Electronic Health Record Concordance with Survey-Reported Military Sexual Trauma Among Younger Veterans: Associations with Health Care Utilization and Mental Health Diagnoses

Introduction

Sexual trauma is associated with myriad health risk factors and conditions, e.g., substance use, posttraumatic stress disorder (PTSD), eating disorders, chronic pain, sexual dysfunction, difficulty managing relationships, and mistrust of health care providers. ^{1 –3} Sexual harassment or assault that occurs during military service is known as military sexual trauma (MST). ⁴ Reports of MST are more prevalent among younger U.S. Veterans who served during recent conflicts in Iraq and Afghanistan (e.g., post-9/11) than among Veterans from earlier service eras. ⁵ MST is over 10× more prevalent among women who served post-9/11 than men from this service era, ⁶ yet based on the totals alone, nearly as many men screen positive for MST as women. MST is known to be a significant factor in the health and health care use of recent Veterans, and is particularly important in the care of women Veterans. ⁷ However, MST is often underreported by men and women alike due to various reasons ⁸ —the associated stigma of sexual assault, pressure not to report MST during military service as military chain of command was tasked with conducting MST investigations until 2022 despite the potential for commanders themselves to be MST perpetrators, and institutional betrayal (i.e., failure of a trusted institution [e.g., the military] on which one relies for safety to respond supportively or an institutional response that exacerbates harm). ^{9 –11} The prevalence of MST may also vary based on intersectional identities (e.g., race, ethnicity, sexual orientation, and gender identity), and MST underreporting is greater for Black, Latino/a, and LGBTQ+-identifying Veterans. ^12,13

The Veterans Healthcare Administration’s (VHA) efforts to manage the health-related impacts of MST have increased over the last few decades. In 1992, Congress passed a law mandating that VHA medical centers provide time-limited psychological treatment for women reporting MST; and by 1995, eligibility for MST care was also extended to men. ¹⁴ In 2000, the VHA mandated screening of all Veterans for a history of MST, and in 2002, VHA implemented a two-question MST screening using an electronic health record (EHR) clinical reminder to ensure that universal MST screening was enacted for all former service members. ¹⁵ Yet, based on the sensitivity of this information and historic norms of military culture that dissuade disclosing sexual trauma or other psychological concerns related to service, ^{16 –18} Veterans underreport MST. ^11,13,19,20 Thus, the VHA EHR does not accurately reflect the prevalence of service-related MST. ²¹ Nondisclosure of MST represents a potential barrier to receipt of appropriate and timely health care for service-related trauma and prevention of the multiple potential downstream health effects. ^7,22 For example, the VHA offers free MST-related care, which is only accessible if MST is disclosed and if Veterans consent to relevant documentation of MST in their EHR. Similarly, without MST disclosure, MST-related disability benefits cannot be accessed from the Veterans Benefit Administration. ²³ Although there have been several previous studies comparing sources of MST disclosure, ^{11 –13,19} the concordance of EHR MST screening with other sources of MST reporting has not been examined in equal groups of younger, men and women post-9/11 Veterans.

It is established that disclosure of MST is associated with increased use of all VHA health services, i.e., primary care, mental health, emergency department, social work, and other outpatient care, ^24,25 with a greater severity of both depression and PTSD symptoms, ²⁵ and also higher, additional costs of care. ²⁴ Yet, no known investigation has determined if there are health-related consequences for those with a history of MST, but who have not reported this experience on the VHA EHR screening. It is plausible that reporting MST on the EHR screen promotes better health care management. Thus, examining how the concordance of MST reports relates to VHA health care use and to the prevalence of trauma-related mental health diagnoses could help to identify more nuanced gaps in MST reporting, receipt of related care, and opportunities for improvement in VHA care delivery.

The primary objective of this cross-sectional investigation was first to replicate prior studies by determining the degree of concordance between MST documentation in the VHA EHR and MST reported on a survey of health risk factors among a geographically diverse sample of men and women post-9/11 Veterans from the Women Veterans Cohort Study (WVCS). Associations were tested separately by gender. We hypothesized that there would be significant discordance between EHR-documented and survey-reported MST for both men and women. As women are disproportionately affected by MST, a second objective was to extend previous work in this area to better understand the potential implications of MST concordance on VHA health care among women. To do so, we explored the relations of EHR and survey MST concordance with VHA health care utilization and the odds of having a diagnosis of PTSD, depression, or anxiety in the EHR. It was hypothesized that women with survey-reported MST only (i.e., without a positive MST screen in the EHR) would show lower health care utilization and a lower likelihood of all mental health diagnoses than those with concordant, positive MST reports in the EHR and the survey.

Methods

WVCS is a nationwide, prospective cohort study that was designed to identify gender-associated disparities in health outcomes and health care utilization among Veterans who were discharged from military service between October 1, 2001, and September 30, 2021, and who established care with a Department of Veterans Affairs (VA) medical center. ⁶ Data from the Department of Defense Manpower Data Center roster were used to derive the WVCS EHR cohort, which included all individuals who served in support of Operations Enduring Freedom, Iraqi Freedom, or New Dawn and enrolled in VHA health care thereafter. Demographic data were linked with diagnoses, pharmacy, and administrative records from VHA inpatient and outpatient encounters, which are housed in the VHA Corporate Data Warehouse.

The WVCS survey cohort consisted of post-9/11 Veterans who had received care at one of 5 VHA systems nationwide (Central Western Massachusetts, Connecticut, Durham, Indianapolis, and Los Angeles). All women and a random sample of men were recruited at a 2:1 ratio to complete the survey by mail. The 2:1 recruitment strategy was used to help ensure that a high proportion of women responded. Baseline surveys were distributed from 2016 to 2019, with 1-year follow-up surveys from 2017 to 2020. Overall, the baseline survey response rate was 24% (51% of whom were women), which is comparable to response rates from other article-based surveys completed by women Veterans. ²⁶ More information about the EHR and survey methods is available elsewhere. ^6,27

In both the EHR screening reminder and the WVCS survey, MST was defined by an affirmative response to at least one of the following: “While you were in the military, a) Did you receive uninvited and unwanted sexual attention such as touching, cornering, pressure for sexual favors, or verbal remarks? B) Did someone ever use force or threat of force to have sexual contact with you against your will?” ²⁸ In the EHR, Veterans may “decline to answer” the MST screening question, data which are coded in the Corporate Data Warehouse as missing. For EHR data, the MST report was based on the most recent positive or negative screening. The date of all MST screening was also included, which was used to calculate the average amount of time from first establishing VHA care to the date that MST disclosure was recorded in the EHR.

Demographic, service, and health-related data abstracted from the EHR included age at VHA enrollment, gender, race, ethnicity, marital status, education, military service branch (i.e., Air Force, Army, Coast Guard, Marines, Navy), service component (i.e., active duty, National Guard, or reserves), and rank (i.e., enlisted or officer). As MST is most likely to be disclosed within the initial two years of establishing VHA care, ¹⁹ health care utilization, abstracted from the EHR, was based on the number of primary care or mental health visits within that timeframe, the settings in which MST is most likely to be screened compared to other VHA clinical settings. ²⁹ PTSD, depression, and anxiety diagnoses were abstracted from the EHR based on International Classification of Diseases, Ninth Revision, Clinical Modification and Tenth Revision [ICD-9-CM and ICD-10-CM] codes and dates. Diagnoses were included if an ICD-9-CM or ICD-10 code was documented during one inpatient or two outpatient encounters. The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guidelines were followed to describe this investigation and the findings (Supplemental Data S1). ³⁰

Results

Discussion

This is the first geographically diverse investigation in which MST documented with the VHA EHR screening tool was compared to surveyed MST in a VA-enrolled sample with similar proportions of men and women post-9/11 Veterans. As hypothesized, ²² MST documented with the VHA EHR screening tool was less prevalent than MST captured with the WVCS survey. This discordance in MST reports was observed for both women (29% versus 64%) and men (2% versus 9%). Therefore, over 30% of women and about 7% of men who reported MST by survey lacked such documentation in their EHR.

Our findings from women were similar to results from two past investigations of women Veterans only: A nationally representative sample of 1287 women (mean age 50 years), in which 35% of women had EHR-documented MST, but 61% of women reported MST by phone, ¹² and a smaller, regional sample of 202 women Veterans (45–64 years of age and cisgender only), ¹¹ in which 40% of women had EHR-documented MST, but 74% of women reported MST on mail and internet-based surveys. In contrast, these results diverged from another study of 697 men and women Veterans (mean age 40 years, 50% women), in which 31% of women and 15% of men who reported MST on a questionnaire or by interview did not have a concordant, positive EHR MST screen. ¹³ However, that group was oversampled for concurrent PTSD diagnosis, a factor that can accompany MST and for which precipitating trauma, including MST, is necessarily assessed. ³² Given the more complex chronic disease and severe mental health comorbidities that accompany MST, ¹⁰ our findings replicate previous results from samples of women Veterans and re-emphasize a wide discordance in MST documentation for both men and women in the post-9/11 cohort. The observed discordance may be due to inaccurate MST reporting by Veterans and also to clinic or facility variations in MST screening and rescreening practices. ³³ All of these factors may prevent men and women from receiving the patient-centered, trauma-focused care that is a VHA priority and hinder clinicians’ opportunities to offer timely health care.

VHA health care utilization is 50% higher among Veterans with a positive versus negative EHR MST screen. ²⁴ Contrary to our hypothesis, among Veterans with a survey-reported MST history, those with and without a concordant, positive EHR MST screen demonstrated a similar number of primary care and mental health visits in the first two years of care, suggesting no difference in opportunities to disclose MST during routine screening in either setting. This result corresponds with those from a previous investigation of MST and VHA primary care utilization among men and women followed in four clinics. ³⁴ Although men and women should be screened at the same rate and each have ample opportunities to report MST in primary care and mental health clinics, a proportion is simply not prepared to do so when they are first screened, and men may be screened less often or less rigorously than women, or have gender-specific reasons for nondisclosure (i.e., greater perceived stigma or shame related to disclosing their trauma), ¹⁸ which could lead to the possible underdiagnosis or misdiagnosis of PTSD among men. There are two future research directions for better understanding the associations between EHR MST disclosure and health care utilization. First, the degree of personal interaction during MST screening may be influential. In another study, women Veterans who were screened for MST by clinicians were more likely to receive a mental health referral than those who completed a trauma questionnaire by self-report rather than a clinical interview. ³⁵ The manner in which MST screening is conducted may also vary by VHA facility, resulting in differences in accuracy and associated health care use. In a second future direction, it is possible that women who are uncomfortable with sharing their MST history may inadvertently downplay their associated mental health symptoms, resulting in similar health care use within the initial period of VHA care. Each of these hypotheses could be tested with future review of progress notes and other more detailed EHR documentation.

As hypothesized, women Veterans with a concordant, positive EHR screen and survey-reported MST were more likely to receive a diagnosis of PTSD and depression than those with survey-reported MST only. Women Veterans are high utilizers of VHA primary care and mental health care. ²⁷ Exploratory analyses comparing the dates of MST screens with dates of psychiatric diagnoses showed that positive MST screens generally precede any mental health diagnosis for women. Thus, MST disclosure could be a significant precipitant of women’s evaluation, referral, and treatment for such diagnoses. There is significant comorbidity between PTSD, depression, and anxiety among Veterans. ³⁶ Interestingly, women with concordant, positive MST reports did not show a greater likelihood of an anxiety disorder diagnosis compared to women with survey-reported MST only. Among post-9/11 Veterans, women are more likely to be treated for depression and anxiety than men, ³⁷ and women are less likely to be diagnosed with PTSD than men. ²⁷ These associations could differ for Veterans with MST. For example, VHA providers may instead be prone to interpret mental health symptoms as PTSD or depression than anxiety, or women who endorse MST in EHR screening may exhibit greater symptoms meeting criteria for PTSD or depression than anxiety, both hypotheses to test in future investigations.

As observed in earlier studies, ^{11 –13,19} rates of survey-reported MST far exceeded those of EHR-documented MST. A systemic explanation for discordant MST records is that Veterans who initially deny a history of MST may not be screened again. ³⁸ However, in our cohort, all Veterans who initially declined EHR MST screening and were later reassessed eventually completed the MST screen. These data align with a VHA report that over 98% of Veterans who received recent care had been screened for MST. ³⁹ Timing of the VHA MST screening is also critical. The EHR screening tool is most often completed at the first primary care visit. At that time, Veterans with a history of MST may expect negative reactions to disclosure, believe that disclosing MST does not pertain to health, and experience feelings of institutional betrayal. ^{9 –11} Other barriers to MST disclosure may include the characteristics of a clinician, including their gender, communication style, and perceived compassion. ²⁰ In this study, we found that the median timeframe for Veterans to disclose MST was 6 months for women and 1.5 years for men. In another random sample of 20,000 men and women Veterans, 13% of women and 39% of men disclosed MST more than 12 months after establishing VHA care. ¹⁹ After rapport and trust have been cultivated with their clinician, patients likely feel differently about disclosing MST, or are at least inclined to complete the MST screening. ²⁰ Ultimately, building trust to complete MST screening can take time.

To address this issue of disclosure, or gap between assessment results, the high discordance of MST reporting in this study and other similar investigations should encourage VHA to consider other, novel MST screening methods. Examples of alternate screening methods include a self-report questionnaire, screening that is completed remotely with a clinician interview or matching the gender of the clinician with the patient, all approaches that could improve patient comfort with disclosure. Education about trauma-informed care is also necessary. In another investigation, Veterans with a positive EHR MST screen also endorsed high satisfaction with MST-related communication, associations that were particularly salient for women. ²⁰ Moreover, continuing comprehensive MST screening over time is essential, perhaps along with other required mental health screening. Recurrent screening for MST could be particularly beneficial for each gender, but perhaps for different reasons. Previous evidence and these results herein both suggest that repeated MST screening improves women’s rates of MST disclosure, ¹³ thereby allowing the provision of mental health services or other preventive care. Finally, Veterans with MST also show a higher prevalence of sexual trauma outside of military service, ^2,40 and VHA should continue improving education about, and use of, trauma-informed care for men and women alike. ⁴¹

Strengths of this investigation include the use of a nationwide sample of men and women Veterans, multiple sources of MST data, and use of the VHA MST definition in the WVCS survey, to ensure consistency between the two data sources. There are also some limitations. First, the findings cannot be generalized to Veterans who did not receive VHA care or to those from other service eras. Although the size of this male group was sufficient for some analyses, the number of men who reported MST was very small; thus, interpretation of findings specific to men should be re-examined with a larger sample. Furthermore, individuals from some demographic subgroups such as sexual minority Veterans show a higher prevalence of MST, ⁴² and associations with health care utilization and psychiatric diagnoses should also be pursued with these vulnerable groups. As only 57% of Veterans receive VHA care, ⁴³ screening for MST in non-VHA settings is also encouraged to provide optimal care for this group. A second limitation is the cross-sectional nature of analyses, meaning that causality between MST, health care utilization, and mental health diagnoses cannot be definitively established. Further, we could not determine the service in which mental health diagnoses were initially determined and any of the diagnoses could have been misclassified. Third, given the retrospective nature of MST reports, there is the potential for response bias. As the survey data were collected 2016–2020 for Veterans discharged as early as 2001, the clinical screening may have occurred up to 19 years before survey completion. Fourth, although EHR MST screening predated the diagnoses, it is unknown if any Veterans re-entered the military after initially establishing VHA care and responding to the EHR MST screening, which could have led to additional trauma not documented in the EHR. Fifth, we did not examine other types of sexual trauma among this group (e.g., intimate partner violence, sexual abuse outside of service), ⁴⁴ data that could be complementary and helpful for understanding the influence of sexual trauma on Veterans’ health and health care. It may also be beneficial to expand VHA MST screening to include several other lifetime trauma types, to provide more comprehensive assessment of trauma to improve related care. ³⁴

For over 20 years, the VHA has devoted significant resources to coordinating and managing the widespread health consequences of MST. This initiative has allowed thousands of trauma survivors to receive better, trauma-centered health care, enabling them to recover from their experiences and live full lives after military service. With the acknowledgment that rates of MST detection in VHA have increased, despite the limitations of the current screening mandate, there was significant discordance between VHA records of MST and survey reports of that trauma among men and women. This discordance significantly limits the ability of VHA providers to offer appropriate health care and other services to palliate the psychological, social, and biological repercussions of sexual trauma. Ultimately, these data reinforce the need for improving the coordinated, two-pronged approach: The U.S. military must better attend to MST prevention and to cultural influences that promote sexual violence, and the VHA must mutually improve Veterans’ awareness of the health benefits of MST disclosure and the utility of MST detection, thereby fulfilling their duty to provide equitable and expedient patient-centered care.