The last three decades have seen an often acrimonious debate about the effects of the military use of herbicides in Vietnam and, in particular, the potential adverse long-term health effects on those who may have been exposed to these herbicides. The Vietnam War was fought in a jungle environment that provided cover to the enemy and made battlefield observations difficult. Military strategists decided to use herbicides to remove foliage along key roads and waterways, defoliate areas surrounding enemy bases and supply and communications routes, improve visibility in heavily canopied hardwood forests and coastal mangrove forests, and destroy enemy subsistence crops such as rice.
Herbicide missions were executed primarily by U.S. Air Force (USAF) personnel involved in Operation Ranch Hand using specially modified C-123 aircraft. From 1962 to 1971, Operation Ranch Hand disseminated over 19 million gallons of herbicides in Vietnam, of which at least 12 million gallons were Agent Orange, the name given to a 50:50 mixture of the herbicides 2,4-D and 2,4,5-T. Health concerns over Agent Orange exposure have centered on a contaminant of 2,4,5-T: 2,3,7,8-tetrachlorodibenzo-p-dioxin; often abbreviated as TCDD and colloquially referred to as dioxin. Operation Ranch Hand was the first (and only) large-scale experience with chemical defoliants in U.S. military operations.
Based on the assumption that Ranch Hand personnel were likely to have been among the most highly herbicide-exposed U.S. service members in Vietnam, this group was deemed desirable as a study cohort for the evaluation of the frequency and nature of adverse health effects related to Agent Orange and other military herbicide exposures. The USAF made a commitment to Congress and the White House in 1979 to conduct an epidemiologic study of this group. Dubbed
the Air Force Health Study (AFHS), it comprises the Ranch Hand veterans as well as a comparison group of USAF personnel who were also in the Vietnam theater, but who were presumed not to have been exposed to herbicides. The study includes a morbidity component with periodic physical examinations and other data collection, a companion examination of reproductive outcomes in subjects’ offspring, and a records-based mortality-only component.
AFHS funding has been provided by direct congressional appropriation as a line item in the Department of Defense budget. Approximately $143 million has been spent or allocated for the study.
Major data gathering activities for the AFHS have been completed, and investigators are finishing research projects. The study is currently scheduled to end on September 30, 2006.
INTENT AND GOALS OF THE NATIONAL ACADEMIES STUDY
Public Law 108-183, the Veterans Benefits Act of 2003, directed the Secretary of Veterans Affairs to contract with the National Academy of Sciences (NAS) to address several questions regarding the appropriate disposition of the AFHS.
Section 602(c) of the law charged the NAS to evaluate the following:
(1) The scientific merit of retaining and maintaining the medical records, other study data, and laboratory specimens collected in the course of the Air Force Health Study after the currently scheduled termination date of the study in 2006.
(2) Whether or not any obstacles exist to retaining and maintaining the medical records, other study data, and laboratory specimens referred to in paragraph (1), including privacy concerns.
(3) The advisability of providing independent oversight of the medical records, other study data, and laboratory specimens referred to in paragraph (1), and of any further study of such records, data, and specimens, and, if so, the mechanism for providing such oversight.
(4) The advisability of extending the Air Force Health Study, including the potential value and relevance of extending the study, the potential cost of extending the study, and the federal or nonfederal entity best suited to continue the study if extended.
(5) The advisability of making the laboratory specimens of the Air Force Health Study available for independent research, including the potential value and relevance of such research, and the potential cost of such research.
This report, prepared by the Committee on the Disposition of the Air Force Health Study, provides responses to these charges.
CHARACTERISTICS OF THE AFHS DATA ASSETS
Most of the AFHS research is focused on the morbidity component of the study, where exceptionally detailed data gathering took place. The initial physical examination and surveying of AFHS subjects (Cycle 1) was conducted in 1982 and subsequent cycles were conducted in 1985, 1987, 1992, 1997, and 2002. Data from the physical examinations include blood draws, urine and semen collections, skin and fat biopsies, stool smears, spirometry, chest X rays, electrocardiograms (ECGs), dermatology and peripheral vascular examinations, neurological assessments, psychological testing, and many other clinical endpoints. Subjects were asked to provide their medical records, as well as those of their children (up to age 17), their present partner, and any previous partner. Questionnaires eliciting information on education; employment; income; marital and fertility history; child and family health; health habits; recreation, leisure, and physical activities; toxic substances exposure; military experience; and wartime herbicide exposure were also administered. In 1982, 1,046 of the so-called Ranch Hands and 1,223 comparison subjects participated in the first of the physical exam cycles. Twenty years later, 777 Ranch Hands and 1,174 comparison subjects completed the sixth cycle. In all, 2,758 subjects participated in at least one cycle exam. Reproductive data were gathered and coded on 9,921 conceptions and 8,100 live births. The mortality component of the study has followed over 20,000 Vietnam War-era veterans with service in Southeast Asia for nearly 25 years.
More than 86,000 biologic specimens have been collected over the course of the study; approximately half of these are serum. Blood was collected in all six cycles and serum stored; semen was obtained in the first cycle and urine in Cycles 1–3 (1982, 1985, and 1987). Adipose tissues collected during Cycle 5 are currently held at UC Davis. In the last (sixth) cycle, whole blood was also stored.
Participation rates in the morbidity study have been relatively high, particularly among Ranch Hand subjects, ranging from 87 percent at the baseline examination to 74 percent at the fifth follow-up. All told, the AFHS has more data points and a higher rate of follow-up than the Framingham Heart Study.
Chapter 2 of the report provides background information on Operation Ranch Hand and on the origins, development, implementation, findings, and impact of the AFHS. It also briefly summarizes the results of AFHS reports and papers.
FINDINGS, CONCLUSIONS, AND RECOMMENDATIONS
Box ES-1 summarizes the committee’s response to the charge contained in the Veterans Benefits Act of 2003. Specifics are discussed below.
There is scientific merit in retaining and maintaining the medical records, other study data, and laboratory specimens collected in the course of the Air Force Health Study after the study’s currently scheduled termination date: no other epidemiologic dataset on Vietnam veterans contains as detailed information over as long a time period, the data appear to be of high quality and the specimens well preserved, and analysis of the assets has contributed to the literature addressing the health of Vietnam veterans.
Obstacles exist to retaining and maintaining the AFHS data assets. These relate to factors intrinsic to the study’s design, resulting from implementation decisions made by the investigators, relating to documentation and organization of the data assets, and addressing the preservation of the privacy of the study subjects and the confidentiality of their personal information. The committee believes that the identified obstacles are surmountable.
Further study of the AFHS medical records, other study data, and laboratory specimens is advisable. This should be accomplished by making these materials available for research via a custodian that takes an active role in fostering use of the assets. Five years after the chosen custodian assumes responsibility, a committee should be convened to evaluate whether any further support should be extended to the maintenance of access to the data or the biospecimens.
The potential value and relevance of further study of the AFHS data assets rest in the application of the results of future health research on the data assets. This research could encompass using novel analysis approaches, employing new technology and techniques, and examining data and outcomes not evaluated to date. The cost of such work will vary greatly, depending on the research question that is addressed.
The committee cannot offer a specific recommendation on the federal or nonfederal entity best suited to continue the study of the AFHS data assets but has identified a number of options that could be pursued successfully.
Independent oversight of future research using the AFHS data assets is advisable, and should be provided through the review of proposals for scientific merit and adherence to ethical, legal, and related considerations by an Institutional Review Board and, separately, an advisory and oversight board. Additionally, research should be carried out in a manner transparent to study subjects, through systematic communication of research plans and results.
The Database and Specimens Repository
Chapter 3 describes the elements comprising AFHS data holdings and their accessibility to outside researchers; Chapter 4 details the AFHS specimen collection, including information related to specimen gathering, processing, and storage, and obstacles and limitations to the collection’s future use. Together, these two chapters recapitulate and expand on the text of the committee’s interim letter report, which was issued in November 2005.
The committee found that while the medical records, other study data, and laboratory specimens collected in the course of the AFHS have been properly maintained, they are not currently organized and documented in a manner that allows them to be easily understood, evaluated, managed, or analyzed by persons outside of the AFHS. It therefore concluded that the present state of the documentation and organization of these assets was an obstacle to retaining and maintaining them. The committee recommended that action be taken prior to the currently scheduled termination date to reorganize and document the study’s medical records, other study data, and laboratory specimens in a form and format that allows them to be easily understood, evaluated, managed, or analyzed by persons outside of the AFHS. Several specific actions to address the shortcomings are detailed in these chapters. Reorganization and documentation of the data assets is a crucial prerequisite to their future use.
The committee believes that it is incumbent on the USAF as the current custodian of the research materials to ensure their proper documentation and organization. It therefore recommended that, if available AFHS program funds are not sufficient to accomplish the actions it proposes, then supplemental funding be provided to carry out such work in a complete and timely manner.
Value of the Research Assets
Chapter 5 builds on the material presented in the earlier chapters to address the scientific merit of retaining and maintaining the AFHS data and specimen holdings. The following three general conclusions are drawn:
The AFHS data assets are unique: no other epidemiologic dataset on Vietnam veterans contains as detailed information over as long a time period.
The data collected by the AFHS appear to be of high quality and the specimens appear to be well preserved.
Analysis of the AFHS data assets has contributed to the literature addressing the health of Vietnam veterans.
It follows that there is scientific merit in retaining and maintaining these resources after the study’s currently scheduled termination date.
Some limitations to the further use of AFHS materials were identified, including those related to the design and execution of the study, and to privacy, security, and other ethical, legal, and social issues. The committee concluded that these concerns and issues are not an intrinsic obstacle to retaining and maintaining the assets after the currently scheduled termination date. However, attention to such concerns and issues must play a central role in decisions regarding their future disposition.
Because there is scientific merit in retaining and maintaining these assets and the identified obstacles to these are believed to be surmountable, the committee concluded that further study of the AFHS medical records, other study data, and laboratory specimens is advisable.
To date, research results published by study investigators have been largely limited to outcomes related to the examination of the possible long-term health effects of wartime exposure to the herbicides used in Vietnam. However, the data assets can be used to examine a far broader range of health questions. The potential value and relevance of extending the study of the AFHS data assets rest in taking full advantage of the available information and in the application of the results of future research on the assets. This research could encompass
reanalysis of outcomes examined by the AFHS using different assumptions and approaches than have been applied to date,
new analyses of the medical records and other study data that examine questions that were not addressed in the AFHS,
new studies of the collected biospecimens that take advantage of advances in technology and science to conduct analyses that were not contemplated in the AFHS protocol,
expansion of the study’s period of analysis through follow-up of the cohorts using publicly available information, and
additional follow-up of health outcomes in AFHS participants.
The AFHS dataset has several weaknesses that limit its utility as a means of evaluating the health impacts of Agent Orange exposure. These include the inherently small size of the cohort,1 the fact that the cohort is unrepresentative of in-theater veterans, lack of any biomarkers of herbicide exposure other than TCDD, little information on subjects’ locations in the theater of operations, unavailability of detailed exposure histories, and possible herbicide exposures in the comparison population. It would thus be a mistake to view the committee’s conclusions as an indication that the AFHS data assets are a definitive source of information on this topic. Instead, the data and specimens are pieces of evidence that can be
put to use in the effort to understand the determinants of good and ill health in Vietnam veterans and to address other scientific questions. Some of these may not be directly related to herbicide exposure but may still be important for furthering knowledge about veterans’ health and the natural history and risks of disease.
While some analyses of AFHS data assets have been carried out by outside researchers, the capacity of the assets to inform the study of other health issues is largely untapped and unknown. Until they are made available to the research community, their worth as a source of information on veterans’ and men’s health will remain open to question.
Options and Recommendations for Further Study
The committee took a broad view of what constituted extension of the AFHS. In Chapter 6, it offers a number of options and recommendations regarding the further study of the data assets. These include observations on the maintenance and use of the assets, the protection of the interests of the study subjects, the characteristics of a good future custodian of the assets, mechanisms for providing oversight of prospective research, alternatives for the management of the assets, and the costs associated with the assets’ maintenance and use. Throughout, it addresses the obstacles to conducting further research and possible means of overcoming them.
The two primary considerations in evaluating the disposition of the AFHS hard copy records are compliance with the Federal Records Act and retention for further research purposes. Many of the hard copy records have been scanned and the images stored in electronic files in Portable Document Format (PDF) that were subjected to quality control checks for readability. The committee recommends that the hard copy originals of these saved documents be destroyed unless the National Archivist concludes that they must be maintained or they are subject to other records retention requirements. No matter what decision is made by the National Archives regarding the retention of the records, the committee recommends that a separate electronic copy of these records should be kept for possible placement with another custodian.
Recommendations intended to facilitate the future use of other nonelectronic assets are also presented. In brief, the committee believes that such data assets should be written to electronic form where a mature technology exists to do so and where the assets have future research value. The X-ray films can be readily converted to digital format—this should be done with acceptable quality control, the resource properly documented, and thereafter the original films disposed of according to standard practices. As no widespread technology exists for ECG strip digitization, the originals should be retained. The AFHS should consult with the National Institute of Dental and Craniofacial Research on whether the video tapes of subjects’ teeth collected in the course of the study have future research
value—if the decision is made to retain them, they should be digitized for ease of long-term management and storage.
Five alternatives for the future of the AFHS data assets were identified:
Render the data assets to the National Archives for evaluation and possible retention.
Identify or establish a research entity for the management and dissemination of existing data assets.
Have that entity also continue passive data collection on AFHS subjects.
Have the entity maintain contact with the AFHS subjects and conduct or facilitate further active data collection.
Extend the AFHS as it is presently constituted.
The medical records and other study data collected in the course of the AFHS will be rendered to the National Archives for evaluation because this is required by statute. Should the National Archives choose to retain them, the records and other study data may be made available to the public as is at some point in the future. The Archives does not, however, have exclusive rights to the data. No matter what it decides, copies of the data may be made available through other channels. If these assets, along with the study’s biospecimen collection, are transferred to a new custodian, other alternatives that would enhance access to the assets would be possible. A custodian could offer research support of various forms that would facilitate their use; or the custodian or outside researchers could supplement the assets with data from publicly available sources or the study participants themselves.
The committee concludes that it is advisable to extend the AFHS by making the database and associated biospecimen collection available for study via a custodian that should take an active role in fostering research on the assets. The committee is thus not recommending that the AFHS be extended in its current form as a research study characterized by periodic, cohort-wide data and biospecimen collection, and analysis by a dedicated team of researchers. Instead, the committee believes that the present data assets should be made available to the whole of the research community. This will allow them to continue to be used to help understand the health consequences of service in Vietnam and also to address other scientific questions related to the health of veterans as well as aging persons.
The committee also concludes that it is advisable to allow for the possibility of collecting additional data and specimens, work that could be conducted by the custodian or by independent researchers. The committee does not explicitly endorse the collection of further morbidity or mortality data but believes that it is appropriate to leave the option open should both the need and support for such work be identified. Developing a future research agenda or setting research priorities is beyond the charge of the committee.
The committee identified several characteristics of a good custodian of the assets. Among these is the ability to satisfy operational requirements related to the supervision of and access to the data and biological materials. Also important is a commitment to address the complex ethical, legal, social, and related issues that arise when managing epidemiologic records and associated biospecimens. The committee specifically recommends that the successor custodian be able to demonstrate in advance its capacity to protect the privacy and security of the research participants and their data to the greatest extent possible. The custodian also should comply with the information and publication restrictions in Section 308(d) of the Public Health Act and should seek a certificate of confidentiality under Section 301(d) of the Act.
The concept of individual informed consent is particularly salient in the setting of the AFHS where thousands of research participants devoted a great deal of time and underwent significant inconvenience and discomfort to contribute to the study. Their wishes must be considered in deciding whether and how to use these data in future research. In recognition of this, the committee recommends that prior to the end of the study, the AFHS should notify the study participants of the following:
The study, as currently constituted, is ending.
The assets will be transferred to a successor custodian, the characteristics of which will be described to the extent feasible.
The successor custodian will seek additional consent for future research involving identifiable data from the research participants, which will cover the topics enumerated below, prior to conducting any research.
This notification should offer the participants the opportunity to decide that information and specimens relating to them should not be transferred.
The committee also recommends that the successor custodian obtain new informed consent that includes
notice of the types of participant data and samples that will be maintained by the custodian;
the procedural protections that will be provided for the data and specimens, including access policies, and oversight;
the specific privacy and security protections that will protect the data and specimens;
the types of studies that the participant is willing to have his data and specimens used for;
whether the participant is willing to be approached in the future by investigators to seek participation in additional studies, including information regarding notice and recontact criteria; and
which individual study results the participant will and will not receive.
If health information regarding the partners and children of the AFHS research participants is made available for future research, consent, and where appropriate, parental permission and assent, must be sought from these individuals. Conducting future research on the data assets is predicated on gaining permission for such work from the subjects. If a significant number of subjects choose not to make their information available, it will greatly limit the future usefulness of the assets.
The committee believes that independent oversight of the AFHS data assets is advisable. Specifically, it recommends that the successor custodian of the AFHS data and biospecimens should commit to reviews of future research involving the assets by an Institutional Review Board constituted in accordance with the Common Rule, create and maintain an independent advisory and oversight board to provide guidance on the conduct of future research and to review and evaluate proposals for use of the data and biologic samples, and develop and implement strategies for apprising the research subjects about the ways in which their information is used and the scientific discoveries that result. These commitments should be a prerequisite to the transfer of the assets. The committee suggests that the advisory and oversight board’s membership consist of 8–12 established scientists representing a range of appropriate disciplines such as herbicide exposure health effects, military health, epidemiology, and aging. Two lay members—study subjects if possible—should be included to provide insights from the participant and veteran perspectives.
The committee identified the following seven options for future management of the data assets:
Use the National Archives data management and distribution mechanisms.
Use the existing AFHS infrastructure within the USAF.
Use Department of Defense (DOD) epidemiologic data management and distribution mechanisms.
Use Department of Veterans Affairs or VA-affiliated epidemiologic data management and distribution mechanisms.
Use National Institutes of Health or Centers for Disease Control and Prevention epidemiologic data management and distribution mechanisms.
Use the Institute of Medicine (IOM) Medical Follow-up Agency.
Identify a new custodian through a competitive process.
The committee does not offer a recommendation on the specific federal or nonfederal entity best suited to assume responsibility for the AFHS data assets. All of the listed options have advantages and disadvantages, and all could be made to work with the appropriate provision of funds, commitment on the part of the custodian, and implementation of the recommendations offered in the report. The committee believes that the AFHS data would be unlikely to realize their research potential if they were only available from the National Archives, which
would neither promote their use nor supported analysis of them and which could not be a home for the specimens. The USAF and DOD manage and analyze other epidemiologic data assets comprising military populations but have not expressed an interest in the AFHS materials, which address a population that is no longer on active duty and exposures that are unlikely to be an issue for future operations. The CDC and NIH have extensive infrastructures in place to manage the dissemination of epidemiologic data and biospecimens in a manner that protects the interests of research subjects. It is unclear, however, whether these bodies would like to assume custody of the data assets or whether the AFHS subjects would be comfortable with any shift in focus in the use of their materials that might be occasioned by a CDC or NIH custodianship. Using a competitive process to identify a new custodian would ensure that a motivated party with an explicit data management plan would assume responsibility for the assets. It would, though, present logistical challenges to carry out this process in a timely manner.
The Department of Veterans Affairs maintains Epidemiological Research and Information Centers (ERICs) under their Cooperative Studies Program. The committee considers two of the centers—the Massachusetts Veterans Epidemiology Research and Information Center (MAVERIC) and the Seattle Epidemiologic Research and Information Center (Seattle ERIC)—to be good candidates for custodianship, in part because they manage data assets that are in some respects similar to AFHS’s and make these assets available to independent, outside researchers under controlled circumstances. The IOM’s Medical Follow-up Agency (MFUA) is also identified as a good candidate. MFUA conducts epidemiologic studies of military and veterans’ health, focusing on records-based research—primarily retrospective cohort studies of morbidity outcomes and mortality. It also maintains a collection of data on study populations of former military personnel that it makes available to independent researchers.
These three candidates share several positive attributes. MAVERIC, Seattle ERIC, and MFUA all have prior experience in veterans’ health studies, collecting and storing epidemiologic data, disseminating it to independent researchers, fostering collaborations with those researchers, maintaining quality control over the studies that use their datasets, and publishing results in the peer-reviewed literature. All have military or veteran health issues as the central focus of their operations, but also conduct other types of studies with their data assets. To the committee’s knowledge, none has experienced a breach in their system of protecting subjects’ privacy and the security of their data. These entities are not, however, without weaknesses. Among the concerns noted by the committee is whether the ERICs have the interest or ability to take on the additional responsibility of the AFHS data assets and how MFUA would manage the biospecimens in the absence of its own repository.
A common issue for these potential custodians is a secure source of funding. Assigning responsibility for the assets to a government entity without also providing a budget allocation would sap funds in a time of constrained resources,
and MFUA or other nongovernment entities could not take on the assets without financial support.
The committee was asked to consider the costs of extending the study of the AFHS data assets. These comprise the costs of maintaining and providing access to the assets and the costs of doing research on them. The primary physical costs of maintaining AFHS’s medical records and other study data and making them available to researchers are associated with storage and handling of the electronic materials. Personnel expenses related to database management are best expressed in terms of workforce needs, as the allocations required to cover direct and indirect costs vary by location and type of institution (government, nonprofit, or commercial) as well as over time. Typical needs would be for time from a principal investigator or project director, data manager, programmers, clerical personnel, and perhaps a webmaster or participant liaison. With the possible exception of a data manager, none of these persons would need to devote their full time to these tasks. The cost for physical maintenance of the AFHS biospecimens will vary considerably, again depending upon the type of institution in which the collection is housed and whether an existing facility is used or a new one created.
It is difficult for the committee to present a bottom-line estimate for the total cost of implementing its recommendations because of the number of uncertainties and variables involved. Roughly speaking, the committee estimates that $150,000–300,000 per year would be required to support the custodian’s database management responsibilities and an additional $200,000 or more per year for proper maintenance of the biospecimens. The actual amount would depend critically on the custodian chosen and the number and type of studies proposed for and conducted with the data assets. First year costs would be higher because a number of one-time-only tasks would have to be performed to set up the operation. The committee assumes that the custodian will impose fees for the provision of database and biospecimen access to recover at least some their costs. It is important, especially in the early years of this new phase in the life of the AFHS resources, that such fees be set at a level that will facilitate use of the resource.
The cost of conducting new studies using the AFHS data assets will greatly depend on their scope and whether they require the use of other databases, analyze biospecimens, or contact subjects. Pilot studies and empirical analyses of the existing AFHS dataset could have relatively modest needs for support—tens of thousands of dollars—while investigations that include collecting additional data and specimens could potentially cost millions of dollars. However, with the exception of the program proposed below, these costs will be the responsibility of prospective researchers, who will need to obtain their own support from funding sources.
The committee believes that funds are needed to establish the AFHS data assets as a resource for external researchers. Its judgment is that a limited-life, small grant program is the best way to accomplish this goal. The committee anticipates that this program will stimulate prospective researchers to also seek ex-
ternal funding from other existing sources for further or more in-depth projects. The committee therefore recommends that Congress should allocate a minimum of $250,000 per year for three years to cover the direct costs of small grants for secondary data analysis or pilot projects using the data assets of the AFHS. This allocation would provide seed money for one to three investigations each year. These funds should be in addition to previously planned or contemplated support of Vietnam veterans’ health research—funding should not be diverted from other research efforts for this initiative.
Although the committee is enthusiastic about the potential for future studies of the AFHS data assets, it understands that the viability of such work is not assured. It therefore believes that it is appropriate to revisit the question of support for further work after the committee’s recommendations have been implemented and have had time to play out in the research realm. A 5-year commitment—that is, two years after the last small grants proposed above are made—should be sufficient to establish whether the AFHS resources have value and relevance as a resource. The committee therefore recommends that five years after the chosen custodian assumes responsibility a committee should be convened to evaluate the potential value and relevance of extending further support to the maintenance of access to the data or the biospecimens collected in the course of the AFHS.
The costs of properly documenting the data assets, making them available to the research community, and implementing the proposed 5-year program to encourage their use are small in comparison with the government’s investment of ~$143 million to date. The potential benefits of these actions, while subject to considerable uncertainty, appear to greatly outweigh the costs.