for Calculating Health Risks - Charge Element (a) 9
3.1.1 Overall Findings 9
3.1.2 Use of Unpublished Methods 9
3.1.3 Check of Numerical Results 10
3.1.4 Absence of Dose Information 10
3.1.5 Information Not Addressed in FOR 13 - Part 1 11
3.1.6 Use of Mortality Data vs Morbidity 12
3.1.7 Absorption Types for Inhaled Radionuclides 13
3.1.8 Assumption of Full-time, Unshielded Exposure to Radionuclides
in Soil 13
3.1.9 Soil Ingestion 13
3.1.10 Appropriateness of Diet Averaging 14
3.2 Assessing the Treatment of Uncertainty - Charge (b) 14
3.2.1 Overall Findings 14
3.2.6 Comprehensive Uncertainty Calculation for Classes of Biokinetic
Models 18
3.3 Assessing the Method for Extending List of Radionuclides - Charge (c) ... 19
3.3.1 Overall Findings 19
3.4 Additional Topics Identified by the Subcommittee 19
3.4.1 Recommended Change in Report Title 20
3.4.2 Contribution of FOR 13 - Part 1 to Regulatory Debate and Ambiguity . . 20
3.4.3 Domain of Applicability 20
3.4.4 Clarification of the Risk Tables 21
3.4.5 Tabulation of Predominant Cancer Types 21
3.4.6 Adequacy of Soil Geometries for Use with the Mill Tailings Standards . . 22
3.4.7 Relationship Between Exposure and Dose 22
3.5 General Conclusions 22
APPENDIX A - DETAILED TECHNICAL COMMENTS A-l
APPENDIX B - ACRONYMS B-l
REFERENCES R-l
1. EXECUTIVE SUMMARY
-------
As part of the EPA Science Advisory Board's (SAB) Radiation Advisory Committee
(RAC), the FOR 13 Review Subcommittee (the Subcommittee) reviewed technical aspects of
"Health Risks from Low-Level Environmental Exposure to Radionuclides," Federal Guidance
Report 13 - Part 1 - Interim Version (FGR 13 - Part 1). This first report in the series provides
tabulations of risk estimates, or "risk coefficients," for cancer morbidity and mortality attributable
to exposure to approximately 100 radionuclides through various environmental media. Radiation
doses are calculated as an intermediate step in estimating risks, but are not tabulated in the report.
The risk coefficients apply to populations that approximate the age, gender, and mortality
experience characterized by the 1989-91 U.S. decennial life tables. FGR 13 - Part 1 is intended
by the Agency to promote consistency in assessments of the risks to health from radiation and to
help ensure that radiological risk assessments are based on sound scientific information.
The Subcommittee was charged by the Agency's Office of Radiation and Indoor Air
(ORIA) to focus its review on the following questions:
a) Is the methodology employed for calculating health risks from radionuclide intakes
and external exposures acceptable?
b) In light of scientific information, have the major uncertainties been identified and
put into proper perspective?
c) Is the proposed method for extending the list of radionuclides to include all those
tabulated in Federal Guidance Reports 11 and 12 reasonable?
The Subcommittee's report addresses these questions and offers some suggestions beyond
the charge. FGR 13, Part 1 is a useful addition to the complement of Federal Guidance reports
relevant to the evaluation of the risks of radiation and makes important improvements in the
calculation of radiation risks. The report is well organized and well written, although difficult to
comprehend for all but highly technical audiences. The careful reader will note that most of the
important caveats to the use of the document's results are included in the text, but not as
prominently as they probably should be. In particular, the magnitudes of the uncertainties in the
computed numbers are difficult to ascertain.
This review focuses on specific technical issues that the Subcommittee believes should be
addressed prior to publication of the final version of FGR 13 - Part 1 and makes suggestions for
future consideration by the Agency. These major findings and recommendations are summarized
below (with references to the corresponding sections of the report appearing in parentheses).
Charge element a) addressed the methodology employed for calculating health risks from
radionuclide intakes and external exposures. The proposed approach was found to be acceptable
in the context of a) current Agency science policy choices, such as the applicability of a linear, no-
threshold dose-response model for cancer risks at low doses of radiation, and b) its intended use
to support broad Federal radiation protection programs such as environmental impact statements
-------
or generic rulemaking. However, as discussed below, the Subcommittee is concerned about
unpublished reports being cited for some of the detailed techniques employed, absence of dose
information, and the incomplete treatment of uncertainty. (Section 3.1.1)
The lack of published reports for some of the techniques employed is a weakness of FGR
13 - Part 1, for both the internal and external exposure calculations. In particular, because dose
calculation (DCAL) software is such an integral part of FGR 13 - Part 1, the Subcommittee
recommends that the DCAL software and a users manual be published in electronic form prior to,
or concurrently with, the final version of FGR 13 - Part 1. (Section 3.1.2)
The Subcommittee considered whether or not it would be possible to check the numerical
results of the risk coefficients in some way. It concluded that a check of the risk estimates, or
merely a verification of the coding of the model equations, would be a formidable task outside the
scope and level of resources available to the Subcommittee. The Subcommittee recommends that
evidence of verification and quality assurance procedures be included in an appendix to the final
version of FGR 13 - Part 1, or be made available in a separate document. (Section 3.1.3)
The Subcommittee is concerned about the absence of radiation dose information in FGR
13 - Part 1. While the direct tabulation of risk estimates may provide a useful shortcut for the
preparation of rulemaking and environmental documents, it may make the report difficult to
follow by those who think of radiation safety in terms of dose and will make it difficult for users
to update the risk coefficients as new information on risk or dosimetry becomes available. This
approach also makes it difficult for users to employ risk models other than the linear-no threshold
model used by the Agency. The Subcommittee recommends that dose information and
corresponding uncertainties be published in electronic form. Alternatively the authors could
identify another source of this information, such as the anticipated publication of dose information
in electronic form by the International Commission for Radiation Protection (ICRP). (Section
3.1.4)
FGR 13 - Part 1 does not use or address some potentially valuable information not
already incorporated into the ICRP models, such as an internal dosimetry model suggested for the
esophagus and trachea. Also, recent calculations of the absorbed energy in the gastrointestinal
tract could replace the very conservative assumptions used in ICRP Publication 30 and FGR 13 -
Part 1. (Section 3.1.5)
Risks calculated in FGR 13 - Part I are based on mortality data. However, information on
cancer incidence (morbidity) is generally much more accurate than is information on cancer
mortality derived from death certificates. Although the Subcommittee understands that a change
to morbidity-based estimation is a major undertaking with potentially significant policy
implications, and therefore may be impossible in the short run, we recommend it be given serious
consideration for future updates of FGR 13 - Part 1 and other radiation risk activities in the
Agency. (Section 3.1.6)
-------
In FGR 13 - Part 1, risk coefficients are calculated only for the default absorption types
given by ICRP and for the absorption types adjacent to the default types, resulting in an
incomplete list of risk coefficients. The Subcommittee recommends that: a) risk coefficients, at a
minimum, be calculated for each absorption type considered by ICRP for all radionuclides, b) the
default absorption types recommended by ICRP be specifically identified in the tables of FGR 13 -
Part 1, and c) preference be given to using material-specific knowledge in lieu of default values
when appropriate and practicable. (Section 3.1.7)
In the risk calculation for direct radiation exposure from contaminated ground and for
immersion in a radioactive gas or vapor, no shielding from structures is assumed. Because most
people spend the majority of their time indoors, the calculated risks in FGR 13 - Part 1 may
overestimate the real risk. The document should provide some general guidance to the user as to
a reasonable range of risk reduction factors and point the user to methods and references for
estimating an appropriate value within that range. (Section 3.1.8)
Soil ingestion has not been considered as an exposure pathway in FGR 13 - Part 1. The
Subcommittee recommends that FGR 13 - Part 1 be expanded at some time in the future to
include soil intake. (Section 3.1.9)
The risk coefficients for radionuclides in food are expressed as lifetime probabilities of
cancer morbidity or mortality from ingestion of 1 Bq of the radionuclide. Except for separate
tables for radioiodines in milk, the radionuclides are assumed to be distributed in foods such that
radionuclide activity intake per unit caloric intake would remain constant with age and gender. In
real exposure scenarios, this assumption might not be correct. Although the Subcommittee is not
suggesting a suite of risk factors for different food types, it recommends that the Agency include
differential food scenarios in the list of risk analyses for which the FGR 13 - Part 1 tables are not
appropriate. (Section 3.1.10)
The second Charge element (b) addressed the treatment of uncertainty. The major
uncertainties have not entirely been identified and put into proper perspective. Although the
document does well in identifying many of the major uncertainties and describing them
qualitatively, it does not provide the reader with a good sense of the overall uncertainty entailed
when a specific risk coefficient is used to predict the risks of low-level exposure to radiation. The
Subcommittee recommends that the Agency attempt to convey better the overall impact of the
multiple sources of uncertainty on the final risk numbers. In particular, it should more
prominently note the debate about the applicability of the linear no-threshold model at very low
doses, even though the Agency's own evaluations may discount the likelihood of alternative
models. It should also be made clearer how the reliability of the calculations varies from such
highly studied radionuclides as 239Pu or 131I to those for which most of the parameters are based
on analogies with surrogates. (Section 3.2.1)
Quantitative estimates of uncertainty are clearly unnecessary for nominal values intended
only for an initial screening calculation. However, the values in FGR 13 - Part 1 are not intended
-------
for screening; rather, they are intended to provide a realistic best estimate of the average risk per
unit exposure. Given the broad application expected for these risk estimates, the Subcommittee
recommends at a minimum that: a) the individual radionuclides of highest concern for
environmental exposures be placed into categories associated with low, moderate, high, or very
high uncertainty; b) these categories be described quantitatively (e.g., a factor of less than three,
three to 10, 10 to 100, and more than 100, respectively; c) these uncertainty categories be listed in
the tables of risk estimates; d) general guidance about the limits of application be provided for
those radionuclides for which information is insufficient to permit placement into a given
uncertainly category, and e) the final version of FGR 13 - Part 1 include an appendix that
describes the steps that may be taken to improve uncertainty estimates for the risk coefficients to
reflect future advances in the state of knowledge of internal and external dosimetry and of the
response to low doses of ionizing radiation. (Section 3.2.2)
The present state of the art in exposure, dose and risk assessment employs the use of
probabilistic methods to propagate the uncertainly of all major model components through to a
final estimate of dose and risk. In the case of FGR 13 - Part 1, the Subcommittee recognizes that
such calculations, due to inherent complexity of the Agency's methodology, could be a major
undertaking requiring extensive new resources. For future improvements of the Agency's reports,
such as revisions of FGR 13, the Subcommittee recommends that quantification of uncertainty be
considered at the outset of the project. (Section 3.2.3)
Although the lack of scientific consensus on the estimation of lifetime cancer risks from
low-level radiation is acknowledged in FGR 13 - Part 1, the Subcommittee recommends adding a
concise, balanced discussion that lays out the arguments for and against the possibility of a
threshold or other non-linear behavior at low dose and low dose rates. (Section 4.4)
Model uncertainties, i.e., uncertainties in the structure of a biokinetic, dosimetric or
dose-response model, are often more difficult to assess than are parameter uncertainties.
Significant uncertainly can be attributed to the models themselves, apart from uncertainties in
parameter values, and model uncertainty should be reflected in the text of FGR 13 - Part 1.
(Section 4.5)
Although making a detailed probabilistic uncertainty evaluation for every element would
be a formidable task, the Subcommittee recommends that a comprehensive uncertainty calculation
be performed for one example radionuclide for each class of biokinetic model. It might be of use
also to show a calculation for the inhalation of specific radon decay products
and some examples for ill-studied radionuclides to illustrate the range of uncertainties that are
present in the calculations. (Section 4.6)
The third and last major element of the Charge (c) dealt with extending the list of
radionuclides to include those incorporated in previous versions of FGR 13 - Part 1. The
radionuclides addressed in the interim report were primarily those included in ICRP Publications
-------
56, 66, 67, 69, and 71, which contained the models used for calculating dose rates to organs and
tissues. The Subcommittee believes that the Agency's proposed method for extending the list of
radionuclides to include those tabulated in FGRs 11 and 12 is reasonable, but recommends that
risk coefficients also be presented for the inhalation of radon and its short-lived decay products,
the largest source of collective radiation exposure (and presumably, radiation risk) to the U.S.
population as a whole. If it is not feasible to include these risk coefficients in the final version of
FGR 13 - Part 1, their omission should be explained and the risk coefficients included in a future
revision of FGR 13 - Part 1. (Section 3.3)
During the course of the public meeting, the Subcommittee identified several additional
areas where improvements or clarifications could be made to the EPA draft report, and has
provided recommendations to that effect. First, the Subcommittee recommends that the title of
the document be changed from "Health Risks from Low-Level Environmental Exposure to
Radionuclides, Federal Guidance Report 13 - Part 1" to "Estimated Health Risks from Low-
Level Environmental Exposure to Radionuclides, Federal Guidance Report 13 - Part 1: Cancer"
(emphasis added). This change conveys two important points: a) The tabulated risk values are
estimates and b) cancer is the only health effect treated in the document. (Section 3.4.1)
The Subcommittee also noted that use of the FGR 13 - Part 1 risk coefficients could
contribute to debate or ambiguity as to whether a regulated site meets regulatory limits, especially
in cases where a regulation is stated in terms of annual (or per incident) dose limits but the
supporting technical documentation, such as an environmental impact statement, might use risk
calculations. The Subcommittee observes that some confusion may result in any attempt to apply
both the FGR 11 and 12 methodologies and the FGR 13 - Part 1 methodology on the same
regulatory issue. This complication could arise even if the Agency's regulatory action itself used
only one method, because an intervener from the regulated community or the public interest
community could use the other method to contest the regulation. (Section 3.4.2)
The Subcommittee is concerned with the potential for misapplication of the FGR 13 - Part
1 document. As noted in the document, its approach to risk analysis is not intended to be applied
to the analysis of risks to individuals associated with either past or future exposures. The
Subcommittee recommends that the preface of FGR 13 - Part 1 include a more thorough
discussion of the potential for misapplication and the limitations of use for FGR 13 - Part 1 risk
estimates. (Section 6.3)
The Subcommittee believes that tables in any document should be understandable without
reference to the text. For instance, the need for separately considering ingrowth of chain
members in an environmental medium before they enter the body (e.g., by ingestion or inhalation)
and the absence of uncertainty bounds in the tables are critically important to explain to users.
(Section 3.4.4)
It is not clear to the Subcommittee that the listing of a single predominant cancer type in
the risk tables is either useful or correct. Alternatives to the current presentation might be: a)
-------
eliminating the columns completely or b) selecting, for example, the three tissues or organs
considered to be at greatest risk, but without giving a quantitative estimate of the risk fraction
associated with those tissues and organs. (Section 3.4.5)
It is stated in FGR 13 - Part 1 that "Because risk coefficients for external exposure to soil
contaminated to 15 cm would differ only slightly from those for concentrations to an infinite
depth, it would not be useful to provide tabulations of risk coefficients for both situations." The
Uranium Mill Tailings Radiation Control Act (UMTRCA) specifies a standard for contaminated
soil of 5 pCi g"1 in the top 15 cm and 15 pCi g"1 below 15 cm. The Subcommittee recommends
that risk coefficients for external exposure, applicable specifically to the UMTRCA standard, be
calculated and tabulated in the final version of FGR 13 - Part 1 or in a separate document.
(Section 3.4.6)
In conclusion, the Subcommittee commends the Agency on its leadership and efforts in
using up-to-date scientific methods and data to make the health risk estimates in FGR 13 - Part 1.
We believe that our recommendations, except those noted for future consideration, can be
incorporated prior to publication of the final version and will strengthen the report and its
credibility. We strongly support the Agency's stated intent to publish supporting information in
electronic form to accompany release of the final version of FGR 13 - Part 1, and we recommend
that it include the data, models, and dose information used in formulating the risk coefficients.
-------
2. INTRODUCTION
2.1 Background
Since the mid-1980s the United States Environmental Protection Agency has issued a
series of Federal guidance documents for the purpose of providing technical information to assist
Federal agencies in their implementation of radiation protection programs (EPA 1984; 1988;
1993; 1998). The newest in the series, collectively to be called Federal Guidance Report 13 (FGR
13), is intended by the Agency to promote consistency in assessments of the risks to health from
radiation and to help ensure that radiological risk assessments are based on sound scientific
information. The document reviewed in this Science Advisory Board (SAB) report is the interim
version of Part 1 of FGR 13 (FGR 13 - Part 1), which is limited to cancer as a health outcome
(EPA, 1998). The FGR 13 documents are to use state-of-the-art methods and models for
estimating the risks to health from internal or external exposure to specific radionuclides, taking
into account the age- and gender-specific aspects of radiation risk. FGR 13 - Part 1 provides
tabulations of risk estimates, or "risk coefficients," for cancer morbidity and mortality attributable
to exposure to each of approximately 100 radionuclides via various environmental media. These
risk coefficients apply to populations that approximate the age, gender, and mortality experience
characterized by the 1989-91 U.S. decennial life tables.
The tabulations in the final version of Part 1 are expected to extend the methodology of
the interim version to the other radionuclides included in FGRs 11 and 12. Subsequent parts of
FGR 13 may extend the exposure pathways and health endpoints addressed. As necessary, these
publications are to be reissued to update the information provided. The Agency chose to issue
Part 1 of FGR 13 as an interim report in order to provide governmental agencies and other
interested parties an opportunity to become familiar with it and its supporting methodology and to
provide comments for the Agency's consideration before publishing the final version.
The Agency intends that the risk estimates tabulated in FGR 13 - Part 1 be used mainly
for prospective assessments of estimated cancer risks from long-term exposure to radionuclides in
environmental media, such as preparation of environmental impact statements and development of
assessments in support of generic rulemaking for control of radiation exposure. Although it is
recognized that these risk coefficients are likely also to be used in retrospective analyses of
radiation exposures of populations, the Agency emphasized that such analyses should be limited
to estimation of total or average risks in large populations. The tabulations are not intended for
application to specific individuals or to age or gender subgroups, for example, children, and
should not be used for that purpose. Also, these risk coefficients should not be applied to
accident cases involving high doses and dose rates, either in prospective or retrospective analyses.
Finally, some risk assessment procedures are established as a matter of policy, and additional steps
may be needed before using these risk coefficients. For example, the Agency recommends that
radiation risk assessments for sites on the National Priorities List under the Comprehensive
-------
Environmental Response, Compensation, and Liability Act (CERCLA) be performed using the
Health Effects Assessment Summary Tables.
In FOR 13 - Part 1, estimated risks are tabulated directly against the radionuclide
quantities without providing radiation dose information to the user. The linear, no-threshold
dose-response hypothesis and the individual organ and tissue risk coefficients are integrated into
the tabulated results. This approach represents a significant departure from the two previous
Federal Guidance Reports (EPA 1988; 1993) in which dose quantities were tabulated as a
function of radionuclide intake or soil concentration, leaving it to the user to apply the risk
coefficients. For users who desire a simple way to estimate risks to populations directly from
radionuclide concentrations, the approach used in FGR 13 - Part 1 represents a useful step
forward. For experts in the field, however, the lack of radiation dose information and other
details in the interim version of FGR 13 - Part 1 is of concern. However, the Agency has stated
its intention to make available detailed supporting information in electronic form, coincident with
the release of the final version of FGR 13 - Part 1.
The RAC was briefed on the subject on March 3, 1998, and subsequently formed the FGR
13 Review Subcommittee (the Subcommittee). The Subcommittee held a public meeting in
Washington, DC on May 6 and 7, 1998, at which time it was briefed by, and had technical
discussions with, the Office of Radiation and Indoor Air (ORIA) staff. In addition, a public
teleconference was held on June 2, 1998 to discuss an internal working draft.
2.2 Charge
The SAB was asked by ORIA to focus on the following questions:
a) Is the methodology employed for calculating health risks from radionuclide intakes
and external exposures acceptable?
b) In light of scientific information, have the major uncertainties been identified and
put into proper perspective?
c) Is the proposed method for extending the list of radionuclides to include all those
tabulated in Federal Guidance Reports 11 and 12 reasonable?
-------
3. DETAILED FINDINGS
3.1 Assessing Methodology for Calculating Health Risks - Charge Element (a)
The first element of the charge asked the Subcommittee to assess the methodology
employed for calculating health risks from radionuclide intakes and external exposures.
3.1.1 Overall Findings
The methodology is acceptable in the context of a) current Agency science policy choices,
such as the applicability of a linear, no-threshold dose-response model for cancer risks at low
doses of radiation and b) its intended use to support broad Federal radiation protection programs
such as environmental impact statements or generic rulemaking. The preface of the draft report
points out that the tabulations are not intended for application to specific individuals or age or
gender subgroups, especially for retrospective analyses. The Subcommittee also cautions that
they may not be appropriate for site- or situation-specific analyses for which the assumptions of
the document may not be valid.
The internal dose assessment and biokinetic models used in the calculations are from the
most recent publications (numbers 56, 67, 68, 69, and 71) of the ICRP (1989; 1993a; 1994b;
1995a; 1995b). In addition, the most recent model of the respiratory system (ICRP Publication
66, 1994a) is used, where appropriate. In some cases, where data are lacking and or incomplete,
the parameters in the several parts of ICRP Publication 30 (1979; 1980;1981; 1988) are used in
the models. The models, many of the important parameter values, and the default situations are
summarized in tables presented in Chapters 4 and 5 of FOR 13 - Part 1. These calculations
appear to have been made using standard techniques and represent "good science." External
exposure calculations also appear to have been made using standard techniques, many of which
have been used in other calculations for the Agency. However, as discussed below, unpublished
reports were cited for some of the detailed techniques employed. Some specific and more
detailed comments on the methodology are given in the following sections.
3.1.2 Use of Unpublished Methods
The lack of published reports for some of the techniques employed is a weakness of this
work, for both the internal and external exposure calculations. This weakness did not hamper our
evaluation because a Subcommittee member had been supplied with draft copies of these reports
by the authors. Nevertheless, this weakness should be addressed before publication of the final
version of FOR 13 -Parti.
In particular all computations of dose and risk were performed using the dose calculation
(DC AL) software developed by staff at Oak Ridge National Laboratory (ORNL), which is to be
published as an ORNL/TM report. DCAL is a comprehensive computer program of biokinetic-
10
-------
dose-risk models and computational systems used for radiation dosimetry and risk analyses.
While DCAL has been extensively tested, and was used to support five published ICRP reports, it
still needs to be published in order to become openly available and widely accepted by the
scientific community. Because DCAL is such an integral part of FOR 13 - Part 1, the
Subcommittee recommends that the DCAL software and a users manual be published in electronic
form prior to, or concurrently with, the final version of FGR-13, Part 1.
3.1.3 Check of Numerical Results
The Subcommittee considered whether or not it would be possible to check the numerical
results of the risk coefficients in some way. Checking the risk estimates in their entirely would
require observing cancer rates in an exposed cohort, which is clearly impossible. Given far more
time and resources than is available to the Subcommittee, it may be possible to check discrete
components of the risk estimates, such as the time-dependent organ burden of a specific
radionuclide as a result of a given intake, or the specific health outcomes resulting from a given
dose to a specific tissue. Another approach would be an exercise through which several qualified
assessment teams independently predict the lifetime risk per Bq of intake along with estimates of
uncertainty, and their results and rationales compared with each other. This type of exercise has
been performed internationally for environmental transfer models (BIOMOVS, 1993; BIOMOVS
II, 1996; IAEA, 1995; IAEA, 1996; Thiessen et al., 1997) but such an inter-assessment
comparison by independent experts has not been undertaken for the disciplines of internal
dosimetry and risk. Other possibilities would be to assign one or two members of the
Subcommittee the task of asking questions and working with the group at ORNL to more
completely understand what was done, or to have another group run the code for a few specific
test cases.
The Subcommittee concluded that a check of the risk estimates, or merely a verification of
the coding of the model equations, would be a formidable task outside the scope and level of
resources available to the Subcommittee. As an alternative, the Subcommittee recommends that
evidence of verification and quality assurance procedures be included in an appendix to the final
version of FOR 13 - Part 1, or be made available in a separate document.
3.1.4 Absence of Dose Information
Although absorbed dose rates were calculated by the authors of FGR 13 - Part 1 as an
intermediate step in the estimation of risk, dose information is not provided in the document. The
Subcommittee understands that the committed effective dose approach was not used in
preparation of FGR 13 - Part 1, but is concerned that the lack of dose information may make the
report difficult to follow by those who think of radiation safely in terms of dose and will make it
difficult for users to update the risk coefficients as new information on risk or dosimetry becomes
available. The lack of dose information will also make it difficult to employ risk models other
than the linear-no threshold model used by the Agency.
11
-------
The Subcommittee recommends that the authors publish comprehensive absorbed dose
rate or other appropriate dose information in electronic form concurrently with the release of the
final version of FGR 13 - Part 1. Alternatively the authors could identify another source of this
information, such as the anticipated publication of dose information in electronic form by the
ICRP. The Subcommittee also recommends that the dose information be accompanied with
estimates of uncertainty. This would allow the reader to evaluate the importance of uncertainty in
the dosimetry model as compared to the uncertainly in the dose response. The Subcommittee
further recommends that the final version of FGR 13 - Part 1 include a discussion that compares
and contrasts the methodologies of FGR Reports 11 and 12 (EPA, 1988; 1993) with the
methodology of FGR 13 - Part 1.
3.1.5 Information Not Addressed in FGR 13 - Part 1
There are instances in which potentially valuable information, not already incorporated
into the ICRP models, is not addressed or used in FGR 13 - Part 1. For example, a dosimetry
model has been proposed for the esophagus and trachea for many years (Lewis and Ellis, 1979).
In Chapter 5 of FGR 13 - Part 1, the authors state that no model of the esophagus has been
incorporated in the mathematical phantom for internal dose calculations, whereas in Chapter 6,
the discussion indicates that the esophagus has been incorporated into the mathematical model for
external dose calculations (but with no literature citation). One would expect there to be some
commonality between the mathematical phantom used for internal dose assessment and the
phantom used in the external dose assessment portion of the work. The Subcommittee
recommends that the final version of FGR 13 - Part 1 include a discussion of why an esophagus
model was not considered appropriate to include in the model for internal dose rate calculations.
The Subcommittee further recommends that a literature citation or a short description be given of
the model of the esophagus used in the mathematical phantom for external exposure.
Recent calculations of the absorbed energy for monoenergetic electrons in the
gastrointestinal tract could provide an improvement over the very conservative assumptions used
in ICRP Publication 30 and FGR 13 - Part 1 (see Poston et al. 1996a, 1996b). These two papers
have provided a better understanding of the energy dependence of the absorbed fractions in the
wall of the gastrointestinal tract. For example, these data show that at 5 MeV, the absorbed
fraction of energy (AF) for electrons in the gastrointestinal tract wall reaches 0.15. For lower
energies, the absorbed fraction can be significantly less. For example, at 1 MeV, the AF is about
0.1 and decreases monotonically to a value of about 2 x 10"5 at 0.01 MeV. These data must be
contrasted with the ICRP approach used in the calculations for FGR 13 - Part 1. It is clear that,
for many beta-emitting radionuclides, the ICRP approach significantly over-estimates the
absorbed energy in the wall of the tract and, therefore, over-estimates dose (and the risk). Here,
the uncertainly in dose and risk is not merely a factor of two or three, but could be orders of
magnitude. The Subcommittee recommends that the approach described in Poston et al. (1996a;
1996b) be considered for the final version of FGR 13 - Part 1.
3.1.6 Use of Mortality Data vs Morbidity Data
12
-------
Mortality and morbidity risk coefficients are both presented in FGR 13 - Part 1. The
mortality risk coefficients were determined based on data from the Japanese atomic bomb
survivors and other study groups. However, the morbidity risk coefficients were calculated by
taking each site-specific mortality risk and dividing it by its respective lethality fraction, that is, the
fraction of radiogenic cancers at that site which are fatal. A better approach would be to develop
morbidity risk coefficients directly from populations such as the Japanese bomb survivors.
Information on cancer incidence (morbidity) has long been recognized to be generally
much more accurate than is information on cancer mortality derived from death certificates.
Typically, incident cancers are histologically confirmed, with the information gathered by
individuals trained in the collection of such data from a wide variety of sources. They are usually
collected in a timely manner following diagnosis and are unaffected by variations in survival
fraction geographically and over time.
Cancer mortality data, gathered from death certificates for most cancers, are often
collected at a time remote from diagnosis, making epidemiologic studies of exposures
problematic. Cause of death is often recorded by those to whom secondary cause of death
information is not available, and it is generally not confirmed histologically or otherwise. Their
accuracy varies substantially with cancer site. All of these problems are expected to hold in some
measure for the principal studies of radiation-induced cancers.
The availability of cancer incidence data from studies of the atomic bomb survivors and
other irradiated populations provides an opportunity to base cancer risk estimates (either for
morbidity or mortality) on these more reliable statistics. Such an endeavor would also have the
advantage of making the radiation cancer risk estimation process more in harmony with that used
for chemicals. Although the Subcommittee understands that a change to morbidity-based
estimation is a major undertaking with potentially significant policy implications, and therefore
may be impossible in the short run, we recommend it be given serious consideration for future
updates of FGR 13 - Part 1 and other radiation risk activities in the Agency.
Additional, more detailed discussions of the advantages of developing risk coefficients
from morbidity data rather than mortality data are given in the SAB/RAC report on uncertainties
in radiogenic cancer risk (SAB, 1998).
3.1.7 Absorption Types for Inhaled Radionuclides
As indicated in FGR 13 - Part 1, there are significant uncertainties in estimating absorption
types for radionuclides in forms likely to be encountered in the environment. In FGR 13 - Part 1,
risk coefficients are calculated only for the default absorption types given by ICRP and for the
absorption types adjacent to the defaults in the ICRP publication. This approach results in an
incomplete list of risk coefficients. An alternative approach, which would provide the user of
FGR 13 - Part 1 the most flexibility, would be at a minimum to calculate the risks for all of the
13
-------
absorption types considered by ICRP for each of the listed radionuclides, and to indicate the ICRP
default type. In this way, the user would be made aware of the recommended ICRP default
values, which would most likely be used in the absence of material-specific data. With more
information, however, the user could also use one of the alternative risk coefficients.
Specifically the Subcommittee recommends that: a) risk coefficients be calculated for
each absorption type considered by ICRP, for all radionuclides, b) the default absorption types
recommended by ICRP be specifically identified in the tables of FOR 13 - Part 1, and c)
preference be given to using material-specific knowledge in lieu of default values when
appropriate and practicable. These additions should make the report easier to use.
3.1.8 Assumption of Full-time, Unshielded Exposure to Radionuclides in Soil
In the risk calculation for direct radiation exposure from contaminated ground and for
immersion in a radioactive gas or vapor, no shielding from structures is assumed. However, most
people spend the majority of their time indoors. Therefore, the calculated risks in FGR 13 - Part
1 may overestimate the real risk by factors ranging from less than two to over ten. The reduction
in dose due to the inherent shielding characteristics of structures depends not only on the type,
thickness and configuration of the building materials but also on the type and energy of the
radiation. The external exposure calculations are made assuming that the exposed individual is
outdoors for the entire period. This assumption could result in additional bias in the range of a
factor of two to three. The Subcommittee realizes that it would be infeasible to incorporate
structural shielding factors into the risk tables. However, the Subcommittee recommends that the
document include some general guidance to the user as to a reasonable range of risk reduction
factors for shielding and point the user to methods and references for estimating an appropriate
value.
3.1.9 Soil Ingestion
Soil ingestion has not been considered as an exposure pathway in FGR 13 - Part 1. For
most radionuclides and soils, the bioavailability from soils would be likely to be lower than for
foods. Applying the ingestion risk factors for food to soil intake is likely to overestimate the
dose. However, the average soil intake rate for children may be higher than for adults. The
relative ratios of food intake by children to adult intake may be skewed if soil intake is included.
That is, the fraction of the lifetime intake attributable to early childhood may be underestimated
when soil is a significant route of exposure. The Subcommittee recommends that FGR 13 - Part 1
be expanded at some time in the future to include soil intake.
3.1.10 Appropriateness of Diet Averaging
The risk coefficients for radionuclides in food are expressed as lifetime probabilities of
cancer morbidity or mortality from ingestion of 1 Bq of the radionuclide, presumed to be spread
evenly over a lifetime or, if acute, spread evenly across a stationary population. The risk
14
-------
coefficients take into account the variations in caloric intake per unit body weight as a function of
age and gender, so that the contribution of a constant concentration of the radionuclide in food
can have differing impacts on risk at different ages. Except for separate tables for radioiodines in
milk, the radionuclides are assumed in FOR 13 - Part 1 to be distributed in foods such that
radionuclide activity intake per unit caloric intake remains constant with age and gender.
In real exposure scenarios, this assumption of constant intake factors is incorrect, as the
Agency has recognized by computing separate risk factors for milk, a food well known to be
consumed differentially by children and adults. In response to Subcommittee questions, the
Agency further stated orally that no other food group stood out as having a significant variability
not predicted by caloric intake. The Subcommittee did not determine what level of food
disaggregation was used, but pointed out that differences could well be significant for some types
of food, such as fish and apple juice. During the discussion, it became clear that the types of
scenarios envisioned in FOR 13 - Part 1 for application of the risk factors were not ones that
would concentrate radionuclides in only a few narrow food groups. Although the Subcommittee
is not suggesting a suite of risk factors for different food types, it recommends that the Agency
include differential food scenarios in the list of risk analyses for which the FGR 13 - Part 1 tables
are not appropriate. To the extent that all sources of exposure have not been identified, this
approach may either overestimate, or underestimate, the actual exposure.
3.2 Assessing the Treatment of Uncertainty - Charge (b)
The second element of the charge asked the Subcommittee to determine if, in light of
scientific information, the major uncertainties have been identified and put into proper perspective
3.2.1 Overall Findings
The major uncertainties have not entirely been identified and put into proper perspective.
Although the document does well in identifying many of the major uncertainties and describing
them qualitatively, it does not provide the reader with a good sense of the overall uncertainty
entailed when a specific risk coefficient is used to predict the risks of low-level exposure to
radiation. It is possible for a user to conclude equally that the estimates are quite reliable or
almost totally uncertain. Without a better understanding of the plausible limits of uncertainly, any
use of the estimates for regulatory purposes could lead to inappropriate decisions and lack of
public trust. Disclosure of quantitative uncertainty is important whenever an estimate of health
risk is intended to be realistic rather than a
conservative screening value.
Although the Subcommittee is aware of the scientific and computational difficulties of
undertaking a detailed quantitative uncertainty analysis that would examine all the potentially
important sources of uncertainty, it recommends that the Agency attempt to convey better the
overall impact of the multiple sources of uncertainty on the final risk numbers, as it is currently
attempting to do for the uncertainties in the risk model (SAB, 1998, report in preparation). In
15
-------
particular, it should more prominently note the debate about the applicability of the linear
no-threshold model at very low doses, even though the Agency's own evaluations may discount
the likelihood of alternative dose-response models. It should also be made clearer how the
reliability of the calculations varies from such highly studied radionuclides as 293Pu or 131I to those
for which most of the parameters are based on analogies with surrogates.
Detailed guidance for uncertainty analysis has been published in NCRP (1996). This
guidance acknowledges that the quantification of the state of knowledge for exposure and risk
assessment models is inherently dependent on subjective judgment. The authors of FGR 13 - Part
1 have extensive experience in the development and application of dosimetric and risk models for
specific radionuclides and should be able to quantify the state of knowledge in a defensible
manner, using subjective probability distributions. Alternatively, expert elicitation can be used to
derive these distributions as recommended in the SAB/RAC review (SAB, 1998) of the December
1997 Draft Addendum on the uncertainty analysis in estimating radiogenic cancer risks (EPA,
1997). It is emphasized that the uncertainty should be characterized and quantified specifically for
the application of the methodology in FGR 13 - Part 1 to the general U.S. population.
When providing estimates of uncertainty, the Agency should also provide guidance and
incentives for the user to incorporate updates of these estimates, given improvements in the state
of knowledge. Presumably, based upon its statement in the preface of the interim version of FGR
13 - Part 1, the Agency will permit updating of the risk estimates and statements of uncertainty in
specific regulatory cases.
Specific and more detailed comments on aspects of uncertainty are given below.
Specification of quantitative uncertainty estimates is increasingly seen as very important
for decision making. Quantitative estimates of uncertainty are clearly unnecessary for nominal
values intended only for an initial screening calculation and when it is already known that the
screening values are not likely to lead to a substantial underestimate of the true but unknown risk
(see NCRP, 1996).
However, the values in FGR 13 - Part 1 are not intended to be used for screening
calculations; rather, they are intended to provide a realistic best estimate of the average risk per
unit exposure. The Subcommittee recommends that, at a minimum, the individual radionuclides
of highest concern for environmental releases and exposures be categorized into those associated
with low, moderate, high, or very high uncertainty. The Agency should define the numerical
bounds of low, moderate, high, or very high uncertainty, but the following guidance is offered:
Low uncertainty might be assigned to any coefficient likely to be in a range less than a factor of
three either side of a best estimate; moderate uncertainty would be greater than a factor of three,
but less than a factor often; high uncertainty would be greater than a factor often, but less than a
factor of 100, and very high uncertainty would exceed a factor of 100. The Subcommittee
recognizes that in some cases, the uncertainty may not be symmetrical about the best estimate. In
this case, two estimates of uncertainty could be provided (e.g., the estimate of calculated risk is
16
-------
unlikely to underestimate the true risk by more than a factor of 3, but under some circumstances
may overestimate the true risk by a factor of 10 to 100).
The Subcommittee also recommends that identification of categories of uncertainty for
radionuclides likely to be of highest concern for environmental releases and exposures be included
in the tables of risk coefficients, rather than in separate tables. The Subcommittee further
recommends that general guidance be given about the limits of application for those radionuclides
for which information is currently insufficient to support their placement into a specific category
of uncertainty and that the final version of FGR 13 - Part 1 include an appendix that indicates
steps that may be taken to improve estimates of uncertainty to account for future advances in the
state of knowledge for internal and external dosimetry and the dose-response for exposure to low
levels of ionizing radiation.
The Subcommittee recognizes that the present state-of-the-art in exposure, dose and risk
assessment employs the use of probabilistic methods to propagate the uncertainly of all major
model components through to a final estimate of dose and risk. The result is usually a subjective
probability distribution from which a 90% or 95% subjective confidence interval (sometimes
referred to as a credibility interval) is obtained. This information is sometimes accompanied by a
table or pie chart that identifies the most important variables that dominate the overall estimate of
uncertainty in the risk per unit intake or per unit concentration in the environment.
In the case of FGR 13 - Part 1, the initial methodology was developed without considering
the need to propagate uncertainly using probabilistic methods. In the current form of the risk
models used in FGR 13 - Part 1, the Subcommittee recognizes that such uncertainty calculations,
due to inherent complexity of the Agency's methodology, could be a major undertaking requiring
extensive new resources. However, if the uncertainty propagation is limited to those model
components already known to dominate the overall result, then the process of uncertainty analysis
can be made more efficient.
For future improvements of the Agency's reports, such as revisions of FGR 13, the
Subcommittee recommends that quantification of uncertainty be considered at the outset of the
project and that the investigators be provided with sufficient funding and resources to characterize
the state of knowledge for uncertain model components in a defensible manner. The computer
algorithms should be designed at the beginning of the project to facilitate the propagation of
uncertainty through to the final result.
The lack of scientific consensus on the estimation of lifetime cancer risks from low-level
radiation is acknowledged in FGR 13, Part 1. Some scientists believe in an effective threshold for
health effects at sufficiently low doses and low dose rates, citing numerous studies that have not
demonstrated a risk at levels within the range of natural background (BEIR, 1990; Modan, 1991).
Others acknowledge the possibility of a non-linear relationship not adequately treated with the
dose and dose-rate effectiveness factor (DDREF) approach. Concluding a discussion on
uncertainty of risk at low dose, the BEIR V report (BEIR, 1990) states "Moreover, epidemiologic
17
-------
data cannot rigorously exclude the existence of a threshold in the millisievert dose range. Thus
the possibility that there may be no risks from exposures comparable to external natural
background cannot be ruled out. At such low doses and dose rates, it must be acknowledged that
the lower limit of the range of uncertainty in the risk estimates extends to zero." The
Subcommittee recommends that a concise, balanced discussion that lays out the arguments for
and against the possibility of a threshold or other non-linear behavior at low dose and dose rates
be included in the final version of FGR 13, Part 1. This issue has been addressed in detail by the
Agency in its unpublished December 1997 Draft Addendum on the uncertainly analysis in
estimating radiogenic cancer risks (EPA, 1997) and these arguments could be summarized in FGR
13 -Parti.
The position that the Agency has taken with respect to the risk of low-LET radiation at
low doses and low dose rates is somewhat comparable to the position taken by the National
Council on Radiation Protection and Measurement's Report Number 126 (NCRP, 1997) in which
the DDREF for whole body radiation spans an interval from one to five with values less than one
and greater than five being given negligible weight. The Agency states that a more careful
consideration of the uncertainly in the DDREF may be needed for cases where the dose is heavily
concentrated in a few specific target tissues. The DDREF approach is defended in FGR 13 - Part
1 (p. 111) by appealing to either the dual-action theory or the saturable repair theory. The
Subcommittee recommends that the authors acknowledge in FGR 13 - Part 1 the possibility of
alternative models in which the reduction factor for low doses and low dose rates might not be the
same, for example, the genomic instability model proposed by Scott (1997) or models that
recognize a continuous dose-dose rate-response surface.
All models by definition are a simplification of reality and hence some uncertainly is
unavoidable. Model uncertainties, i.e., uncertainties in the structure of a biokinetic, dosimetric or
dose-response model, are often more difficult to assess than are parameter uncertainties. The
current efforts by the Human Alimentary Tract Model Task Group of ICRP Committee 2 to
revise the gastrointestinal tract model point out the recognized need to update the Eve model
(Eve, 1966; Dolphin and Eve, 1966).
The previous respiratory tract model (TGLD 1966) was not adopted by the ICRP until
1979, 13 years after its publication (ICRP, 1979), and was not codified in the United States
(10CFR20) until 1991, 25 years after its publication. By that time, deficiencies in the TGLD
(1966) model were well recognized. These deficiencies, plus the large amount of new research
results, were the impetus for the development of the newer ICRP Publication 66 model (ICRP,
1994a).
Uncertainties in systemic biokinetic models probably vary among elements, depending on
our understanding of the metabolism of the individual elements and on the amount of data
available to construct and parameterize them. Model uncertainties also depend on the specific
models that are to be used. For example, none of the current generation of biokinetic models are
18
-------
physiologically based toxicokinetic models, in which the biokinetics are described based on
models constructed with true physiological variables and values.
The important point is that there is significant uncertainty that can be attributed to models
themselves, apart from uncertainties in parameter values. Continuing scientific review will
elucidate strengths and weaknesses of the models, which will need to be taken into account as
they are identified. The Subcommittee recommends that model uncertainty be discussed in the text
of FOR 13 - Part 1, most logically in the uncertainty section.
3.2.6 Comprehensive Uncertainty Calculation for Classes of Biokinetic Models
There is no question that making a detailed probabilistic uncertainty evaluation for every
element would be a formidable task. However, the Subcommittee recommends that a
comprehensive uncertainty calculation be performed for one example radionuclide for each class
of biokinetic model. Possibilities include well-studied radionuclides such as 1311,90Sr, 239Pu, 137Cs,
3H, and perhaps 238U and 226Ra. It might be of use also to show a calculation for the inhalation of
specific radon decay products. To the extent feasible, the Agency should also consider some
examples for less well-studied radionuclides to illustrate the range of uncertainties that are present
in the calculations. This effort is necessary for the user to appreciate the relative magnitude of the
uncertainties associated with these calculations.
Also, the Subcommittee recommends that uncertainty calculations using the ICRP generic
model for "calcium-like" elements (ICRP, 1993a) be presented. This model is used extensively in
FOR 13 - Part 1, and an uncertainty evaluation would be extremely informative. This task would
not be easy but is feasible.
Usually, a parameter value is chosen from a range of possible values found in the open
literature. It would be helpful to present examples showing the range of values and indicating the
specific value chosen for each parameter. In this regard, the Agency might benefit from the model
of uncertainty analysis used in the NRC (1998) uncertainty analysis of internal dosimetry.
3.3 Assessing the Method for Extending List of Radionuclides - Charge (c)
The third element of the charge asked for the Subcommittee's assessment of the proposed
method for extending the list of radionuclides to include all those tabulated in Federal Guidance
Reports 11 and 12 (EPA, 1988; 1993).
3.3.1 Overall Findings
FGR 13 - Part 1 is not very explicit about how the extension to other radionuclides would
be accomplished, but the Subcommittee did receive a briefing on the Agency's plans for the
extension at the public meeting. Although the proposed method for extending the lists of
radionuclides to those in FGRs 11 and 12 is reasonable, the Subcommittee recommends that risk
19
-------
coefficients also be presented for the inhalation of radon and its short-lived decay products.
These radionuclides, when inhaled, are the largest source of collective radiation exposure (and
presumably, radiation risk) to the U.S. population as a whole. In determining the radon risk
coefficients, consideration should be given to the risk estimates found in the BEIR VI report
(BEIR 1998) and in the ICRP Publication 65 on protection against radon (ICRP 1993b). The
Subcommittee recommends that these risk coefficients be included in the final version of Part 1, if
feasible.
The interim version of the report does not include a tabulation of risk coefficients for all
of the radionuclides covered in FGRs 11 and 12. The radionuclides addressed in the interim
report are primarily those included in ICRP Publications 56, 66, 67, 69, and 71 (ICRP 1989;
1994a; 1993a; 1995a; 1995b), which contained the models used for calculating dose rates to
organs and tissues.
The Agency proposes to calculate risk coefficients for intakes of the remaining
radionuclides in FGR 11 and for external exposures to those remaining in FGR 12. The models in
ICRP Publication 72 (ICRP 1996) will be used to calculate dose rates for intakes of the remaining
radionuclides in FGR 11. These dose rates will be used to calculate risk coefficients in the same
manner as for the radionuclides in the interim report. The methodology for calculating the
external exposure risk coefficients will also be the same as used in the interim report.
3.4 Additional Topics Identified by the Subcommittee
The following items are outside the charge to the SAB but were identified as important by
the Subcommittee.
3.4.1 Recommended Change in Report Title
The Subcommittee recommends that the title of the document be changed from "Health
Risks from Low-Level Environmental Exposure to Radionuclides, Federal Guidance Report 13 -
Part 1" to "Estimated Health Risks from Low-Level Environmental Exposure to Radionuclides,
Federal Guidance Report 13 - Part 1: Cancer" (emphasis added). This change conveys two
important points: 1) The tabulated risk values are estimates and 2) cancer is the only health effect
treated in the document. Note that the "Blue Book" (EPA 1994), from which the health risks
used in FGR 13 - Part 1 were derived, is titled "Revised EPA Methodology for Estimating
Radiogenic Cancer Risks. " (emphasis added)
3.4.2 Contribution of FGR 13 - Part 1 to Regulatory Debate and Ambiguity
Public comments on the document expressed concern that use of the FGR 13 - Part 1 risk
coefficients could contribute to debate or ambiguity as to whether a regulated site meets
regulatory limits, especially in cases where a regulation is stated in terms of annual (or per
20
-------
incident) dose limits but the supporting technical documentation (such as an environmental impact
statement) might be required to use risk calculations. The Agency responded orally that it had
clearly stated that dose-based regulations would still use FGRs 11 and 12 for implementing the
dose calculations and that use of the FOR 13 - Part 1 risk coefficients was not mandated for any
purpose. The Subcommittee observes that some confusion may result in any attempt to apply
both the FGR 11 and 12 methodologies and the FGR 13 - Part 1 methodology on the same
regulatory issue, because the projection of the dose calculations to risk using standard dose-to-
risk conversion factors can result in large differences from the FGR 13 - Part 1 method, as
acknowledged by Agency staff. This complication could arise even if the regulatory agency
action itself used only one method, because an intervener from the regulated community or the
public interest community could use the other method to contest the regulation. An example is
Superfund risk assessments in which dose-based limits are being used as "Applicable or Relevant
and Appropriate Requirements" (ARARs) but risks are also calculated. Although the
Subcommittee makes no recommendation regarding the propriety of having potentially conflicting
guidance in place, it does suggest that the caveats in FGR 13 - Part 1 be strengthened to minimize
any such potential for conflicts to occur.
3.4.3 Domain of Applicability
The Subcommittee is concerned with the potential for misapplication of FGR 13 - Part 1.
As noted in the document, its approach to risk analyses is not intended to be applied to the
analysis of risks to individuals associated with either past or future exposures. The Subcommittee
recommends that the preface of FGR 13 - Part 1 include a more thorough discussion of the
potential for misapplication and the limitations of use for FGR 13 - Part 1 risk estimates.
The Preface of FGR 13 - Part 1 (page iv, second full paragraph, beginning with: "The risk
estimates—" through to the bottom of the page) contains an important statement describing the
intended use and limitations to the applicability of FGR 13 - Part 1 methodology. Since users may
fail to read the preface, the Subcommittee recommends that this important paragraph be repeated
in the main body of the text at least once.
3.4.4 Clarification of the Risk Tables
The Subcommittee believes that tables in any document should be understandable without
reference to the text. In FGR 13 - Part 1, however, the explanatory material required may be
more than can reasonably be placed in the titles and footnotes. As an alternative, each of the risk
coefficient tables could have a note, perhaps in parentheses under the title, that states "Refer to
the explanation of entries in the text preceding this table for important information necessary for
interpreting the table entries."
Two explanations were identified by the Subcommittee as critically important for users of
the tables:
21
-------
a) Ingrowth of chain members in the environmental medium is not incorporated into
the risk estimates; and
b) Because scientists disagree regarding the reliability of estimates of lifetime cancer
risk from low-level exposure to radiation, and because the error in such estimates
may vary substantially from one radionuclide to another and from one exposure
scenario to another, no attempt is made in the tables to characterize the overall
uncertainty associated with any given risk coefficient. (Obviously this last
statement would be modified if the authors of FGR 13 - Part 1 are able to
implement some of the suggestions made in section 3.2 regarding uncertainty.)
The Subcommittee recommends that these two statements be linked clearly with the tables
of risk coefficients in the final version of FGR 13 - Part 1.
3.4.5 Tabulation of Predominant Cancer Types
It is not clear to the Subcommittee that the listing of a single predominant cancer type in
the risk tables is either useful or correct. The scientific basis for making such predictions is
lacking in cases where the tissues at risk are limited, the type of radiation and its dose patterns
(spatial and temporal) are different from those used to develop the risk coefficients (which are
based predominantly on the atomic bomb survivor data), and radionuclide-specific data are sparse
or absent. Alternatives to the current presentation might be: a) eliminating the columns
completely, and b) selecting, for example, the three tissues or organs considered to be at greatest
risk, but without giving a quantitative estimate of the risk fraction associated with those tissues
and organs. This approach at least would dampen the criticism that the predominant cancer type
is over-specified. The Subcommittee's recommendation is to eliminate the predominant cancer
columns because of insufficient knowledge to specify the tumor locations reliably.
3.4.6 Adequacy of Soil Geometries for Use with the Mill Tailings Standards
It is stated in FGR 13 - Part 1 that "Because risk coefficients for external exposure to soil
contaminated to 15 cm would differ only slightly from those for concentrations to an infinite
depth, it would not be useful to provide tabulations of risk coefficients for both situations" (p. 56,
last sentence). The Uranium Mill Tailings Radiation Control Act (UMTRCA) specifies a standard
for contaminated soil of 5 pCi g"1 in the top 15 cm and 15 pCi g"1 below 15 cm. Because of the
wide application of this standard, even beyond its original intent, it would be useful to provide
tabulations of risk coefficients for dose rates from soil that is uncontaminated for the first 15 cm,
but contaminated to an infinite depth below 15 cm. These risk coefficients could then be added to
those for soils contaminated to an infinite depth to readily obtain overall risk coefficients for sites
where the UMTRCA standards apply. This approach would need to be developed only for
radionuclides and their decay products to which the UMTRCA standard is applied. The
Subcommittee recommends that risk coefficients for external exposure, applicable specifically to
22
-------
the UMTRCA standard, be calculated and tabulated in the final version of FGR 13 - Part 1 or in a
separate document.
3.4.7 Relationship Between Exposure and Dose
In example 4 of Appendix E in FGR 13 - Part 1, a risk is derived from a measured
exposure rate (in uR h"1) using a factor of 1 rem per roentgen. Use of this factor results in an
overestimate of the risk, since some self-shielding of critical organs is provided by the body itself.
For example, if the sum of the dose conversion factors for an infinitely thick layer of soil for the
238U decay series found in FGR 12 (EPA 1993) is compared to the exposure rate from an infinite
soil source with 238U in equilibrium with its decay products found in Huffert and Miller (1995), a
conversion factor from uR h"1 to urem h"1 of approximately 0.7 is inferred. Some realistic
guidance for converting measured exposure to risk rates should be provided in FGR 13 - Part 1 if
such examples are to be included.
3.5 General Conclusions
The interim version of FGR 13 - Part 1 is a useful addition to the complement of Federal
Guidance reports relevant to the evaluation of the risks of radiation and makes important
improvements in the calculation of radiation risks for appropriate regulatory decisions. The
report is well organized and well written, although difficult to comprehend for all but highly
technical audiences. The inclusion of sample calculations in the appendices should be useful for
those attempting to understand the details of the extensive calculations, especially if they need to
modify the computed values to fit a situation for which the standard calculations are not suited.
The careful reader will also note that most of the important caveats to the use of the document's
results are included somewhere in the text. However, the caveats are not as prominent as they
probably should be, given the potential for misuse or misinterpretation of the document's risk
estimates. In particular, the magnitudes of the uncertainties in the computed numbers are difficult
to ascertain.
The Subcommittee commends the Agency on its leadership and efforts in using up-to-date
scientific methods and data to develop the health risk estimates in FGR 13 - Part 1. We believe
that our recommendations, except those noted for future consideration, can be incorporated prior
to publication of the final version and will strengthen the report and its credibility. We strongly
support the Agency's stated intent to publish supporting information in electronic form to
accompany release of the final version of FGR 13 - Part 1, and we recommend that it include the
data, models, and dose values used in formulating the risk coefficients.
23
-------
APPENDIX A - DETAILED TECHNICAL COMMENTS
The Subcommittee offers the following specific technical comments:
a) Suitability of Assumed Particle Diameter: P. 53: An activity median
aerodynamic diameter (AMAD) of 1 //m is assumed for the calculations.
Presumably, situations will arise in which the AMAD will be different. The
Subcommittee recommends that comment on this issue be made in the document
and that a procedure for adjusting the coefficients (or perhaps the air
concentration) appropriately be included. Another possibility is to include this
information in the electronic files to be issued by the Agency.
b) Gastrointestinal Absorption Factors: P. 62-63: The gastrointestinal (GI)
absorption factors for ingestion are usually equal to that for inhalation of
absorption type F compounds. Presumably, elements are sometimes chemically
bound in a relatively insoluble form, even in food or water. The document should
explain how the ingestion absorption factors were selected and, if appropriate,
suggest when and how an adjustment should be made if conditions indicate a lower
absorption factor.
c) Compound Speciation: P. 71: Surrogate information for the biokinetic models is
described as being based on a "chemical analogue of the element in humans."
Compound speciation may be important for an element. FOR 13 - Part 1 should
state to what extent the surrogate data are based on an appropriate analogue
compound (as opposed to element).
d) Effective Alpha Particle RBE: The document states that an "effective" alpha
particle RBE of 1 is used for leukemia (P. 98). It is not clear whether another such
adjustment has been made for the effect of radium on bone, where the unadjusted
ICRP methods do not predict radium risks correctly. More broadly, the difficulties
with the whole RBE structure are not sufficiently discussed in this section.
e) Selection of Units of Risk: A statement about the selection of the unit in which
risk is expressed should be given. The Subcommittee offers the following for
consideration by the authors: "Risk coefficients are presented in units of Bq"1
because the Bq is the fundamental SI unit of activity."
A-l
-------
APPENDIX B - ACRONYMS
AF Absorbed Fraction of Energy
AMAD Activity Median Aerodynamic Diameter
ARARS Applicable or Relevant and Appropriate Requirements
BEIR Biological Effects of Ionizing Radiation
BIOMOVS Biospheric Model Validation Study
Bq Becquerel
CERCLA Comprehensive Environmental Response Compensation and Liability Act
Ci Curie
cm Centimeters
Cs-137 Cesium-137
DCAL Dose Calculation (Software)
DDREF Dose and Dose-Rate Effectiveness Factor
EPA U.S. Environmental Protection Agency (U.S. EPA)
FGR Federal Guidance Report (Includes FGR 11, 12, 13)
GI Gastrointestinal
g Gram
h Hour
HRTM Human Respiratory Tract Model
131I Iodine-131
IAEA International Atomic Energy Agency
ICRP International Commission on Radiological Protection
LET Linear Energy Transfer
m Meter
MeV Million electron Volts
NCRP National Council on Radiation Protection and Measurements
NRC U.S. Nuclear Regulatory Commission (U.S. NRC)
NUREG Nuclear Regulatory (U. S. NRC Documents)
ORNL Oak Ridge NationalJ^aboratory
ORIA Office of Radiation and Indoor Air (U. S. EPA)
p Pico (one trillionth, e.g., 10"12 Ci is a picocurie )
239Pu Plutonium-239
226Ra Radium-226
R Roentgen
RAC Radiation Advisory Committee (RAC/U. S. EPA/SAB/RAC)
RBE Relative Biological Effectiveness
rem the unit of the quantity dose equivalent
SAB Science Advisory Board (U.S. EPA/SAB)
B-l
-------
SI Siysteme International d' Unites (international system of units)
90Sr Strontium-90
238U Uranium
UMT Uranium Mill Tailings
UMTRCA Uranium Mill Tailings Radiation Control Act
U.S. United States
// Micro (A millionth of given unit) [e.g., //m - Micrometer (A millionth of a m);
//rem - Microrem (A millionth of a rem); //R - Micro Roentgen (A millionth of a
R)]
B-2
-------
REFERENCES
BEIR (Committee on the Biological Effects of Ionizing Radiation). 1998 (to be published).
Health Effects of Exposure to Radon (BEIR VI), National Academy Press, Washington,
DC. (draft has been released by NAS)
BEIR (Committee on the Biological Effects of Ionizing Radiation). 1990. Health Effects of
Exposure to Low Levels of Ionizing Radiation (BEIR V), National Academy Press,
Washington, DC.
BIOMOVS II (Biospheric Model Validation Study). 1996. An Overview of the BIOMOVS II
Study and its Findings. Stockholm: Swedish Radiation Protection Institute, BIOMOVS
II Technical Report No. 17.
BIOMOVS (Biospheric Model Validation Study). 1993. Final Report. Stockholm: Swedish
Radiation Protection Institute. BIOMOVS Technical Report 15.
Dolphin, G.W. and IS. Eve. 1966. Dosimetry of the gastrointestinal tract. Health Phys. 12,
pp. 163-72.
EPA (Environmental Protection Agency). 1998. Health Risks from Low-Level Environmental
Exposure to Radionuclides, Interim Version of Federal Guidance Report No. 13, Part 1.
EPA-402-R-97-014 (Oak Ridge National Laboratory, Oak Ridge, TN; U. S.
Environmental Protection Agency, Washington, DC.)
EPA (Environmental Protection Agency). 1997. Estimating Radiogenic Cancer Risks Draft
Addendum: Uncertainty Analysis (Draft). U.S. Environmental Protection Agency
Washington, DC
EPA (Environmental Protection Agency). 1994. Estimating Radiogenic Cancer Risks. EPA402-
R-93-076 (U. S. Environmental Protection Agency, Washington, DC.)
EPA (Environmental Protection Agency). 1993. External Exposure to Radionuclides in Air,
Water, and Soil, Federal Guidance Report No. 12 EPA-402-R-93-081 (Oak Ridge
National Laboratory, Oak Ridge, TN; U. S. Environmental Protection Agency,
Washington, DC.)
R-l
-------
EPA (Environmental Protection Agency). 1988. Limiting Values of Radionuclide Intake and Air
Concentration and Dose Conversion Factors for Inhalation, Submersion, and Ingestion,
Federal Guidance Report No. 11 EPA-520/1 -88-020 (Oak Ridge National Laboratory,
Oak Ridge, TN; U. S. Environmental Protection Agency, Washington, DC.)
EPA (Environmental Protection Agency). 1984. The Radioactivity Concentration Guides,
Federal Guidance Report No. 10. EPA520/1-84-005 (Oak Ridge National Laboratory,
Oak Ridge, TN; U. S. Environmental Protection Agency, Washington, DC.)
Eve, IS. 1966. A review of the physiology of the gastrointestinal tract in relation to radiation
doses from radioactive materials. Health Phys. 12, pp. 131-61.
Huffert, A. M., and K. M. Miller. 1995. Measurement Methods for Radiological Surveys in
Support of New Decommissioning Criteria. Draft Report for Comment. NUREG-1506,
Office of Nuclear Regulatory Research. (U. S. Nuclear Regulatory Commission
Washington, DC)
IAEA (International Atomic Energy Agency). 1996. Validation of models using Chernobyl
fallout data from southern Finland—Scenario S. Second Report of the VAMP Multiple
Pathways Assessment Working Group. Vienna: IAEA-TECDOC-904.
IAEA (International Atomic Energy Agency). 1995. Validation of models using Chernobyl
fallout data from the Central Bohemia region of the Czech Republic—Scenario CB. First
Report of the VAMP Multiple Pathways Assessment Working Group. Vienna:
IAEA-TECDOC-795.
ICRP (International Commission on Radiological Protection). 1995a. Age-Dependent Doses to
Members of the Public from Intake of Radionuclides, Part 3. ICRP Publication 69.
Pergamon Press, Elsevier Science Ltd., Oxford, UK.
ICRP (International Commission on Radiological Protection). 1995b. Age-Dependent Doses to
Members of the Public from Intake of Radionuclides, Part 4. ICRP Publication 71.
Pergamon Press, Elsevier Science Ltd., Oxford, UK.
ICRP (International Commission on Radiological Protection). 1994a. Human Respiratory Tract
Model for Radiological Protection. ICRP Publication 66. Pergamon Press, Elsevier
Science Ltd., Oxford, UK.
R-2
-------
ICRP (International Commission on Radiological Protection). 1994b. Dose Coefficients for
intakes of radionuclides by workers. ICRP Publication 68. Pergamon Press, Elsevier
Science Ltd., Oxford, UK.
ICRP (International Commission on Radiological Protection). 1993a. Age-Dependent Doses to
Members of the Public from Intake of Radionuclides, Part 2. ICRP Publication 67.
Pergamon Press, Elsevier Science Ltd., Oxford, UK.
ICRP (International Commission on Radiological Protection). 1993b. Protection against radon-
222 at home and at work. ICRP Publication 65. Annals of the ICRP (2) 23. Pergamon
Press, Elsevier Science Ltd., Oxford, UK.
ICRP (International Commission on Radiological Protection). 1989. Age-Dependent Doses to
Members of the Public from Intake of Radionuclides, Part I. ICRP Publication 56.
Pergamon Press, Elsevier Science Ltd., Oxford, UK.
ICRP (International Commission on Radiological Protection). 1988. Limits for intake of
radionuclides by workers: An Addendum, ICRP Publication 30, Part 4. Pergamon Press,
Elsevier Science Ltd., Oxford, UK.
ICRP (International Commission on Radiological Protection). 1981. Limits for intake of
radionuclides by workers, ICRP Publication 30, Part 3. Pergamon Press, Elsevier Science
Ltd., Oxford, UK.
ICRP (International Commission on Radiological Protection). 1980. Limits for intake of
radionuclides by workers, ICRP Publication 30, Part 2. Pergamon Press, Elsevier Science
Ltd., Oxford, UK.
ICRP (International Commission on Radiological Protection). 1979. Limits for intake of
radionuclides by workers, ICRP Publication 30, Part 1. Pergamon Press, Elsevier Science
Ltd., Oxford, UK.
Lewis, C.A. and R.E. Ellis. 1979. Additions to the Snyder Mathematical Phantom. Phys. Med.
Biol. 24, pp. 1019-1024.
Modan, B. 1991. Low-dose radiation epidemiological studies: An assessment of methodological
problems, in Risks Associated With Ionising Radiations. Annals of the ICRP 22 (1)
R-3
-------
NCRP (National Council on Radiation Protection and Measurements). 1997. Uncertainties in
Risk Estimates for Fatal Cancer Induction by Low LET Radiation. NCRP Report No.
126. Bethesda, MD.
NCRP (National Council on Radiation Protection and Measurements). 1996. A Guide for
Uncertainty Analysis in Dose and Risk Assessments Related to Environmental
Contamination. NCRP Commentary No. 14. Bethesda, MD.
NCRP (National Council on Radiation Protection and Measurements). 1980. Influence of Dose
and Its Distribution in Time on Dose-Response Relationships for Low-LET Radiations.
NCRP Report No. 63. Bethesda, MD.
NRC (Nuclear Regulatory Commission). 1998. Probabilistic Accident Consequence Uncertainty
Analysis. Uncertainly Assessment for Internal Dosimetry. NUREG/CR-6571, Vols. 1&2.
States Nuclear Regulatory Commission, Washington, DC.
Poston, J.W., Jr., K.A. Kodimer, W.E. Bolch, and J.W. Poston, Sr. 1996a. Calculation of
Absorbed Energy in the Gastrointestinal Tract. Health Physics. 71 (No. 3) pp. 300-306.
Poston, J.W., Jr., K.A. Kodimer, W.E. Bolch, and J.W. Poston, Sr. 1996b. A Revised Model for
the Calculation of Absorbed Energy in the Gastrointestinal Tract. Health Physics. 71 (No.
3), pp. 307-314.
SAB (Science Advisory Board). 1998. SAB Radiation Advisory Committee (RAC). Review of
Office of Radiation and Indoor Air Draft Document Estimating Cancer Risks Draft
Addendum: Uncertainty Analysis.
Scott, B.R. 1997. A mechanistic model for neoplastic transformation of cells by high LET
radiation and its implications for low dose, low dose rate, risk assessment. Radiat. Protect.
Dosim. 72, pp.105-117.
Thiessen, K. M., Hoffman, F. O., Rantavaara, A., and S. Hossain. 1997. Environmental models
undergo international test: The science and art of exposure assessment modeling were
tested using real-world data from the Chernobyl accident. Environmental Science &
Technology 31(8), pp. 358A-363A.
R-4
-------
DISTRIBUTION LIST
Deputy Administrator
Assistant Administrators
EPA Regional Administrators
EPA Laboratory Directors
Deputy Assistant Administrator for Office of Policy, Planning and Evaluation
Director, Office of Strategic Planning and Environmental Data
Director, Office of Policy Analysis
Director, Office of Regulatory Management and Evaluation
Deputy Assistant Administrator for Air and Radiation
Director, Office of Radiation and Indoor Air
Director, Office of Radiation Programs
EPA Headquarters Libraries
EPA Regional Libraries
Library of Congress
National Technical Information Service
------- |