Veterans' Benefits: Independent Review Could Improve Credibility of
Radiation Exposure Estimates (Letter Report, 01/28/2000, GAO/HEHS-00-32).

Pursuant to a congressional request, GAO provided information on the
validity of dose reconstruction as a tool for determining veterans'
eligibility for benefits, focusing on: (1) reviewing studies that
assessed the validity of dose reconstruction for estimating veterans'
radiation exposure and discussing the issue with experts in the field
and other knowledgeable individuals; (2) determining what activities are
in place to oversee the dose reconstruction process; and (3) examining
alternatives for deciding veterans' claims for compensation related to
radiation exposure.

GAO noted that: (1) some veterans, veterans' organizations, and experts
GAO talked with do not have confidence in the Department of Defense's
(DOD) dose reconstruction program; (2) they question the completeness of
DOD's data and methodology and believe the DOD's involvement in
estimating radiation doses from exposure to tests for which it was
responsible presents a conflict of interest; (3) however, independent
studies by the National Research Council and the Institute of Medicine
have validated the dose reconstruction process that DOD uses for
deciding radiation claims; (4) these studies point out that DOD's
reconstruction process tends to overestimate both external and internal
does--an outcome that would increase the likelihood that a claim would
be decided in a veteran's favor; (5) DOD conducted separate studies to
determine the accuracy of dose reconstruction and found that the
external radiation dose estimates obtained through reconstruction
methods were about the same as the readings directly measured by film
badges worn by other participants at the same tests; (6) some experts
interviewed, including Health Physics Society representatives, also
support the use of dose reconstruction for claims decisions; (7)
although studies appear to validate DOD's dose reconstruction program
for deciding claims, the agency is not providing for independent
oversight of the program; (8) the Institute of Medicine has been
critical of the program's lack of quality control, including the lack of
a peer review process; (9) the National Research Council has also
suggested that dose reconstruction be reviewed, or subjected to peer
review, by outside independent scientists; (10) a VA official told GAO
it was not VA's responsibility to establish a process to oversee a DOD
program; (11) a DOD official explained that there had been no direct
recommendation to DOD for a peer review process and the program did not
include one when it was designed; (12) GAO did not identify any better
alternatives available for deciding claims that dose reconstruction;
(13) some suggest expanding the list of presumptive diseases as an
alternative to dose reconstruction; and (14) others favor adding all
human cancers to the list, but some experts, including officials at the
National Cancer Institute, find little or no evidence that would link
many cancers to relatively low levels of radiation, such as those levels
to which most veterans were exposed.

--------------------------- Indexing Terms -----------------------------

 REPORTNUM:  HEHS-00-32
     TITLE:  Veterans' Benefits: Independent Review Could Improve
	     Credibility of Radiation Exposure Estimates
      DATE:  01/28/2000
   SUBJECT:  Radiation safety
	     Radiation exposure hazards
	     Nuclear radiation monitoring
	     Nuclear weapons testing
	     Veterans disability compensation
	     Veterans benefits
	     Military benefits claims
	     Atomic energy defense activities
	     Data collection
IDENTIFIER:  DOD Exercise Desert Rock VII
	     Marshall Islands
	     Operation Plumbob

******************************************************************
** This file contains an ASCII representation of the text of a  **

** GAO report.  Delineations within the text indicating chapter **
** titles, headings, and bullets are preserved.  Major          **
** divisions and subdivisions of the text, such as Chapters,    **
** Sections, and Appendixes, are identified by double and       **
** single lines.  The numbers on the right end of these lines   **
** indicate the position of each of the subsections in the      **
** document outline.  These numbers do NOT correspond with the  **
** page numbers of the printed product.                         **
**                                                              **
** No attempt has been made to display graphic images, although **
** figure captions are reproduced.  Tables are included, but    **
** may not resemble those in the printed version.               **
**                                                              **
** Please see the PDF (Portable Document Format) file, when     **
** available, for a complete electronic file of the printed     **
** document's contents.                                         **
**                                                              **
** A printed copy of this report may be obtained from the GAO   **
** Document Distribution Center.  For further details, please   **
** send an e-mail message to:                                   **
**                                                              **
**                                            **
**                                                              **
** with the message 'info' in the body.                         **
******************************************************************

Report to the Committee on Veterans' Affairs, U.S. Senate

January 2000

VETERANS' BENEFITS

Independent Review Could Improve Credibility of Radiation Exposure
Estimates
*****************

*****************

GAO/HEHS-00-32

Letter                                                                     3

Appendixes

Appendix I:Scope and Methodology

                                                                         18

Appendix II:Comments From the Defense Threat Reduction Agency

                                                                         19

Appendix III:Comments From the Department of Veterans Affairs

                                                                         21

Appendix IV:GAO Contact and Staff Acknowledgments

                                                                         22

DOD     Department of Defense

VA      Department of Veterans Affairs

VBA     Veterans' Benefits Administration

VHA     Veterans' Health Administration

                                                     Health, Education, and
                                                             Human Services

B-282605

January 28, 2000

The Honorable Arlen Specter 
Chairman
The Honorable John D. Rockefeller IV
Ranking Minority Member
Committee on Veterans' Affairs 
United States Senate

Between 1945 and 1962, about 200,000 active military personnel
participated in atmospheric nuclear tests, and some were exposed to
potentially harmful doses of radiation./Footnote1/ Over the past 2
decades, veterans with diseases they believe have been caused by radiation
exposure have filed claims for monetary compensation with the Department
of Veterans Affairs (VA). For 16 types of cancer, veterans can receive
compensation without documenting the radiation dose because VA presumes
these cancers to be directly linked to the veterans' exposures. For other
cancers and certain nonmalignant ailments, the Department of Defense (DOD)
estimates, or "reconstructs," the radiation dose using information such as
the duties and locations of veterans and their units during the
atmospheric testing, the type and quantity of radioactivity released from
the explosions, and readings from film badges worn by the veterans and
from other devices that measure radiation dose in the vicinity. The
results of these dose reconstructions are used by VA to decide
compensation claims.

Because there has been significant controversy surrounding the way
radiation doses are estimated, you asked us to review the validity of dose
reconstruction as a tool for determining veterans' eligibility for
benefits. Specifically, we (1) reviewed studies that assessed the validity
of dose reconstruction for estimating veterans' radiation exposure and
discussed the issue with experts in the field and other knowledgeable
individuals, (2) determined what activities are in place to oversee the
dose reconstruction process, and (3) examined alternatives for deciding
veterans' claims for compensation related to radiation exposure.

To address these issues, we interviewed VA and DOD officials and
contractors at headquarters and field locations and reviewed applicable
government policies and procedures. We also interviewed radiation and dose
reconstruction experts in private industry, educational institutions, and
government agencies as well as veterans' service organization
representatives. In addition, we reviewed relevant studies and literature
related to dose reconstruction and the health effects of radiation,
including those by the National Research Council and the Institute of
Medicine. (For the details of our methodology, see app. I.) We performed
our review between November 1998 and October 1999 in accordance with
generally accepted government auditing standards.

Results in Brief

Some veterans, veterans' organizations, and experts we talked with do not
have confidence in DOD's dose reconstruction program. They question the
completeness of DOD's data and methodology and believe that DOD's
involvement in estimating radiation doses from exposure to tests for which
it was responsible presents a conflict of interest. However, independent
studies by the National Research Council and the Institute of Medicine
have validated the dose reconstruction process that DOD uses for deciding
radiation claims. These studies point out that DOD's reconstruction
process tends to overestimate both external and internal doses--an outcome
that would increase the likelihood that a claim would be decided in a
veteran's favor. DOD conducted separate studies to determine the accuracy
of dose reconstruction and found that the external radiation dose
estimates obtained through reconstruction methods were about the same as
the readings directly measured by film badges worn by other participants
at the same tests. Some experts we interviewed, including Health Physics
Society representatives, also support the use of dose reconstruction for
claims decisions./Footnote2/

Although studies appear to validate DOD's dose reconstruction program for
deciding claims, the agency is not providing for independent oversight of
the program. The Institute of Medicine has been critical of the program's
lack of quality control, including the lack of a peer review process. The
National Research Council has also suggested that dose reconstruction be
reviewed, or subjected to peer review, by outside independent scientists.
It has reported that such review could result in greater public confidence
in dose reconstruction. Given the controversy surrounding the program,
independent review could help resolve concerns about the integrity of the
program. A VA official told us it was not VA's responsibility to establish
a process to oversee a DOD program. A DOD official explained that there
had been no direct recommendation to DOD for a peer review process and
that the program did not include this feature when it was designed. We
therefore recommend that DOD establish a process for independent review of
its dose reconstructions.

Finally, we did not identify any better alternatives currently available
for deciding claims than dose reconstruction. For example, although some
suggest expanding the list of presumptive diseases as an alternative to
dose reconstruction, deciding which cancers to add to the presumptive list
is controversial. Some advocate a restricted approach, such as adding only
those cancers that statistically occur more frequently in the exposed
veteran population than in other similar groups of nonexposed people. Such
an approach would result in the addition of few, if any, cancers or
ailments to the presumptive list because few cancers have been shown to be
more prevalent among nuclear test participants than among nonparticipants.
Others favor adding all human cancers to the list, but some experts,
including officials at the National Cancer Institute, find little or no
evidence that would link many cancers to relatively low levels of
radiation, such as those levels to which most veterans were exposed.

Background

Between 1945 and 1962, the United States conducted atmospheric tests of
nuclear weapons--primarily at the Pacific Proving Grounds in the Marshall
Islands and at the Nevada Test Site in southern Nevada. The tests were
conducted for several reasons, including to determine the effects of
atomic bombs on military installations and equipment and to provide
training to crews in military tactics for using atomic bombs. Some
veterans who participated in these tests were exposed to varying doses of
radiation, and some subsequently developed a cancer or nonmalignant
ailment. All veterans who have participated in a nuclear test can receive
health care for conditions recognized by VA as potentially caused by
radiation, including all forms of cancer. These veterans may also be
eligible for disability compensation if their conditions are determined to
be service-connected. For example, a veteran without a spouse or
dependents could generally receive from $98 to $2,036 a month, depending
on the extent of disability.

Veterans who seek compensation can file a claim with VA's Veterans'
Benefits Administration (VBA). To resolve these claims, VBA uses one of
two processes, depending on the veteran's cancer or ailment. If the
veteran was a participant at an atmospheric test and has any one of 16
specific cancers, the veteran can be compensated, under 38 C.F.R.
****ITCCentury Book:xa4**** 3.309(d), without further proof that the
cancer was caused by radiation exposure. These 16 cancers are presumed,
under 38 U.S.C. ****ITCCentury Book:xa4**** 1112(c)(2), to have been
caused by radiation that the veterans were exposed to as test
participants./Footnote3/

The second claims process, under 38 C.F.R. ****ITCCentury Book:xa4****
3.311, is used for veterans with all other cancers and certain
nonmalignant ailments: tumors of the brain and central nervous system,
nonmalignant thyroid nodular disease, posterior subcapsular cataracts, and
parathyroid adenoma. This second claims process relies on dose
reconstruction to decide the claim. Specifically, DOD, under an agreement
with VA, uses scientific and personnel data to estimate the amount of
radiation the veteran received during the atmospheric tests. Both external
doses and internal doses--those associated with inhalation or ingestion of
radioactive material--can be estimated. For example, external radiation
can be estimated from available radiation measurement devices or through
calculations based on the proximity of personnel to radiation sources.
Internal radiation can be estimated on the basis of the airborne
concentration of radioactivity, the duration of exposure to airborne
fallout or suspended particles in contaminated areas, and how the
radioactivity enters and is transported through the body. When historical
records documenting the exact radiation dose are incomplete, inaccurate,
or missing, dose reconstruction is used to either complement or substitute
for dose records.

VA's Veterans' Health Administration (VHA) then compares the reconstructed
doses with reports showing dose levels that have at least a small
reasonable possibility of being as likely as not to have caused the
cancer./Footnote4/ For example, one report shows that exposure to 17 rads
of radiation at age 20 or to 33.1 rads of radiation at age 30 would have
at least a 1-percent chance of being as likely as not to cause colon
cancer./Footnote5/ In another example, exposure of a 20-year-old nonsmoker
to 4.3 rads of radiation would have at least a 1-percent chance of being
as likely as not to cause lung cancer. VHA's medical opinions are
considered by VBA in making decisions on benefit awards.

Since 1977, about 9,200 of the 200,000 nuclear test participants have
filed claims with VA for compensation for illnesses that they believe were
caused by exposure to radiation during atmospheric nuclear
tests./Footnote6/ Not all of these veterans, however, have had dose
reconstructions. For example, those with cancers on the presumptive list
would not need a dose estimate. The applications for compensation of
others may have lacked evidence of an ailment or appropriate military
service. In all, DOD has used historical records to conduct dose
reconstructions for about 2,900 veterans./Footnote7/ Although VA does not
keep data on the disposition of claims involving dose reconstructions, it
estimated that there have been relatively few awards. According to VA
officials, the agency denied the majority of claims because the radiation
doses to which the veterans were exposed were too low to have been likely
to cause their cancer or ailment.

Dose Reconstruction Is Considered Valid for Deciding Claims

Veterans and veterans' service organizations have expressed concern over
the completeness of data used by DOD and the methodology it uses to
estimate doses, particularly doses from inhaled or ingested radioactive
particles. Some are also skeptical about DOD's ability to be unbiased in
the dose reconstruction process, since DOD was responsible for the
atmospheric testing that exposed the veterans to radiation.

However, studies conducted by DOD, the National Research Council, and the
Institute of Medicine have generally shown that DOD's dose reconstruction
approach is valid for providing dose estimates for VBA to use in deciding
claims. DOD has studied the validity of dose reconstruction by using the
process to estimate doses for cases in which the actual dose is well
documented. In such studies, DOD has found, for example, that
reconstructed doses are similar to doses recorded on film badges, or
radiation measurement devices, worn or used by some nuclear test
participants. Independent reviews conducted by the National Research
Council in 1985 and by the Institute of Medicine in 1995 showed that DOD's
reconstructed estimates are typically high, which would generally favor
veterans in the claims process. In addition, representatives from the
Health Physics Society and other experts express confidence that dose
reconstruction is capable of producing dose estimates that are adequate
for claims decision purposes.

Some Question the Reliability of Dose Reconstruction
----------------------------------------------------

A number of veterans and representatives of veterans' service
organizations told us that they do not have confidence in the process DOD
uses to estimate radiation exposure. Specifically, they question the
completeness of the data used and certain assumptions used to estimate
doses.

Several veterans and veterans' organizations we interviewed were skeptical
about DOD's knowledge of personnel activities and locations. For example,
a veterans' service organization official told us that military personnel
swam in a lagoon contaminated by radiation shortly after a nuclear test,
but that he doubted that such information would be recorded in documents
available to DOD. Yet a DOD official told us that veterans are expected to
include these types of assertions in their claims, and regulations require
DOD to accept these assertions as fact unless DOD can prove them incorrect.

Some veterans and veterans' organizations also questioned the assumptions
that DOD makes when information on a veteran's exposure is incomplete or
missing. These individuals believe that if DOD had more specific facts
about veterans' duties, dose estimates could be higher. For example, for
one dose reconstruction for a veteran who washed contaminated aircraft,
DOD estimated the radiation dose to the veteran's skin on the basis of
assumptions about the distance between the veteran and the aircraft. DOD
officials told us that they believe assumptions that they are required to
make are reasonable because specific evidence of exact activities does not
always exist and that any inaccuracies in these assumptions generally do
not significantly alter the dose estimates.

Some veterans also questioned DOD's methodology for estimating internal
radiation from inhaled or ingested radioactive particles. According to an
expert at Oak Ridge National Laboratory, internal radiation doses are
typically much smaller than external doses and are also very difficult to
accurately estimate. In addition, a former health physicist from the
Nevada Test Site believes that internal doses have been underestimated
because particles in the soil from previous tests at the Nevada site could
have been suspended in the air by subsequent nuclear detonations without
corresponding indications of this radiation from previous tests appearing
on film badges or other measuring devices. According to an official in
DOD's dose reconstruction program and to written dose reconstruction
procedures, however, fallout from prior nuclear tests is factored into
dose reconstructions.

Evidence From Scientific Studies Supports DOD's Dose Reconstruction
-------------------------------------------------------------------

We identified five studies conducted since 1979 that assessed the validity
of dose reconstruction in estimating radiation exposure. Three of these
studies were conducted by the DOD contractor responsible for performing
dose reconstructions. To determine the process' validity, the contractor
compared known doses recorded on film badges with estimated doses
determined through the dose reconstruction process./Footnote8/ The DOD
contractor reported in 1979, 1980, and 1981 that reconstructed doses were
similar to those found on film badges for external doses./Footnote9/ For
example, in 1981, the contractor prepared an analysis of nuclear radiation
exposure of personnel in a Marine brigade during their participation in
Exercise Desert Rock VII, Operation Plumbbob, at the Nevada Test Site in
1957. The report notes that "film badge dosimetry and other records of
activity are sufficient to identify the personnel who had the greatest
potential for exposure and to determine their specific activities."
Further, the report concludes that dosimetry data, or data from radiation
measurement devices, "correlate well with calculated doses, thus providing
the necessary confidence in the calculated doses, and group activities,
for those personnel with no dosimetry records."

Studies by the National Research Council and by the Institute of Medicine
have shown that dose reconstruction is a valid process for estimating
radiation doses for VBA to use in deciding claims because dose
reconstruction tends to overestimate doses, giving the benefit of the
doubt to the veteran. In turn, this favors veterans in the claims process.
Specifically, the National Research Council issued a report in 1985 on the
comprehensiveness and scientific validity of the methods and procedures
DOD's Nuclear Test Personnel Review organization uses to estimate internal
and external doses of radiation./Footnote10/ The report concluded that
"the methods used by the [Nuclear Test Personnel Review] team to assign
external . . . doses are generally reasonable and make appropriate use of
available data." Although the National Research Council found that the
methods used to estimate internal doses were based on unsupported
assumptions, it nevertheless found that internal doses are relatively
insignificant when compared with external doses, regardless of the
estimate. The National Research Council also reported that if any bias
exists in the estimates, it is probably a tendency to overestimate doses,
especially internal doses. Similarly, the Institute of Medicine noted in a
1995 letter to the Nuclear Test Personnel Review organization that DOD has
resolved uncertainties in dose reconstructions by assigning veterans
higher rather than lower doses./Footnote11/ 

According to the Health Physics Society, dose reconstruction is a standard
practice for estimating the amount of radiation exposure when more direct
evidence, such as film badge readings, is not available. Other health
physicists and experts with whom we spoke shared this view.

Independent Review Could Enhance Confidence in the Program

Public and veteran confidence in dose reconstruction as a tool to decide
compensation claims could be improved through independent oversight and
validation of dose reconstruction results. However, ongoing independent
reviews are currently not being conducted to validate program results.
Biological testing to verify dose reconstructions is currently being
investigated by the scientific community, and some scientists believe that
this approach may be useful in the future.

DOD's Dose Reconstruction Program Is Not Independently Reviewed
---------------------------------------------------------------

Some individuals we interviewed believe that DOD's involvement in dose
reconstruction poses a conflict of interest, since DOD was responsible for
the atmospheric tests that exposed the veterans to radiation. In addition,
one expert told us that DOD did not sufficiently monitor the nuclear tests
and misled participants about the potential health effects of radiation.
Allegations from veterans that data from nuclear tests are inaccurate and
that DOD's estimates of exposure are too low further reflect veterans'
distrust. Independent reviews of DOD's dose reconstruction program could
help mitigate some of these criticisms.

According to VA regulations, a veteran who disagrees with the dose
estimate provided by DOD can obtain, at the veteran's expense, an
independent estimate from a reliable source. If the independent estimate
is at least twice as large as the DOD estimate, VA contracts with a third
party to reconcile the two estimates and uses the reconciled estimate to
make compensation decisions. Officials in VBA's Compensation and Pension
Office told us that they could not remember an instance during the last 5
years when any veteran presented an independent estimate. A dosimetry
expert selected by VA to reconcile DOD dose estimates with veterans'
independent estimates told us that he completed four such reconciliations
in the early 1990s. For one of the reconciliations, his estimate was
higher than the DOD estimate but lower than that provided by the veteran;
he generally agreed with DOD's estimates for the other three
reconciliations. 

In a 1995 letter to the Nuclear Test Personnel Review organization, the
Institute of Medicine criticized DOD's quality assurance program for dose
reconstruction as limited and noted that there has been little peer review
of the methods used and of the actual dose estimates. The Institute also
noted that no evidence exists that dose estimates were verified by an
independent source other than by the National Research Council in the
1980s. According to 1995 National Research Council guidance on dose
reconstruction for epidemiological uses, dose reconstructions should
undergo peer review if they are to be viewed as credible./Footnote12/ The
guidance is primarily directed at epidemiological studies of populations
located near nuclear weapons plants. It suggests that all dose
reconstruction studies be reviewed by groups of scientists and public
health officials who are not directly involved in the study, either as
participants or as advisors, and that time and resources be allocated for
resolving any discrepancies in the results. Applying the same principle,
independent validations of a sample of individual dose reconstructions
could enhance confidence in the DOD program.

VA officials told us that they did not think it was VA's responsibility to
establish a dose reconstruction review program; if a program were to be
established, they said it would be DOD's responsibility to establish it.
An official in DOD's dose reconstruction program told us that there is no
requirement for peer review of the program. In addition, he noted that the
program was examined by the National Research Council in 1985 and the
Institute of Medicine in 1995 and that there had been no direct
recommendation to DOD for a peer review process. The DOD official also
told us that a peer review program would not likely be acceptable to
veterans if DOD had any involvement in the process.

Biological Technology Has Potential for Validating Dose Reconstruction
Results in the Future
---------------------------------------------------------------------------

Some scientists believe biological measurement techniques could be useful
in estimating radiation exposure and validating the results of dose
reconstruction techniques. Once biological techniques are fully studied
and tested, it is conceivable that they might eventually be used as a
substitute for dose reconstruction. Unlike DOD's reconstruction process
and other dosimetric methods, biological techniques do not depend on the
availability of data on military activities, locations, and atmospheric
conditions. Instead, biological techniques essentially use the body as a
dosimeter and may, for example, measure chromosomal aberrations in the
blood or the effects of radiation on teeth. In 1985, the National Research
Council suggested that biological techniques be used in the dose
reconstruction program and that biological testing be used to measure the
presence of plutonium and strontium in veterans' urine.

As suggested by the National Research Council, DOD recently started a
program to test for veteran exposure to plutonium, a radioactive element
found in fallout from atmospheric nuclear testing. Under the program, 100
veterans who participated in nuclear tests volunteered to provide urine
specimens that have been tested for plutonium by a national laboratory.
Test results are being analyzed, and final results will be reported early
in 2000. According to a DOD fact sheet, results from DOD's dose
reconstruction program will be reevaluated for any veterans whose
plutonium levels are found through urine testing to be significantly
elevated. However, no reconstructed doses will be lowered as a result of
biological testing. According to a DOD official, waiting for the results
of separate feasibility studies by the National Institutes of Health,
refining the testing technique, and obtaining funding have contributed to
the delay in implementing the National Research Council suggestion.

Other biological measurement techniques for identifying radiation exposure
include blood tests for aberrant chromosomes. Although the National
Research Council's 1995 guidance on conducting dose reconstruction studies
cited this technique as promising for measuring past radiation exposure,
it also reported that such testing was more accurate at higher rather than
lower levels of radiation exposure. Scientists also suggested that
determining the extent of radiation exposure in the teeth of test
participants might be used to validate dose reconstructions. These
biological measurement techniques, however, are still in the developmental
stage.

Better Alternatives to Dose Reconstruction Are Not Currently Available

We identified no better alternatives for deciding claims than dose
reconstruction. Veterans and veterans' service organizations advocate
expanding the presumptive list of 16 cancers. However, the experts with
whom we spoke do not agree on which additional cancers or ailments, if
any, are caused by various levels of radiation exposure.

Some experts have recommended adding to the presumptive list those cancers
that occur more frequently in veterans who have been exposed to radiation
than in similar groups that have not been exposed. This type of comparison
is made in mortality studies. Theoretically, basing compensation on the
results of these studies could potentially increase confidence that the
cancers added to the presumptive list were statistically correlated with
participation in atmospheric testing. But, in fact, the statistically
significant health effects found by mortality studies vary. For example,
many findings in a recent study by the Institute of Medicine comparing the
mortality of about 70,000 individuals who participated in one of five
nuclear tests with the mortality of about 65,000
nonparticipants/Footnote13/ differed from the results of prior studies
conducted by other researchers. The Institute of Medicine study found, for
example, that the increased mortality from leukemia among test
participants was not statistically significant, while other mortality
studies found increases in mortality from leukemia that were significant.
Leukemia is on VA's presumptive list. Conversely, the Institute of
Medicine study found statistically significant increases in mortality from
nasal and prostate cancers. Other mortality studies have not found
statistical elevations in mortality from these cancers, and they are not
on the current presumptive list./Footnote14/

Other experts advocate including all known cancers on the presumptive
list, arguing that the science is not clear-cut on radiation-induced
cancer or exposure levels that cause disease. For example, a professor of
molecular and cell biology--who is also an expert in nuclear and physical
chemistry--told us that there is no evidence showing that radiation is
safe at any level. Other experts, including officials from the National
Cancer Institute, contend that the relationship between radiation and
cancer is indeed well understood and that there is no evidence that links
low levels of radiation to many cancers.

Conclusions

Available scientific studies indicate that dose reconstruction is a valid
method for estimating veterans' radiation exposure to decide disability
claims, and we have not identified a better alternative. However, the dose
reconstruction program lacks an independent review process. According to a
DOD official, there has not been a direct recommendation for DOD to
establish such a process and peer review was not included as part of the
dose reconstruction program when it was established. Given the controversy
surrounding the program, an independent review process for validating
DOD's dose reconstruction methods and estimates could mitigate concerns
about the integrity of the program. In the future, DOD may be able to
validate its dose reconstruction program by biological measurement
methods, such as counting chromosomal aberrations or measuring radiation
exposure in teeth, although such methods are still in the early stages of
development.

Recommendation

We recommend that the Secretary of Defense establish an independent review
process for the dose reconstruction program under which independent
verifications of a sample of individual dose reconstructions are made.

Agency Comments

We provided a draft copy of this report to DOD and VA. Both agencies
provided technical comments, which we incorporated where appropriate. The
agencies' comments concerning our recommendation and our response follow.
DOD's and VA's comments appear in appendixes II and III, respectively.

Defense Threat Reduction Agency
-------------------------------

The Director of the Defense Threat Reduction Agency agreed in principle
with our recommendation. However, he stated that an organization separate
from DOD, such as the National Institutes of Health, should support and
direct the independent review process to avoid any appearance of conflict
of interest and to build veterans' confidence in the review process. We
believe that DOD, as program manager, should maintain responsibility for
program integrity, including establishing a review program. DOD could
arrange for an independent entity, such as the National Institutes of
Health, to conduct a peer review program and ensure the entity chooses a
methodology that is independent of DOD influence. We would expect that DOD
and any independent entity would consider the views of others, including
veterans and veterans' organizations, on how the program would be set up
in order to enhance their confidence in the program. 

Department of Veterans Affairs
------------------------------

The Assistant Secretary for Planning and Analysis agreed with our
recommendation that the Secretary of Defense establish an independent
review process for the dose reconstruction program and said that VA would
not be the appropriate organization to perform such reviews.

We are sending copies of this report to the Honorable William S. Cohen,
Secretary of Defense; the Honorable Togo D. West, Jr., Secretary of
Veterans Affairs; appropriate congressional committees; and other
interested parties. We will also make copies available to others upon
request. 

Please contact me on (202) 512-7111 if you or your staff have any
questions. Another GAO contact and other staff who made major
contributions to this report are listed in appendix
IV.
*****************

*****************

Stephen P. Backhus
Director, Veterans' Affairs and
  Military Health Care Issues

--------------------------------------
/Footnote1/-^In addition to those veterans who participated in atmospheric
  tests, approximately another 200,000 were potentially exposed to
  radiation as a result of their presence in Japan after World War II.
  These veterans were excluded from our review because a single dose
  reconstruction was done for all.
/Footnote2/-^The Health Physics Society is a nonprofit scientific
  organization composed of 6,800 scientists, physicians, engineers,
  lawyers, and other professionals representing academia, industry,
  government, national laboratories, trade unions, and other
  organizations. The society promotes the practice of radiation safety and
  attempts to ensure that health physicists have information and
  capabilities to control the beneficial use of radiation in medical,
  research, and power generation activities to protect workers and the
  public from potential hazards. 
/Footnote3/-^In 1988, 13 cancers--leukemia (except chronic lymphocytic
  leukemia); multiple myeloma; lymphomas (except Hodgkin's disease);
  primary liver cancer; and cancers of the thyroid, breast, pharynx,
  esophagus, stomach, small intestine, pancreas, bile ducts, and gall
  bladder--were legislatively deemed to be service-connected for veterans
  who participated in an atmospheric nuclear test. In 1992, two additional
  cancers--cancers of the salivary gland and urinary tract--were added to
  the list, and in 1999 bronchiolo-alveolar carcinoma was added.
/Footnote4/-^Dose levels have historically been developed by groups such
  as the National Research Council and the Committee on Interagency
  Radiation Research and Policy Coordination, which was chartered through
  the Federal Coordinating Council for Science, Engineering and
  Technology, Office of Science and Technology Policy, Executive Office of
  the President. The National Cancer Institute recently began an effort to
  update the dose levels.
/Footnote5/-^Doses from this report are based on a 99-percent lower
  credibility or confidence limit that has a probability of causation of
  50 percent. A rad, or radiation absorbed dose, is a measurement of
  ionizing radiation energy that is absorbed. When DOD estimates a range
  of potential exposures for a veteran, VHA will use the upper limit.
/Footnote6/-^On the basis of statistics from recent mortality studies, a
  DOD official told us that an estimated 70 to 75 percent of the 200,000
  exposed veterans are presumed alive. In addition to compensating
  veterans for radiation-related claims, VA may also compensate the
  survivors of deceased veterans. 
/Footnote7/-^According to a DOD official, the 2,900 dose reconstructions
  may include some that were done at the request of individual veterans
  who might not have filed claims with VA.
/Footnote8/-^In 1989, the National Research Council studied the
  reliability of film badges that were available for some participants in
  nuclear tests and found that film badge data were generally reliable.
  See National Research Council, Committee on Film Badge Dosimetry in
  Atmospheric Nuclear Tests, Energy Engineering Board, Commission on
  Engineering and Technical Systems, Film Badge Dosimetry in Atmospheric
  Nuclear Tests (Washington, D.C.: National Academy Press, 1989).
/Footnote9/-^Science Applications, Inc., Analysis of Radiation Exposure
  for Task Force Warrior--Shot Smoky--Exercise Desert Rock VII-VIII
  Operation Plumbbob, DNA 4747F (1979); Analysis of Radiation Exposure for
  Task Force Big Bang, Shot Galileo, Exercise Desert Rock VII-VIII
  Operation Plumbbob, DNA 4772F (1980); and Analysis of Radiation
  Exposure, 4th Marine Corps Provisional Atomic Exercise Brigade, Exercise
  Desert Rock VII, Operation Plumbbob, DNA 5774F (1981).
/Footnote10/-^National Research Council, Committee on Dose Assignment and
  Reconstruction for Service Personnel at Nuclear Weapons Tests, Board on
  Radiation Effects Research, Commission on Life Sciences, Review of the
  Methods Used to Assign Radiation Doses to Service Personnel at Nuclear
  Weapons Tests (Washington, D.C.: National Academy Press, 1985).
/Footnote11/-^National Academy of Sciences, Institute of Medicine, A
  Review of the Dosimetry Data Available in the Nuclear Test Personnel
  Review (NTPR) Program, An Interim Letter Report of the Committee to
  Study the Mortality of Military Personnel Present at Atmospheric Tests
  of Nuclear Weapons to the Defense Nuclear Agency (Washington, D.C.:
  Institute of Medicine, May 15, 1995).
/Footnote12/-^National Research Council, Committee on an Assessment of CDC
  Radiation Studies, Board of Radiation Effects Research, Commission on
  Life Sciences, Radiation Dose Reconstruction for Epidemiologic Uses
  (Washington, D.C.: National Academy Press, 1995).
/Footnote13/-^Medical Follow-up Agency, Institute of Medicine, The Five
  Series Study: Mortality of Military Participants in U.S. Nuclear Weapons
  Tests (Washington, D.C.: National Academy Press, 1999).
/Footnote14/-^The Institute of Medicine study reported that its finding
  about nasal and prostate cancers has not been reported in other
  mortality studies or in studies of Japanese atomic bomb survivors.
  Furthermore, the Institute study explained the divergent findings on
  prostate cancer--which it acknowledges is generally not linked to
  radiation exposure--as possibly the result of veterans' concerns about
  cancers caused by exposure. Such concerns may have led to more
  identifications of prostate cancer and a subsequent increase in reported
  deaths due to prostate cancer.

SCOPE AND METHODOLOGY
=====================

To review assessments of the validity of dose reconstruction as a method
for estimating radiation exposure, we relied on various studies and the
expertise of scientists in this area. We conducted literature searches for
journal articles, scholarly papers, and other similar materials related to
estimating radiation exposure. In particular, we reviewed empirical
studies that investigated the degree to which reconstructed dose estimates
correlate with actual known doses. We also reviewed reports from those who
have specifically critiqued the Department of Defense's (DOD) dose
reconstruction program. In addition, we reviewed statements from witnesses
who testified at the 1998 hearing on ionizing radiation, veterans' health
care, and related issues before the Senate Committee on Veterans' Affairs.
We then contacted experts associated with scientific organizations,
including the National Academy of Sciences, the National Research Council,
the National Cancer Institute, and the Health Physics Society, who were
knowledgeable about radiation exposure to get other expert opinions and to
discuss conflicting views on the topic. In addition, we obtained the views
of representatives from veterans' groups, such as the National Association
of Atomic Veterans, and veterans' service organizations, including the
American Legion, Disabled American Veterans, and Veterans of Foreign Wars.

To determine the oversight activities associated with the dose
reconstruction program, we spoke with DOD officials responsible for such
activities, including officials in the Defense Threat Reduction Agency,
the DOD agency with responsibility for the program, as well as the
contractor who actually performs the dose reconstructions. In evaluating
the strength of DOD's oversight activities, we reviewed National Research
Council guidance on conducting dose reconstructions and compared DOD's
processes with this guidance. Additionally, we considered comments by the
Institute of Medicine on DOD's procedures.

In researching alternatives to dose reconstruction for deciding claims, we
relied on existing scientific studies and the knowledge of the scientific
community. In particular, when considering the use of presumptive lists,
we reviewed studies of the health effects of radiation exposure, such as
those prepared by the Committee on Interagency Radiation Research and
Policy Coordination, the National Institutes of Health, and the National
Research Council. We also reviewed studies that evaluated increases in
mortality for radiogenic diseases for nuclear test participants. We also
obtained the views of veterans' organizations on using presumptive lists
for deciding claims. 

COMMENTS FROM THE DEFENSE THREAT REDUCTION AGENCY
=================================================

*****************

*****************

*****************

*****************

COMMENTS FROM THE DEPARTMENT OF VETERANS AFFAIRS
================================================

*****************

*****************

GAO CONTACT AND STAFF ACKNOWLEDGMENTS
=====================================

GAO Contact

Ronald J. Guthrie, (303) 572-7332

Staff Acknowledgments

John A. Borrelli, Joseph J. Buschy, Nancy L. Crothers, Deborah L. Edwards,
Chariti E. Gent, George L. Lorenzen, B. Behn Miller, and Karen M. Sloan
also made key contributions to this report.

(105771)

*** End of document. ***