Epilepsy is a substantial concern among United States service members and veterans (hereafter referred to as veterans as service members ultimately become veterans). Among post-9/11 veterans, prior studies have found a prevalence of epilepsy of approximating 10.6/1000 [1]. In particular, posttraumatic epilepsy (PTE), which occurs following a traumatic brain injury (TBI), is more common among veterans who served during Operation Enduring Freedom or Operation Iraqi Freedom (post-9/11) because of the higher incidence of TBI and blast injuries compared with earlier conflicts [1-3]. The annual cost of epilepsy care in the United States has been estimated at US $12.5 billion [4], with higher costs for those newly diagnosed, those with seizures refractory to anticonvulsant medication treatment, and those with comorbid health conditions [5,6]. Indirect costs (eg, caretaker’s time away from work) comprise most of the total cost of epilepsy care [7,8]. Estimates of caregiver-related costs based on self-report are high [9].

Poorly controlled epilepsy is associated with injury, disability, mortality, and poor quality of life [10-15]. Veterans with epilepsy have also reported role limitations caused by impairments related to physical, mental, emotional, and social functions [16,17]. Of the post-9/11 deployed veterans, 11% to 23% experienced TBI, most of which were mild TBIs (mTBIs) [2,18-21]. Even among those with mTBI, the risk of developing PTE is elevated (adjusted odds ratio 1.28, 95% CI 1.07-1.53) [1]. With hundreds of thousands of veterans exposed to mTBI, even a small elevated risk has the potential to impact thousands of individuals, caregivers, and families.

Our understanding of epilepsy and the impact of PTE in post-9/11 combat veterans is limited. Data from the Department of Veterans Affairs’ (VA) comprehensive TBI evaluation have shown that veterans with PTE have significantly higher cognitive, affective, somatosensory, and vestibular symptom burdens than those with mTBI only. A survey-based study found that veterans with PTE had significantly lower quality of life, social support, and family resilience scores compared with those with epilepsy, mTBI, and control participants with neither TBI nor epilepsy [22]. Given the relatively young age of post-9/11 veterans, the costs of care (both personal and financial) may profoundly affect the lives of these veterans, their caregivers or families, and the health care systems on which they depend for care.

Informal caregivers—family and friends who provide assistance for people living with chronic health conditions or disabilities living in the community—in general, have significantly higher levels of stress, depression, and lower levels of physical health than noncaregivers [23-25]. Although positive effects of caregiving have been reported [26], caregiving can negatively influence caregiver health, and these effects are likely unique to specific caregiver populations. The effects of caregiving for veterans with PTE are understudied. Qualitative interviews indicated that caregivers experience stress, anxiety, and secondary posttraumatic stress disorder because of frequent seizure or suicide watch when caring for a veteran with PTE. These data suggest unmet needs for support and services owing to a lack of tailored programs and support for veterans with PTE and their caregivers. Most prior studies that have assessed the dyadic relationship between caregivers and care recipients have focused on children with epilepsy; therefore, assessments of the quality of life for veterans with PTE and their caregivers are needed.

The available evidence regarding military caregiving suggests that these caregivers may experience a health decline for a longer period than any other caregiver population [27-29]. The young age and long life span of veterans with TBI, PTE, and other catastrophic injuries make this population unique in light of the existing literature, which has focused on family caregivers of patients in hospice or patients with a diagnosis of cancer, dementia, Parkinson disease, Alzheimer disease, and other diseases or disabilities [30-32]. The full range of experiences and needs of caregivers of veterans with TBI in general and of those with PTE in particular has not been fully explored. For example, other researchers have highlighted the need to understand family and relationship quality [33], caregiver resilience [34], and positive aspects of the caregiver experience [35] to support veterans with TBI and their caregivers.

Given these gaps in knowledge, the purpose of this study is to longitudinally examine factors associated with quality of life and health outcomes for veterans with epilepsy or PTE and their caregivers. To develop appropriate interventions to improve health outcomes for veterans and their families, we must understand the veteran-caregiver dynamics.

On the basis of the existing literature and our preliminary studies, we hypothesize that caregivers of veterans with PTE encounter unique challenges in their caregiving role, resulting in more intense caregiving and a higher burden than caregivers of other post-9/11 US veterans, including those with TBI only or nontraumatic epilepsy (NTE). Furthermore, we hypothesize that caregivers of veterans with PTE and associated epilepsy comorbidities (eg, depression and headache) have a higher burden and poorer health than caregivers of veterans with PTE but without associated comorbidities. Therefore, we expect that the health and burden trajectories of caregivers of veterans with PTE will differ from caregivers of veterans without PTE and that there will be distinctions between these groups based on whether the veteran has associated comorbidities. Finally, we hypothesize that the caregiver’s burden and own health trajectory will influence the health status of the veteran at baseline and their health trajectories during follow-up.

This study will evaluate the health, well-being, and quality of life of veterans and their caregivers with the following aims: (1) describe and compare the health and quality of life of veterans and caregivers of veterans with and without PTE; (2) evaluate the change in available support and unmet needs for services among caregivers of post-9/11 veterans with PTE over a 2-year period and compare supports and unmet needs with those without PTE; and (3) identify veteran and caregiver characteristics associated with the 2-year health trajectories of caregivers and veterans with PTE compared with veterans without PTE.

This study is guided by a biopsychosocial model of caregiving [36] and the Military and Veteran Caregiver Experience Map [37]. Figure 1 [36,37] is an adaptation of the Military and Veteran Caregiver Experience Map developed over the course of several years by the Elizabeth Dole Foundation and military and veteran caregiver (MVC) stakeholders [37] and Raina’s synthesis of theoretical frameworks for the impact of caregiving [36]. The model posits that baseline characteristics (ie, social and economic status, family structure, cultural support systems, and veteran health and wellness) influence the caregiver’s ability to meet the new demands of caregiving, which may lead to caregiver stress or strain depending on the extent to which their current identity and roles are altered by caregiving requirements. Over time, the caregiver may shift priorities and seek help within current social or family circles or the health care system and may develop new coping skills. When the caregiver is unable to shift priorities or obtain the needed support within social relationships or the health care system, there is a negative impact on caregiver well-being, continued dysfunction, and diminished veteran, caregiver, and family function, which can then lead to a negative impact on baseline influences and a negative spiral of health and well-being. When the caregiver is able to adapt or cope with new roles and responsibilities, there is a positive impact on veterans, caregivers, and family well-being, as well as a positive impact on baseline characteristics with a positive spiral for health and well-being for veterans, caregivers, and families. Survey and ecological momentary assessment (EMA) data collection will obtain data that address each component of this model and use modeling techniques to explicate the trajectories of veteran and caregiver health or well-being and quality of life.

We will conduct a prospective cohort study of the health and quality of life among 4 groups of veterans and their caregivers: (1) PTE, (2) NTE, (3) TBI only, and (4) neither epilepsy nor TBI. We will recruit participants from previous studies of PTE or TBI and MVC outcomes conducted by Dr Pugh and Dr Delgado; specifically, we will sample from among respondents who agreed to be contacted for future research studies. Data sources will include existing study data from these prior projects supplemented with primary data collected for this study in 2 forms: (1) survey data collected at baseline; (2) daily brief reporting using EMA [38] over the course of 30 days, with the option to continue this data collection for up to 2 years; and (3) qualitative data collected via interviews.

We established a Service Member, Veteran, and Caregiver Community Stakeholders Group for this study, which consists of 3 veterans with the goal of having 1 with TBI, 1 with epilepsy, and 1 without TBI or epilepsy, and 3 caregivers of veterans with TBI or epilepsy. We recruited locally using existing relationships and used state and national service members, veterans, and caregiver organizations as recruitment sources (eg, Wounded Warrior Programs, American Red Cross’ Military and Veteran Caregiver Network, and Iraq and Afghanistan Veterans of America). This group will help identify special issues in addressing veteran and caregiver needs and ensure that our approach is consistent with community needs and values. Members will receive stipends for their participation. We will consult with the community stakeholders to finalize the veteran and caregiver surveys and EMA items, and we will discuss findings with them to assure our interpretation of the results resonate with their experiences and understanding of living with and providing care to veterans with TBI and PTE. Full community stakeholders group meetings are planned quarterly throughout the study period.

We will recruit veteran and caregivers for this study from among a group of participants who agreed to be recontacted from the Veterans Posttraumatic Epilepsy Study (VPES; 189/210, 90% response rate in follow-up interviews), the pilot TBI Caregiver Study, and the Military and Veteran Caregiver Health and Well-being (MVCHW) study. We will request participation from the primary caregiver for VPES participants, veterans for the MVCHW, and recontact both veterans and the primary caregiver from the pilot TBI Caregiver Study. The group of veterans with neither TBI nor epilepsy includes people with primary mental health comorbidities or devastating diseases such as amyotrophic lateral sclerosis. These groups will allow us to examine variations in caregiver burden, stressors, resources used, and unmet needs based on epilepsy and comorbidity patterns.

As we are recruiting from 3 prior studies, the inclusion criteria and sampling frame for each of these studies vary slightly and are summarized below.

The VPES included veterans who were deployed in support of post-9/11 conflicts and received ≥2 years of VA care between fiscal year 2002 and fiscal year 2014 with at least one of those years after 2007, when VA began mandatory TBI screening. Epilepsy was identified using our validated algorithm (International Classification of Disease-9 or -10 diagnosis of epilepsy or convulsion and subsequent use of antiseizure medication) [1], with subsequent validation by medical chart abstraction. TBI was identified using the International Classification of Disease-9 or -10 algorithm developed by the Armed Forces Health Surveillance System [39], and self-reported lifetime TBI history based on the Ohio State TBI Identification measure [40]. Veterans with PTE, NTE, TBI, and controls (veterans with neither epilepsy nor TBI) were randomly selected for survey administration. Only those veterans who responded and agreed to be recontacted for future research and also reported having a caregiver were included in this study (269/326, 82.5% reported having a caregiver).

The MVCHW study recruited participants using email invitations sent out by MVC support organizations such as the Elizabeth Dole Foundation and Hearts of Valor in April 2017. Although no denominator was provided to the study team, 476 individuals responded and completed the web-based survey within a 3-month period without any incentive. Caregivers self-reported the types of conditions for which they provided care.

The pilot TBI Caregiver Study invited 186 veterans who were diagnosed with penetrating TBI (per the Armed Forces Health Surveillance System algorithm) and their caregivers via email. Upon validation of TBI severity using the Ohio State TBI Identification measure [40], it was found that all had moderate or severe TBI. Of the 186 invites, 66 (35.4%) veterans and caregivers responded, and 65 (34.9%) dyads agreed to be recontacted for future research.

For this study, we will use data on TBI and epilepsy history to classify all veteran-caregiver dyads into 1 of the 4 study groups. Only caregivers aged >18 years will be included. Although we will inquire about caregiving provided by children, we will not engage minors in the study.

We expect to achieve a response rate of 75% given that we will sample from among veterans who have already participated in a study and expressed a willingness to be recontacted. We also expect a high participation rate among caregivers, as the veteran has already participated and presumably will have told the caregiver about the study. We plan to enroll 97 veterans with PTE, 45 with NTE, 323 with TBI only, and 145 with no PTE and no TBI and their caregivers, totaling 610 veterans and 610 caregivers. Table 1 shows the number of participants in each of the 4 study groups.

We plan to compare veterans or caregivers of veterans with PTE to those with (1) NTE, (2) TBI only, and (3) no epilepsy and no TBI. We anticipate adequate statistical power for planned analyses using the survey data. For example, we have >80% power to detect a 2-point difference between veterans with PTE and those with NTE, the minimum size considered clinically meaningful for the Veterans RAND 12-Item Health Survey, with an SD as high as 4.25, even if we account for the correlation between veterans and caregivers (assumptions: cluster size of 2, intraclass correlation coefficient of 0.8 between veterans and caregivers, and α=.05) [41]. We assume similar variability among caregivers and therefore also anticipate adequate power when comparing the experiences of caregivers of veterans with PTE and NTE. We may not have adequate power to detect differences between groups when making comparisons based on health trajectory groups, given that we anticipate 3 groups, 2 of which may be moderate to small. However, our qualitative analysis will allow us an additional perspective in comparing PTE and NTE.

We will collect survey data from both veterans and caregivers following the Dillman Tailored Design Method for mail and web-based surveys to minimize nonresponse errors [42]. This approach involves a combination of data collection strategies to achieve a higher response rate. Our specific strategy will involve either mail or email, depending on what contact information we have for the veteran and caregiver, and is outlined in Figure 2.

To encourage participation, we will provide an incentive of US $25 for the baseline survey and US $2 per response for the first 30 days of EMA data collection. In addition, we will offer incentives for the remainder of the 2-year EMA period by entering participants in a drawing for 5 US $500 gift cards and 10 US $100 gift cards per quarter for veterans and caregivers. This approach was used in a previous study and improved response rates by over 60% (from 30% to 48%).

We will collect survey data from both veterans and their primary caregivers at baseline. Our caregiver survey was modeled based on the RAND Military Caregivers Study [23], the Behavioral Risk Factor Surveillance System (BRFSS) Caregiver Module [43], and the National Alliance for Caregiving survey [44]. The goal is to develop a comprehensive survey that assesses various aspects of the respondent’s health, quality of life, and service needs in a reasonable amount of time (≤30 minutes). The Service Member, Veteran, and Caregiver Community Stakeholders Group will review the baseline survey and recommend changes before it is fielded. Veterans and caregivers who complete the baseline survey will be invited to participate in EMA data collection for up to 2 years.

We will mail or email information on how to use the LifeData EMA platform to those who agreed to participate in the EMA. Participants will also be given contact information of study personnel who can assist the participant with the platform setup. Once set up, each participant will receive approximately 2 prompts daily to respond to for 30 days. The EMA will capture different measures daily. Participants can also add information to the EMA platform on concerns or unmet needs at any time using a pull mechanism where they initiate the data collection at the time of their choosing. Participants will receive US $2 per assessment completed for 30 days and also be eligible for drawings of larger incentives for continued EMA participation. The EMA platform (LifeData) is Health Insurance Portability and Accountability Act compliant and includes iOS, Android, and web-based platforms that can be used depending on the preference of the participant [45].

We will conduct qualitative interviews based on extreme variation sampling of results from early EMA data. Our descriptive phenomenological approach will be guided by the following research questions: (1) (Veteran) What are the epilepsy and TBI characteristics (ie, exposure, mechanism, and history) of veterans in a subsample of the cohort? (2) (Caregiver) How do veteran caregivers experience and perceive health, and what are the characteristics and unique factors that affect their overall health?

The goal of the qualitative component of this mixed methods study is to describe the insights and experiences of participants and to identify important elements of caregiving in this particular group that will help develop strategies for preventive care.

We plan to begin collecting baseline surveys and EMA data from January 2022. Baseline data collection will continue for approximately 9 months. EMA data collection will begin shortly after baseline surveys are launched and will continue for 2 years after baseline data collection is complete, or nearly 3 years in total. Depending on recruitment and saturation, qualitative interviews will begin approximately 6 months after EMA data collection begins and will continue for up to 2 years thereafter. The project will be complete by September 30, 2024.

We have drafted the baseline surveys, EMA measures, and interview guides and have begun working to secure human subject approval. We have started recruiting the Service Member, Veteran, and Caregiver Community Stakeholders Group members.

The University of Utah, Salt Lake City VA, and the University of Texas Health Science Center, San Antonio, approved this study. The study is currently being reviewed by the Human Research Protection Office of the United States Army Medical Research and Development Command. All data with protected health information or personal identifiable information will remain on a server that has federally approved encryption at VA Salt Lake City. Analyses of the deidentified data may occur at other sites.

The Institute of Medicine identified the evaluation of quality of life in epilepsy as a priority. Concern about the profound impact of epilepsy led to the Institute of Medicine (now the National Academy of Medicine) conducting an in-depth study of the public health dimensions of epilepsy. The study concluded that “comprehensive, timely, and accurate epilepsy surveillance data are needed to provide a better understanding of the burden of epilepsy, its risk factors and outcomes, and health service needs” [93]. This study will address this priority and the existing gap in the literature related to the needs and health impacts of caregiving for veterans with epilepsy, including both PTE and NTE.

Our long-term goal is to implement the findings of the proposed study in practice to improve the health and well-being of veterans with PTE and epilepsy and their caregivers. We follow the RAND Corporation’s blueprint for MVC research [94] as we develop our strategy. Specifically, of the 10 research objectives identified by their stakeholder panels, we will directly address the following 5 objectives: (1) describe caregivers (and unique characteristics of caregivers of PTE or epilepsy); (2) assess how the needs of care recipients change over time; (3) document the effects of caregiving on care recipient outcomes; (4) document the effects of caregiving on caregiver outcomes; and (5) examine factors associated with caregiver and care recipient harm.

We also believe that it is critical to plan for implementation as we conduct the research. To this end, we have developed a pathway for clinical implementation and improved health outcomes to guide the work, which considers benefits to both veterans and caregivers. Figure 4 illustrates the process by which we plan to implement the findings of the proposed study into practice to improve the health and well-being of veterans with PTE or epilepsy and their caregivers.

Our baseline survey will include validated measures that have been used in many previous studies and therefore will enable comparisons to other veteran and caregiver populations. In addition, EMA data will enhance our ability to use real time experiences of veterans and their caregivers to identify health outcomes, quality of life, unmet needs, and supports that may inform programs and interventions. By collecting numerous random assessments over a period of time, the EMA provides a representation of how individuals’ experiences and behaviors vary over time and across situations. For this study, the method focuses on symptoms, experiences with caregiving and adaptive behaviors, and aims to map the fluctuations of daily function within both caregivers and care recipients. This method is considered suitable for understanding daily changes in mood, stress, and medical symptomology [38,45,95]. EMA is a useful tool in research for collecting data which are often unobtainable by other methods and providing ecologic validity based on a reduction of recall bias and therefore assessment error. It has also been shown to provide clarity on individual pattern assessments that can otherwise be misunderstood. EMA will allow unprecedented information on unmet needs among MVCs and veterans with PTE.

To our knowledge, our analyses will be novel, as no one has examined longitudinal dyadic data using the changes in the extension of the bivariate dual latent change score model. The partitioning of prior levels and prior changes as distinct predictors of later changes is invaluable in helping us understand the underlying mechanisms that can be targeted for intervention. In addition, the study includes a broad range of measures that represent physical, mental, and emotional health of dyads. These data will enable us to not only evaluate individual-level impact on key outcomes but also to better understand the bidirectional health-related influences between veterans and caregivers. We expect that interventions designed to leverage relational strengths or address interpersonal barriers will ultimately result in better outcomes for both care recipients and caregivers [96-98].

Study limitations include the fact that the sample may not represent all geographic regions of the United States or all subgroups of veterans. In addition, whereas the study is longitudinal, follow-up will occur over a 2-year period, and therefore, we will not be able to assess the veterans’ and caregivers’ experiences over a longer period of time. As noted in the methods, our sample size may be too small to compare all 4 study groups on some measures, including health trajectory groups, depending on the distribution of responses and experiences. To keep the survey length manageable, we used screening versions of validated questionnaires for anxiety and depressive symptoms and therefore do not have more detailed measures of these experiences.

This study will fill a critical gap in the literature related to the needs, activities, and positive aspects of veterans with PTE and their caregivers. It will also strengthen our understanding of the relationship and positive aspects of caregiving for veterans with TBI and other injuries. We have designed the study using integrative frameworks and are leveraging a multidisciplinary team to use this work as the foundation for future interventions that will strengthen caregiver resilience and improve the health and quality of life of post-9/11 veterans and their family members who support them.