Long-term Health Consequences of Military Service

Military service can affect health years after military service, but studying these long-term health effects is challenging. US Department of Defense (DoD) records may not span sufficient time, because most service members leave active duty after only a few years in service. Long-term studies are needed on veterans who use US Department of Veterans Affairs (VA) health care. Now is an opportune time to conduct these studies, in part because the DoD and VA are moving toward electronic health record (EHR) interoperability. However, many veterans use private sector health care in addition to, or instead of, VA health care. The growing adoption of EHRs in the federal government and private sector and the development of a federal infrastructure for health information exchange offer an opportunity to study the long-term health consequences of military service.

We propose that the DoD and VA invite all members who are departing service to participate in a long-term observational study of their health that would span military service and postservice periods. The initiative would use an existing health information exchange program to capture health-related data from military service records and health care data from veterans’ place of care (ie, the DoD, VA, or private sector). It would impose minimal burden on participants but could provide substantial benefits to them, to other veterans, and to service members. It would require only modest additional resource commitments from the VA and DoD.

Challenges to Understanding the Long-term Health Effects of Military Service

The VA has long recognized that military service–related experiences may result in delayed health effects. Veterans may be eligible for VA health care for conditions that are possibly associated with long-past environmental exposures, such as exposure to Agent Orange during the Vietnam War; to toxic air contaminants released from an incinerator near Naval Facility Atsugi, Japan, between 1985 and 2001; and to contaminated drinking water on US Marine Corps Base Camp Lejeune, North Carolina, from the 1950s through the 1980s.

Concerns about the long-term health effects from environmental exposures also have surfaced for more recent military operations. For example, VA clinicians identified possible long-term consequences of war experiences in Afghanistan and Iraq, such as burn-pit exposure, traumatic brain injury, and mental health problems. ^{1 –3} Prompted by concerns about possible exposure to radiation in March 2011 during the reactor emergency in Fukushima, Japan, Congress required the DoD in 2014 to report on the adverse medical conditions of sailors serving on the USS Ronald Reagan during the postdisaster response. ⁴

For these environmental exposures and other service-related exposures (ie, experiences with possible health implications), studying long-term health effects is challenging. First, uncertainty about who was exposed and the magnitude of exposure is a challenge that can be attributed, in part, to the quality of personnel records, especially during war. For example, records of troop location in Vietnam during exposure to Agent Orange were notoriously poor. In determining health care benefits, the VA assumed that any service member who set foot in Vietnam between 1962 and 1975 or who served on a ship on inland waterways during this time was exposed. Record keeping on personnel location has improved since the Vietnam War, but more precise epidemiologic assessments are not always possible. For example, the Institute of Medicine could not definitively say whether or not burn-pit emissions in Afghanistan and Iraq contributed to adverse health events, in part because of inadequate data on troop location. ⁵

The DoD does, however, maintain several databases that are useful for the epidemiologic study of potentially harmful service-related exposures. The Defense Medical Surveillance System integrates data from many of these databases to facilitate health surveillance of DoD populations. The Defense Medical Surveillance System includes data on demographic and occupational characteristics, immunizations, clinical medical encounters (outpatient and inpatient), microbiological and other laboratory testing associated with medical encounters, deployment, pre- and postdeployment health assessments, reportable medical event reports, and casualties of all service members. ⁶

The Armed Forces Health Surveillance Branch (Surveillance Branch) of the Defense Health Agency, which is charged with conducting comprehensive health surveillance for the DoD, ⁷ maintains and uses the Defense Medical Surveillance System for routine public health surveillance and outbreak investigations. The Surveillance Branch also maintains the DoD Serum Repository (the Repository), which contains more than 55 million specimens that are linked to Defense Medical Surveillance System data, collected, and stored after mandatory HIV testing (every 2 years) and mandatory pre- and postdeployment blood draws. ⁶ Using the data on biological indicators of exposure to potentially harmful agents in the Repository and Defense Medical Surveillance System, military public health and research personnel conduct epidemiologic investigations. The Surveillance Branch and its partners distribute results of Defense Medical Surveillance System or the Repository analyses to military organizations and in peer-reviewed journals, including the Medical Surveillance Monthly Report, ⁸ which is published by the Surveillance Branch.

Several recent initiatives aim to improve the documentation of service-related exposures with possible adverse health effects. For example, the Individual Longitudinal Exposure Record is a pilot project led by the DoD and the VA to consolidate exposure information for each service member (from the EHR, Defense Medical Surveillance System, incident-specific reports, and other databases) for incorporation into the DoD and VA EHRs. Anticipated applications include the diagnosis and treatment of exposure-related conditions, the development of exposure cohorts for epidemiologic studies, and the determination of VA benefit eligibility.

Another effort to improve documentation of exposures is the Surveillance Branch-sponsored proof-of-concept research to identify molecular indicators of exposure to environmental and occupational hazards in the Repository specimens. Such biomarkers could allow more precise estimates of individual exposure to environmental hazards, enabling epidemiologic studies of individual-level exposure and health-related events in linked Defense Medical Surveillance System data.

As DoD records needed to assess potentially harmful exposures improve, however, a second major challenge is identifying long-term health effects, including the lag between exposure and effect. Defense Medical Surveillance System and other sources of surveillance data are useful for identifying short-term exposure effects for active duty service members, who use DoD-funded health care almost exclusively. Assessing long-term effects with DoD data only, however, is problematic. ⁹ Most service members separate from the active duty force before longer-term health effects may manifest. During fiscal years 2002-2013, the median age of separating service members was 25 years, and the median length of service was 4 years; only about 10% of service members were in active duty status for 20 years or longer (unpublished data, Defense Manpower Data Center).

For separating service members who use VA health care, interoperable DoD and VA EHRs could provide a longitudinal record spanning sufficient time to capture data on the long-term effects of military service. However, this interoperable record system would not capture data on all veterans, because many veterans use non-VA health care. For example, although more than 1.8 million veterans of the campaigns composing the Afghanistan and Iraq wars (ie, Operation Enduring Freedom, Operation Iraqi Freedom, and Operation New Dawn) are eligible for VA health care, only about 60% have used it. ¹⁰

Another approach to studying the long-term health effects of military service is the use of epidemiologic research studies, which examine exposures and outcomes in defined cohorts. The DoD and VA sponsor several such studies. For example, the Millennium Cohort Study is the DoD’s largest-ever prospective epidemiologic study. Initiated in 2001 to evaluate the long-term health effects of military service and deployment, it includes nearly 150 000 participants, half of whom served in Operation Enduring Freedom, Operation Iraqi Freedom, or Operation New Dawn. The study identified mental health outcomes, environmental effects, health behaviors, and other health effects associated with military service and deployment. ¹¹

The VA’s epidemiologic research portfolio includes several studies that focus on burn-pit exposures in Afghanistan and Iraq, a study of veterans who have traumatic brain injury or posttraumatic stress disorder, and the National Health Study for a New Generation of US Veterans, which began in 2009 and enrolls 60 000 veterans who served between 2001 and 2008. ¹² In a related approach, the VA also invites veterans to participate in VA registries for several environmental exposures to monitor the health of these populations.

Although these epidemiologic studies and registries facilitate long-term monitoring of participants, they do have limitations. First, adding assessments of new exposures or outcomes after initiation of the study may not be feasible. Second, study populations may not be large enough to assess all subgroups of concern. Third, resources may not always be available to initiate a new study each time that a cohort is identified that may have had harmful exposures.

Lifetime Surveillance Through Electronic Health Information Exchange

We propose a hybrid approach that combines features of routine health surveillance with features of epidemiologic research, which exploits recent advances in health information technology to further understanding of the long-term health effects of military service.

The proposal builds on an ongoing project: the Virtual Lifetime Electronic Record (Record), ¹³ a VA-DoD effort to enable the exchange of data for service members and veterans among VA, DoD, and civilian providers who care for these populations. (The initiative is related to, but distinct from, DoD and VA efforts to improve interoperability of their EHRs.) To facilitate data exchange, the Record uses the Nationwide Health Information Network (the Network). Sponsored by the Office of the National Coordinator for Health Information Technology in the US Department of Health and Human Services, the Network is a set of standards, services, and policies to enable secure health information exchange via the Internet. Ultimately, the goal is to gather data on health, benefits, and personnel information in the Record from the date of veterans’ entrance into military service, through their military career, and after leaving military service and receiving health care through the VA system or the private sector. ¹³

The VA and DoD have piloted the Record in several locations with private sector health care providers who agreed to participate, beginning in San Diego, California, in 2009. The Record is now available at all VA medical centers; veterans who receive VA health care may elect to share their health care information with participating private sector providers. DoD participation is growing, with pilots underway to enroll families of service members.

Although the Record was developed to improve continuity of care as veterans move among health care providers, it also provides an unprecedented opportunity to study the long-term health effects of military service. Upon separation from active duty status, service members could be asked to enroll in an observational study of their health. The study would use data already available in the veterans’ Record, as well as EHR data added to the Record after military service. Beyond EHR data, the DoD’s contribution to veterans’ Record could include data on deployment, occupational and environmental exposures, and possible service-related exposures, such as those already provided to the Defense Medical Surveillance System and new data sources under development (although the DoD EHR may ultimately include these data).

A database integrating all of these records, from the military service and postservice periods, would offer unique opportunities to assess the relevance of distant service-related exposures to medical conditions among veterans. If consent to participate also included de-identified use of the Repository specimens collected while on active duty, insights into biological precursors of medical conditions might be possible years before they manifest.

Assessing the long-term health effects of military service combines favorable features of routine health surveillance and epidemiologic research approaches. Similar to routine surveillance, this approach would require no additional burden on participants in the research study beyond that associated with participation in the Record. The only additional resources needed would be the new database that integrates data already collected as part of existing programs, such as the Record, and personnel required to maintain and use the database. Monitoring would not end after a predetermined number of years, as with a research study, but would continue to collect data on new cohorts as military service–related exposures change. Similar to a cohort study, however, the proposed approach would follow veterans for a long time—years or even decades—to assess long-term health effects.

The program could lead to cost savings if VA compensation for the health effects of harmful exposures could more precisely target the populations that may have been affected. Regardless, however, the need is clear for a more systematic and comprehensive approach to identifying potentially harmful service-related exposures and using that knowledge for treatment and prevention. When hypotheses are made about service-related exposures, it will be important to implement expedient and valid methods of investigating the relationship and determining whether additional study is needed or whether the data already support treatment and prevention measures. The need exists for a third option that bridges expensive epidemiologic studies that could last years and analyses of insufficient existing data sets.

The proposed program could provide this third option and complement epidemiologic studies. It could identify harmful exposures that otherwise would not have met the threshold for a dedicated epidemiologic study based on existing data sources or that may not have come to light because of the delayed or complex relationship between exposure and outcome, which can be identified through only a comprehensive data analysis. It could enable analysts to investigate associations in a rigorous manner using readily available data.

We propose that the VA and DoD establish a joint activity to lead the program. It would build on collaborative efforts already underway at VA and DoD. For example, a retrospective analysis combining EHR data from VA and DoD facilities in North Chicago, Illinois, demonstrated improved outbreak detection. ¹⁴ Based on this assessment, a subsequent project combined data from similar VA and DoD systems for all facilities nationwide. It found that the relative timeliness of detecting health events of regional and national importance improved by 92% in geographic areas with access to both DoD and VA facilities. ¹⁵

These and related efforts demonstrate the capability to merge DoD and VA health information and improve our understanding of common, immediate-term public health threats. Our proposal extends this model to longer-term health surveillance.

The initiative that we propose would face several challenges. One is securing participation from separating service members. The VA and DoD would need to communicate the potential benefits of the initiative to the veteran and service member populations. A possible benefit is that as knowledge of the health effects of military service improves, the health information exchange infrastructure could accelerate application of this knowledge to clinical practice at the point of care. For example, based on the veteran’s service history or medical status, prompts to the health care provider could suggest questions to ask or medical tests to administer to evaluate whether or not the veteran experienced an exposure of concern. The system could incorporate clinical guidance for health care providers as best management practices for veteran-specific issues are defined, which could be especially useful for private sector providers who are less familiar than VA and DoD providers with veterans’ health care needs.

Another challenge is establishing ethically appropriate and technically feasible procedures for obtaining informed consent to participate. Emerging models for informed consent—developed to cover broad possible uses of EHR data or biological samples in large populations—may be useful. ¹⁶ For example, in the dynamic consent model for sharing routinely collected EHR data, patients use secure Internet technology to electronically control consent over time and are made aware of how their data are being used. ¹⁷ If implemented in the proposed long-term veterans’ surveillance program, participants would not be asked at enrollment to consent to a broad range of possible uses of data and specimens. Rather, they could consent to specific uses of data and specimens and revise their consent whenever they choose.

Another challenge is the need for political will to implement the program. The joint VA-DoD activity that we propose, although modest in resource requirements, is an additional direction for a high-profile VA-DoD collaboration. A mandate for this new direction (ie, from Congress, the president, or the secretaries of the VA and DoD) likely would improve its chances of overcoming bureaucratic obstacles and boost its credibility among prospective participants.

Conclusion

Currently, a small portion of the medical record is shared among VA, DoD, and private sector providers and at only a few sites. However, if the Record achieves its vision, it will include the complete EHR and possibly other records to enable health information exchange across the Network. Health information exchange infrastructure is maturing rapidly in the United States, and now is the time for the VA and DoD to recognize long-term, population-level health monitoring as a needed Record capability and to begin defining the technical and programmatic requirements to achieve this goal.

A new generation of veterans needs care for conditions related to military service. For some of these conditions, knowledge of effective long-term management is limited. The VA and DoD have an opportunity to produce that knowledge and ensure its application wherever veterans seek care.