Gulf War Illnesses: Federal Research Strategy Needs Reexamination
(Testimony, 02/24/98, GAO/T-NSIAD-98-104).

GAO discussed its evaluation of the federal strategy to research Gulf
War illnesses.

GAO noted that: (1) the government was not proactive in researching Gulf
War illnesses; (2) the government's early research emphasized stress as
a cause for Gulf War veterans' illnesses and gave other hypotheses, such
as multiple chemical sensitivity, little attention; (3) in contrast, the
private sector pursued research on the health effects of low-level
exposures to certain chemical warfare agents or industrial chemical
compounds; (4) government research used an epidemiological approach, but
little research on treatment was funded; and (5) most of the ongoing
epidemiological research focusing on the prevalence or causes of Gulf
War-related illnesses will not provide conclusive answers, particularly
in identifying risk factors or potential causes due to formidable
methodological and data problems.

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

 REPORTNUM:  T-NSIAD-98-104
     TITLE:  Gulf War Illnesses: Federal Research Strategy Needs 
             Reexamination
      DATE:  02/24/98
   SUBJECT:  Medical research
             Military personnel
             Disease detection or diagnosis
             Diseases
             Armed forces abroad
             Health research programs
             Chemical warfare
             Hazardous substances
             Biological warfare
IDENTIFIER:  Persian Gulf War
             Gulf War Syndrome
             DOD Persian Gulf Registry of Unit Locations
             
******************************************************************
** 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.                         **
******************************************************************


Cover
================================================================ COVER


Before the Subcommittee on Human Resources, Committee on Government
Reform and Oversight, House of Representatives

For Release on Delivery
Expected at
10:00 a.m., EST
Tuesday,
February 24, 1998

GULF WAR ILLNESSES - FEDERAL
RESEARCH STRATEGY NEEDS
REEXAMINATION

Statement of Donna Heivilin, Director of Planning and Reporting,
National Security and International Affairs Division

GAO/T-NSIAD-98-104

GAO/NSIAD-98-104T

Gulf War Illnesses

(713019)


Abbreviations
=============================================================== ABBREV

  CIA - Central Intelligence Agency
  VA - Veterans Affairs
  DOD - Department of Defense
  PGVCB - Persian Gulf Veterans' Coordinating Board
  PTSD - posttraumatic stress disorder

============================================================ Chapter 0

Mr.  Chairman and Members of the Subcommittee: 

I am pleased to be here today to discuss our evaluation of the
federal strategy to research Gulf War illnesses.  We reported our
findings on this strategy in June 1997 as part of our response to a
congressional mandate regarding the government's clinical care and
medical research programs relating to illnesses suffered by Gulf War
veterans.\1 I will first summarize our findings and provide some
background information on the government's research program before
giving you the details on our findings. 


--------------------
\1 Gulf War Illnesses:  Improved Monitoring of Clinical Progress and
Reexamination of Research Emphasis Are Needed (GAO/NSIAD-97-163, June
23, 1997). 


   RESULTS IN BRIEF
---------------------------------------------------------- Chapter 0:1

In short, we found that

(1) the government was not proactive in researching Gulf War
illnesses;

(2) the government's early research emphasized stress as a cause for
Gulf War veterans' illnesses and gave other hypotheses, such as
multiple chemical sensitivity, little attention;

(3) in contrast, the private sector pursued research on the health
effects of low-level exposures to certain chemical warfare agents or
industrial chemical compounds;

(4) government research used an epidemiological approach, but little
research on treatment was funded; and

(5) most of the ongoing epidemiological research focusing on the
prevalence or causes of Gulf War-related illnesses will not provide
conclusive answers, particularly in identifying risk factors or
potential causes due to formidable methodological and data problems. 


   BACKGROUND
---------------------------------------------------------- Chapter 0:2

U.S.  troops were reportedly exposed before, during, and after the
Gulf War to a variety of potentially hazardous substances.  These
substances include decontaminating and protective compounds used
without proper safeguards (particularly decontaminating solution 2,
or DS2, and chemical agent resistant coating); diesel fuel used as a
sand suppressant in and around encampments, fuel oil used to burn
human waste; fuel in shower water; and leaded vehicle exhaust used to
dry sleeping bags.  Other potential hazards included infectious
diseases (most prominently leishmaniasis, a parasitic infection);
pyridostigmine bromide and vaccines (to protect against chemical and
biological weapons); depleted uranium (contained in certain
ammunition and in residues from the use of this ammunition);
pesticides and insect repellents, chemical and biological warfare
agents; and compounds and particulate matter contained in the
extensive smoke from the oil-well fires at the end of the war.  Over
100,000 of the approximately 700,000 Gulf War veterans have
participated in health examination programs that the Department of
Defense (DOD) and the Department of Veterans Affairs (VA) established
between 1992 and 1994.  Of those veterans examined by DOD and VA,
nearly 90 percent have reported a wide array of health complaints and
disabling conditions, including fatigue, muscle and joint pain,
gastrointestinal complaints, headaches, depression, neurologic and
neurocognitive impairments, memory loss, shortness of breath, and
sleep disturbances.  Some of the veterans fear that they are
suffering from chronic disabling conditions because of exposure
during the war to substances with known or suspected health effects. 

The federal government, primarily through DOD and VA, has sponsored a
variety of research on Gulf War veterans' illnesses.  DOD's research
is one component of a broader agenda coordinated under the aegis of
the Persian Gulf Veterans' Coordinating Board (PGVCB), which
comprises the Secretaries of the Department of Health and Human
Services, VA, and DOD.  The details of this agenda are described in
the PGVCB publication entitled
A Working Plan for Research on Persian Gulf Veterans' Illnesses.\2
This agenda was developed in response to an Institute of Medicine
conclusion that the DOD and VA should determine specific research
questions that need to be answered and design epidemiologic research
to these questions.  Accordingly, most of the research sponsored
under this agenda is characterized by PGVCB as epidemiological. 

The objectives of epidemiologic research are to determine the extent
of diseases and illness in the population or subpopulations, the
causes of disease and its modes of transmission, the natural history
of disease, and the basis for developing preventive strategies or
interventions.\3 To conduct such research, investigators must follow
a few basic generally accepted principles. 

First, they must specify diagnostic criteria to (1) reliably
determine who has the disease or condition being studied and who does
not and (2) select appropriate controls (people who do not have the
disease or condition). 

Second, the investigators must have valid and reliable methods of
collecting data on the past exposure(s) of those in the study and
possible factors that may have caused the symptoms.  The need for
accurate, dose-specific exposure information is particularly critical
when low-level or intermittent exposure to drugs, chemicals, or air
pollutants is possible.  It is important not only to assess the
presence or absence of exposure but also to characterize the
intensity and duration of exposure.  To the extent that the actual
exposure of individuals is misclassified, it is difficult to detect
any effects of the exposure.  Another means of linking environmental
factors to disease is to determine whether or not evidence shows that
as the exposure increases, the risk of disease also increases. 
However, this dose-response pattern can be detected only if the
degree of exposure among different groups can be determined. 

Finally, in addition to specific case definition and dose-specific
exposure information with known accuracy, it is important that a
sufficient number of persons be studied to have a reasonable
likelihood of detecting any relationship between exposures and
disease.  To the extent that this relationship is subtle or obscured
in particular investigations by "loose" case definition (that is, a
case definition that is too broad and encompasses different types of
illnesses) or problems in measuring exposure, larger samples would be
required.  For example, the Institute of Medicine noted that "very
large groups must be studied in order to identify the small risks
associated with low levels of exposure, whereas a relatively small
study may be able to detect the effect of heavy or sustained exposure
to a toxic substance.  In this way, a study's precision or
statistical power is also linked to the extent of the exposure and
the accuracy of its measurement.  Inaccurate assessment of exposure
can obscure the existence of such a trend and thus make it less
likely that a true risk will be identified."\4 Similarly, if an
exposure had an effect only on a particular birth defect for example,
this effect might be missed by studying all birth defects as a group. 


--------------------
\2 A Working Plan for Research on Persian Gulf Veterans' Illnesses
(First Revision), Department of Veterans Affairs, November 1996. 

\3 A.  M.  Lilienfeld and D.  E.  Lilienfeld, Foundations of
Epidemiology (New York:  Oxford University Press, 1980). 

\4 Veterans and Agent Orange:  Update 1996 (Washington, D.C.: 
Institute of Medicine, 1996), pp.  99-100. 


   GOVERNMENT WAS NOT PROACTIVE IN
   RESEARCHING CAUSES OF GULF WAR
   VETERANS' ILLNESSES
---------------------------------------------------------- Chapter 0:3

Although Gulf War veterans' health problems began surfacing in the
early 1990s, the vast majority of research was not initiated until
1994 or later.  And much of that research responded to legislative
requirements or external reviewers' recommendations.  As noted by
external reviewers, since federal research goals and objectives were
not identified until 1995, after most research activities had been
initiated, the research reflects a rationalization of ongoing
activity rather than a research management strategy. 

The government's 3-year delay complicated the researchers' tasks and
limited the amount of completed research available.  Of the 91
studies receiving federal funding, over 70 had not been completed at
the time of our review.  The results of some studies will not be
available until after 2000. 

By the time research was accelerated and broadened, opportunities had
been missed to collect critical data that researchers cannot
accurately reconstruct.  Even efforts to measure the chemical content
of the oil-fire smoke, begun only 2 months after the fires began
burning, were initiated after most troops had left the affected areas
and the climatological dynamics were different.  Consequently,
researchers had to use statistical models of the behavior of smoke
plumes in order to infer the ground-level exposures experienced by
the large numbers of troops who had departed by the time they began
collecting data.  Even if such models could accurately explain the
behavior of the smoke plumes, they had not been validated as measures
of individual exposure, and their accuracy for this purpose could not
be presumed.  Similar and even more serious problems were caused in
the measurement of other exposures by the failure to collect data
promptly and maintain adequate records.\5

The delay in starting research has also hindered accurate reporting
of exposures by Gulf War veterans.  At the time of our review, 6
years after the war ended, questionnaires were being distributed
requesting information from veterans on their exposures to certain
agents during the war. 


--------------------
\5 See Defense Health Care:  Medical Surveillance Has Improved Since
the Gulf War, but Results in Bosnia Are Mixed (GAO/NSIAD-97-136, May
13, 1997) and Institute of Medicine, Health Consequences of Service
During the Persian Gulf War:  Recommendations for Research and
Information Systems, p.  5 (Washington, D.C.:  National Academy
Press), 1996. 


   INITIAL GOVERNMENT RESEARCH
   EMPHASIZED STRESS; OTHER
   HYPOTHESES WERE NOT PURSUED
   UNTIL LATER
---------------------------------------------------------- Chapter 0:4

Early federal research appeared to emphasize risks associated with
psychological factors such as stress.  To support this emphasis, DOD
pointed out that the psychological state of mind can influence
physical well-being.  DOD also pointed to a recent argument that from
the American Civil War onward (and perhaps even earlier), a small
number of veterans have reacted to the stress of war by suffering
symptoms similar to those reported by some Gulf War veterans.\6

Of the 19 studies initiated before 1994, roughly half focused on
exposures to stress or the potential for posttraumatic stress
disorder (PTSD) among returning troops.\7 As late as December 1996,
the Presidential Advisory Committee noted that "stress is the risk
factor funded for the greatest fraction of total - 32 studies (30
percent)."\8

While research on exposures to stress received early emphasis, other
hypotheses have received scant support.  In its Final Report, the
Institute of Medicine discusses the evidence for a number of disease
hypotheses, including multiple chemical sensitivity, fibromyalgia,
and organophosphate-induced delayed neuropathy.  However, the federal
research program has supported only one study of the relationship
between symptoms reported by veterans and fibromyalgia.  In addition,
prior to October 1996, only one of the studies initiated in response
to Gulf War veterans' illnesses focused on the health effects of
potential exposures to chemical warfare agents.\9 While multiple
studies of the role of stress in the veterans' illnesses have been
supported with federal research dollars, other hypotheses have been
pursued largely outside the federal research program. 

Although veterans raised concerns about potential chemical exposures
soon after the war, the federal research plan was not modified to
include an investigation of these concerns until 1996, when DOD
acknowledged potential exposures to chemical agents at Khamisiyah,
Iraq.  The failure to fund such research cannot be traced to an
absence of investigator-initiated submissions.  According to DOD
officials, three recently funded proposals on low-level chemical
exposure had previously been rejected.\10


--------------------
\6 K.C.  Hyams et al., "War Syndromes and Their Evaluation:  From
Civil War to the Persian Gulf War," Annals of Internal Medicine, vol. 
125 (1996), pp.  398-405. 

\7 An additional 3 of the 19 studies did not provide information
about veterans' illnesses but were instead building databases or
methods to be used in later studies.  Notably, according to PGVCB,
none of these three studies had been completed as of June 1997. 

\8 Presidential Advisory Committee on Gulf War Veterans' Illnesses,
Final Report, p.  34 (Washington D.C.:GPO), December 1996. 

\9 This study of the impacts of sulfur mustard agent is a
collaborative effort between the Portland VA Medical Center and the
Oregon Health Sciences University.  The principal investigator for
the study pointed out that the possibility of chemical warfare
exposure seemed plausible even in 1994 when he sought initial funding
for this research. 

\10 The three previously unfunded proposals address central nervous
system targets for organophosphates, development of a DNA-based
method for assessing exposures to mustard agent, and work on the
pharmacokinetics of the nerve agent VX. 


   PRIVATE SECTOR PURSUED VARIETY
   OF HYPOTHESES
---------------------------------------------------------- Chapter 0:5

A substantial body of research suggests that low-level exposures to
chemical warfare agents or chemically related compounds, such as
certain pesticides, are associated with delayed or long-term health
effects.  For example, abundant evidence from animal experiments,
studies of accidental human exposures, and epidemiologic studies of
humans shows that low-level exposures to certain organophosphorus
compounds, including sarin nerve agents to which our troops may have
been exposed, can cause delayed, chronic neurotoxic effects.  This
syndrome is characterized by clinical signs and symptoms manifested 4
to 21 days after exposure to organophosphate compounds.  The symptoms
of delayed neurotoxicity can take at least two forms:  (1) a single
large dose may cause nerve damage with paralysis and later spastic
movement and (2) repetitive low doses may damage the brain, causing
impaired concentration and memory, depression, fatigue, and
irritability.  These delayed symptoms may be permanent. 

As early as the 1950s, studies demonstrated that repeated oral and
subcutaneous exposures to neurotoxic organophosphates produced
delayed neurotoxic effects in rats and mice.  In addition, German
personnel who were exposed to nerve agents during World War II
displayed signs and symptoms of neurological problems even 5 to 10
years after their last exposure.  Long-term abnormal neurological and
psychiatric symptoms as well as disturbed brain wave patterns have
also been seen in workers exposed to sarin in sarin manufacturing
plants.\11 The same abnormal brain wave disturbances were produced
experimentally in primates by exposing them to low doses of sarin.\12

Delayed, chronic neurotoxic effects were also seen in animal
experiments after the administration of organophosphates.\13 These
effects include difficulty in walking and paralysis.  In recent
experiments, animals given a low dosage of the nerve agent sarin for
10 days showed no signs of immediate illness but developed delayed
chronic neurotoxicity after
2 weeks.\14

It has been suggested that the ill-defined symptoms experienced by
Gulf War veterans may be due in part to organophosphate-induced
delayed neuropathy.\15 This hypothesis was tested in a privately
supported epidemiological study of Gulf War veterans.\16 In addition
to clarifying the patterns among veterans' symptoms by use of
statistical factor analysis, this study concluded that vague symptoms
of the ill veterans are associated with objective brain and nerve
damage compatible with the known chronic effects of exposures to low
levels of organophosphates.\17 It further linked the veterans'
illnesses to exposure to combinations of chemicals, including nerve
agents, pesticides in flea collars; DEET and highly concentrated
insect repellents; and pyridostigmine bromide tablets. 

Finally, research that we reviewed also indicates that agents like
pyridostigmine bromide, which some Gulf War veterans took to protect
themselves against the immediate, life-threatening effects of nerve
agents, may alter the metabolism of organophosphates in ways that
activate their delayed, chronic effects on the brain.\18

Moreover, exposure to combinations of organophosphates and related
chemicals like pyridostigmine or DEET has been shown in animal
studies to be far more likely to cause morbidity and mortality than
any of the chemicals acting alone.\19

Despite the fact that in 1994, Congress directed DOD and VA to
research treatments for ailing Gulf War veterans, such research has
largely not taken place.  While 61 of the 91 federally sponsored
studies (67 percent) were classified as epidemiological by the PGVCB,
only three of the studies had focused primarily on identification and
improvement of treatments for these illnesses. 


--------------------
\11 F.  H.  Duffy et al., "Long-Term Effects of an Organophosphate
Upon the Human Electroencephalogram," Toxicology and Applied
Pharmacology, vol.  47 (1979), pp.  161-176, and F.R.  Sidell,"Soman
and Sarin:  Clinical Manifestations and Treatment of Accidental
Poisoning by Organophosphates," Clinical Toxicology, vol.  7 (1979),
pp.  1-17. 

\12 J.  L.  Burchfield et al., "Persistent Effect of Sarin and
Diodrin Upon the Primate Electroencephalogram," Toxicology and
Applied Pharmacology, vol.  35 (1976), pp.  365-379. 

\13 M.  B.  Abou-Donia, "Organophosphorus Ester-induced Delayed
Neurotoxicity," Annual Review of Pharmacology Toxicology, vol.  21
(1981), pp.  511-548, and M.  K.  Johnson, "The Target for Initiation
of Delayed Neurotoxicity by Organophosphorus Esters:  Biochemical
Studies and Neurotoxicological Applications," Review of Biochemistry
and Toxicology, vol.  4 (1982), pp.  141-212. 

\14 K.  Husain et al., "Assessing Delayed Neurotoxicity in Rodents
after Nerve Gas Exposure," Defense Science Journal, vol.  44 (1994),
pp.  161-164; K.  Husain et al., "Delayed Neurotoxic Effects of Sarin
in Mice After Repeated Inhalation Exposure," Journal of Applied
Toxicology, vol.  13 (1993), pp.  143-145; and K.  Husain et al., "A
Comparative Study of Delayed Neurotoxicity in Hens Following Repeated
Administration of Organophosphorus Compounds," Indian Journal of
Physiology and Pharmacology, vol.  39 (1995), pp.  47-50. 

\15 R.  W.  Haley et al., "Preliminary Findings of Studies on the
Gulf War Syndrome," Presentations to the Intergovernmental
Coordinating Board for the Gulf War Illnesses and the Staff of the
Presidential Advisory Committee on Gulf War Veterans' Illnesses,"
September 16, 1995, and R.  W.  Haley, "Organophosphate-Induced
Delayed Neurotoxicity," Internal Medicine Grand Rounds, University of
Texas Southwestern Medical Center, Dallas, Texas, October 10, 1996. 

\16 This research, conducted at the University of Texas Southwestern
Medical Center, has been supported in part by funding from the Perot
Foundation. 

\17 R.  W.  Haley et al., "Is There a Gulf War Syndrome?  Searching
for Syndromes by Factor Analysis of Symptoms," Journal of American
Medical Association, vol.  277 (1997), pp.  215-222; R.  W.  Haley et
al., "Evaluation of Neurologic Function in Gulf War Veterans:  A
Blinded Case-Control Study," Journal of American Medical Association,
vol.  277 (1997), pp.  223-230; and R.  W.  Haley et al.,
"Self-reported Exposure to Neurotoxic Chemical Combinations in the
Gulf War:  A Cross-sectional Epidemiologic Study," Journal of
American Medical Association, vol.  277 (1997), pp.  231-237. 

\18 C.  N.  Pope and S.  Padilla, "Potentiation of Organophosphorus
Delayed Neurotoxicity," Journal of Toxicology and Environmental
Health, vol.  31 (1990), pp.  261-273. 

\19 M.  B.  Abou-Donia et al., "Increased Neurotoxicity Following
Concurrent Exposure to Pyridostigmine Bromide, DEET, and
Chlorpyrifos," Fundamental of Applied Toxicology, vol.  34 (1996),
pp.  201-222 and M.  B.  Abou-Donia et al., "Neurotoxicity Resulting
From Coexposure to Pyridostigmine Bromide, Deet, and Permethrin,"
Journal of Toxicology and Environmental Health, vol.  48 (1996), pp. 
35-56. 


   FORMIDABLE METHODOLOGICAL
   PROBLEMS HAVE HAMPERED RESEARCH
---------------------------------------------------------- Chapter 0:6

Our review indicated that most of the epidemiological studies have
been hampered by data problems and methodological limitations and
consequently may not provide conclusive answers in response to their
stated objectives, particularly in identifying risk factors or
potential causes. 


      MEASUREMENT OF EXPOSURES IS
      PROBLEMATIC
-------------------------------------------------------- Chapter 0:6.1

The research program to answer basic questions about the illnesses
that afflict Gulf War veterans has at least three major problems in
linking exposures to observed illness or symptoms.  First, it is
extremely difficult to gather information about unplanned exposures
(for example, oil-fire smoke and insects) that may have occurred in
the Gulf.  And DOD has acknowledged that records of planned or
intentional exposures (for example, the use of vaccines and
pyridostigmine bromide to protect against chemical/biological warfare
agents) were inadequate.  Second, the veterans were typically exposed
to a wide array of agents with commonly accepted health effects,
making it difficult to isolate and characterize the effects of
individual factors or to study their combined effects.  Third, the
passage of time following these exposures has made it increasingly
difficult to have confidence in any information gathered through
retrospective questioning of veterans.\20

In part, the latter difficulty was created by the delayed release of
information about detection of chemical warfare agents during the war
as well as the delayed collection of exposure data.  Five years
passed before DOD acknowledged that American soldiers may have been
exposed to chemical warfare agents shortly after the war ended in
1991 (at the Khamisiyah site).  Moreover, although chemical
detections by Czech forces are regarded as valid by DOD, the origin
of the detected chemical agents has not been identified by either DOD
or the Central Intelligence Agency (CIA).  In the face of denials by
DOD officials, several researchers told us that they had considered
it pointless to pursue hypotheses that the symptoms may have been
associated with exposures to chemical weapons. 

When we asked investigators responsible for federally funded
epidemiological research how they were collecting data on the various
elements to which Gulf veterans may have been exposed, they indicated
that they had no means other than self-reports for measuring most of
these elements.  This reliance on self-reports was not much less for
elements such as vaccines, for which the opportunity for record
keeping clearly existed.\21

Two problems are associated with reliance on self-reports for
exposure assessments.  First, recalled information may be inaccurate
or biased after such a long time period; that is, some veterans may
not remember that they were exposed to particular factors, while
others may not have been exposed but nonetheless inaccurately report
that they were.  Information also may be biased if, for example,
veterans who became sick following the war recalled their exposures
earlier, more often, or differently from veterans who had not become
sick.  Second, there is often no straightforward way to test the
validity of self-reported exposure information, making it impossible
to separate bias from actual differences in exposure frequency. 

Several investigators were also relying on a model developed by the
U.S.  Army Environmental Hygiene Agency for assessing exposures to
components of oil-fire smoke through the combination of unit location
data with information from models of the distribution of oil-fire
smoke.  However, this model requires the use of unit location as a
proxy for exposure, and the validity of this approach is unknown. 
The Presidential Advisory Committee has noted, "DOD's Persian Gulf
Registry of Unit Locations lacks the precision and detail necessary
to be an effective tool for the investigation of exposure incidents."


--------------------
\20 Large numbers of veterans questioned during their participation
in the VA's health registry examination program reported they did not
know whether they were exposed to certain agents.  "Don't know"
responses were greatest for nerve gas (64.9 percent), mustard gas
(60.2 percent), depleted uranium (52.5 percent), chemical-agent
resistant coating (47.8 percent), microwaves (32.8 percent), paints
or solvents (24.9 percent), and pyridostigmine (21.1 percent).  To
the extent that a response of some kind reflects greater certainty,
veterans were more confident in their reports regarding smoke from
tent heaters, passive smoking, diesel or other petrochemical fumes,
skin exposure to fuel, pesticides in cream or spray form, and burning
trash or feces, each of which resulted in fewer than 11 percent of
respondents reporting "don't know." While such confidence does not
necessarily mean that the reports are accurate, the lack of
confidence in responding to questions about some exposures raises
questions about studies relying on self-reports to assess these
exposures. 

\21 Defense Health Care:  Medical Surveillance Improved Since Gulf
War, but Mixed Results in Bosnia (GAO/NSIAD-97-136). 


      CASE DEFINITION IS
      COMPLICATED BY PRESENCE OF
      NONSPECIFIC SYMPTOMS
-------------------------------------------------------- Chapter 0:6.2

Another major hurdle to the development of a successful research
agenda has been the difficulty in classifying symptoms into one or
more distinct illnesses.  Some veterans complain of gastrointestinal
pain, others report musculoskeletal pain or weakness, and still
others report emotional or neurological symptoms.  As explained
previously, development of one or more specific case definition is
essential to conducting certain types of epidemiological studies. 

The VA collected some data on symptoms beginning in 1992 with the
initiation of its registry.  However, these efforts to collect
information about symptoms and exposures from registry participants
were limited and nonspecific.  This constrained VA's potential use of
the information for improving understanding of the patterns of
veterans' complaints.  These data limitations were unfortunate, as
detailed information about symptoms and exposures might have yielded
earlier, more reliable analyses of the nature and causes of veterans'
complaints and could have also assisted in developing working case
definitions. 

We also found that both the federally supported projects and the
federal registry programs have generally failed to study the
conjunction of multiple symptoms in individual veterans.  Articles
and briefing documents that we obtained from DOD and VA reported
findings that addressed only the incidence of single symptoms and
diagnoses.  There were two exceptions.  First, for an Air National
Guard unit in Pennsylvania, the Center for Disease Control and
Prevention developed an operational case definition, which was quite
similar to the case definition of chronic fatigue syndrome.  Second,
the studies conducted by Haley et al.  also focused on identifying
symptom clusters. 

For those ongoing, epidemiological studies that were built on
case-control designs, we asked about how a case was defined.  The
specificity of this definition is important because a vague case
definition can lead to considering multiple kinds of illnesses
together.  When this is done, it is not surprising to find no
commonality of experience among the cases.  Moreover, the use of
specific case definition is particularly critical to achieving
meaningful results within this type of research design.  At the same
time, for the case definition to be relevant, it must fit the
symptoms described by an important portion of the group being
studied. 


      SAMPLE SIZE
-------------------------------------------------------- Chapter 0:6.3

Most of the investigators we interviewed took steps to estimate the
size of the sample they would require to have a reasonable
expectation of detecting the effects of exposures to hazardous
substances.  However, many other variables were involved in such
calculations, for example, the prevalence of exposures, some of which
were unknown at the time the studies were planned.  Thus, they had to
make estimates within somewhat broad parameters. 

Although steps were clearly taken to plan for an adequate sample
size, some investigators reported difficulty in locating subjects due
to factors beyond their control, such as the rate of referrals from
VA examination centers or the rate of identification of subjects that
fit highly specific case definitions.  Moreover, other studies, such
as those on specific birth defects, required extremely large samples. 


   CONCLUSIONS
---------------------------------------------------------- Chapter 0:7

The ongoing epidemiological research cannot provide precise,
accurate, and conclusive answers regarding the causes of veterans'
illnesses because of researchers' methodological problems as well as
the following: 

  -- Researchers have found it extremely difficult to gather
     information about many key exposures.  For example, medical
     records of the use of pyridostigmine bromide tablets and
     vaccinations to protect against chemical/biological warfare
     exposures were inadequate. 

  -- Gulf War veterans were typically exposed to a wide array of
     agents, making it difficult to isolate and characterize the
     effects of individual agents or to study their combined effects. 

  -- Most of the epidemiological studies on Gulf War veterans'
     illnesses have relied only on self-reports for measuring most of
     the agents to which veterans might have been exposed. 

  -- The information gathered from Gulf War veterans years after the
     war may be inaccurate or biased.  There is often no
     straightforward way to test the validity of self-reported
     exposure information, making it impossible to separate bias in
     recalled information from actual differences in the frequency of
     exposures.  As a result, findings from these studies may be
     spurious or equivocal. 

  -- Classifying Gulf War veterans' symptoms and identifying their
     illnesses have been difficult.  From the outset, the symptoms
     reported have been varied and difficult to classify into one or
     more distinct groups.  Moreover, several different diagnoses
     might provide plausible explanations for some of the specific
     health complaints.  It has thus been difficult to develop one or
     more working case definitions to describe veterans undiagnosed
     complaints. 


   RECOMMENDATIONS
---------------------------------------------------------- Chapter 0:8

Because of the numbers of veterans who have experienced illnesses
that might be related to their service during the Gulf War, we
recommended in our report that the Secretary of Defense, with the
Secretary of Veterans Affairs, give greater priority to research on
effective treatment for ill veterans and on low-level exposures to
chemicals and other agents as well as their interactive effects and
less priority to further epidemiological studies. 


-------------------------------------------------------- Chapter 0:8.1

Mr.  Chairman, that concludes my prepared remarks.  I will be happy
to answer any questions you may have. 

*** End of document. ***