Caring for Service Members Who Have Been Sexually Assaulted: The Military Health System

ABSTRACT Introduction Reports of sexual assault (SA) in the U.S. Military have increased in recent years. Given the deleterious effects of military SA, there remains a need for large-scale studies to assess SA-related health care utilization among active duty service members (ADSMs). The present study, therefore, utilized Military Health System (MHS) data to determine the prevalence of SA-related care, sociodemographic characteristics of ADSMs receiving said care, and the type of provider seen during the initial SA-related health encounter. Materials and Methods Utilizing the MHS Data Repository and Defense Enrollment Eligibility Reporting System, all ADSMs from the Air Force, Army, Navy, and Marine Corps during fiscal years (FY) 2016-2018 were identified. Those with an International Classification of Diseases diagnostic code related to SA during the study period were isolated. Descriptive statistics and multivariable logistic regression analyses were conducted. The study was exempt from human subjects review. Results A total of 1,728,433 ADSMs during FY 2016-2018 were identified, of whom 4,113 (0.24%) had an SA-related health encounter. Rates of SA-related health care encounters decreased each FY. Women (odds ratio [OR] = 12.02, P < .0001), those in the Army (reference group), and enlisted personnel (OR = 2.65, P < .0001) were most likely to receive SA-related health care, whereas ADSMs aged 18-25 years had lower odds (OR = 0.70, P < .0001). In addition, higher odds of SA-related care were observed among those identifying as American Indian/Alaskan Native (OR = 1.37, P = .02) and “Other” race (e.g., multiracial) (OR = 4.60, P < .0001). Initial SA-related health encounters were most likely to occur with behavioral health providers (41.4%). Conclusions The current study is the first large-scale examination of health care usage by ADSMs in the MHS who have experienced SA. Results indicated that rates of SA-related care decreased throughout the study period, despite the increasing rates of SA documented by the DoD. Inconsistent with previous research and DoD reports indicating that younger ADSMs are at the highest risk for SA, our study observed lower rates of SA-related care among those aged 18-25 years; additional research is warranted to determine if there are barriers preventing younger ADSMs from seeking SA-related health care. Behavioral health providers were most frequently seen for the initial SA-related encounter, suggesting that they may be in a unique position to provide care and/or relevant referrals to ADSMs who have experienced SA. The present study provides key insights about the prevalence of SA-related care within the MHS, not yet reported in previous literature, which could help inform MHS screening practices. The strengths of the study are the inclusion of the entire active duty population without the need for research recruitment given the utilization of de-identified TRICARE claims data. The study is limited by its use of health care claims data, general SA International Classification of Diseases codes as a proxy indicator for military SA, and lack of data on ethnicity. Future research utilizing MHS data should examine mental health outcomes following the documentation of SA and disruptions in SA-related care due to SARS-CoV-2.


BACKGROUND
Information from both DoD formal reports and anonymous surveys indicate that sexual assault (SA) in the U.S. Military has been rising in recent years, with this trend continuing throughout fiscal years (FY) 2016-2018. [1][2][3][4] The experience of military sexual trauma (MST) has been linked to adverse outcomes-including suicidality, poorer physical health, and increased risk-taking behavior. [5][6][7][8][9][10][11] The prevalence of MST is lower for men (1%-3%) than for women (25%-33%); however the overall number of service members who have experienced MST is fairly comparable, given that men have comprised approximately 84% of the total military force since 2004. 12,13 Although research consistently indicates that younger age and enlisted rank are associated with greater risk for MST, other factors such as race are not as clear. 3,[14][15][16][17] Additionally, the DoD annual report on new SA cases formally reported in the military demonstrated that the Army (the largest service branch) had the highest rate of MST, 14 whereas studies with veterans have indicated that those from the Navy and Marine Corps were more likely to experience MST. 16,18 The Workplace Gender Relations Study is an anonymous survey distributed biannually throughout the DoD. From 2016 to 2018, there was a statistically significant increase in the number of service members who had experienced SA in the previous year (14,900 in FY 2016to 20,500 in FY 2018. 4 Based on the information from these data and in comparison to DoD reports, the researchers estimated that only about a third of service members who experienced SA made a formal report. When making a formal report of SA, service members can elect to have their report be either Restricted or Unrestricted. 3 Restricted reports remain confidential, whereas Unrestricted reports allow official investigations to be launched and a greater amount of information to be documented, tracked, and shared. Therefore, although the DoD publishes data of new SA reports made each year, there is no coordinated health care information collected for these reports. Furthermore, a majority of the MST research thus far has been conducted with veteran populations, or with health care services outside of the Military Health System (MHS), such as the Veteran Health Administration. [19][20][21] As such, little is known about the actual health service utilization of active duty service members (ADSMs) who have experienced SA.
Given the adverse effects of MST and the threat it poses to military safety and readiness, there remains a need for large-scale studies to assess health care utilization among ADSMs who have experienced SA. Additionally, this information could be valuable to inform MHS screening practices. The present investigation is the first population-level study to examine the prevalence of SA-related health care delivered to ADSMs in the MHS and the type of provider seen during the initial SA-related health encounter.

METHOD
Utilizing the Defense Enrollment Eligibility Reporting System, all ADSMs aged 18 and older from the Air Force, Army, Navy, and Marine Corps during FY 2016-2018 were identified. Given the study focus on care provided in the MHS, Active National Guard and Reserves members were excluded due to their inconsistent access to the MHS.
The Military Health System Data Repository (MDR)-the centralized data repository that captures Defense Health Agency health care data worldwide-was used to identify information on health care usage during the study period. The MDR is a single source for all health care encounter data for MHS beneficiaries receiving care at military treatment facilities, also known as direct care, and at civilian fee-for-service facilities accessible through TRICARE benefits, also known as private sector care. TRICARE is the DoD insurance product providing coverage to 9.6 million beneficiaries, including ADSMs, retired personnel, and their families. 22 The MDR has proven to be an abundant source for research on care utilization and behavioral health for ADSMs and all MHS beneficiaries. 23,24 The MHS and TRICARE do not include care delivered by the Veterans Affairs system or services paid out of pocket by ADSMs.
Utilizing the International Classification of Diseases, 10th revision codes related to SA, we identified all SA-related claims by ADSMs in both the direct and private sector care settings during the study period. If an ADSM had multiple claims related to SA, only the first record during the study period was retained.
Demographic and clinical variables examined included: Age, gender, race, service branch (Army, Navy, Air Force, Marine Corps), and military rank (Enlisted, Officer, and Other for cadets or unknown and unspecified rank). Age was stratified into three groups: Young Adult (18-25 years), Adult (26-40 years), and Older Adult (41 years and above). Race is reported in the MDR as White, Black, Asian/Pacific Islander, American Indian/Alaskan Native, and Other; information on ethnicity was not available for this study.
Provider types were categorized using taxonomy codes established by the Health Insurance Portability and Accountability Act and grouped into four categories: Physicians (i.e., doctor of medicine and doctor of osteopathic medicine); behavioral health and social service providers (e.g., psychologists, social workers, and counselors); physician assistants (PAs), nurses, and nurse practitioners (NPs); and a remaining "other" group for those that did not fit the previous categories (e.g., occupational therapist, clinical medical laboratory, and rehabilitation hospital). Given that Health Insurance Portability and Accountability Act taxonomy codes group a variety of medical specializations (e.g., brain injury medicine, epilepsy, addiction medicine, and pain medicine) into a singular "Psychiatry and Neurology" code, for the purpose of the current analyses, these providers were included within the physician category.
Descriptive statistics were performed on the ADSMs' demographics and the type of provider seen during the initial SA-related health encounter. To maintain patient anonymity, data with cell counts of less than 11 were suppressed. SArelated encounters per 1,000 ADSMs were calculated from 2016 to 2018 for trend analysis. Multivariable logistic regressions were performed to assess the associated risks of SArelated care among ADSMs. Those with incomplete data were excluded from the logistic regression analyses, resulting in a total of 1,683,727 ADSMs (2.6% loss). All analyses were performed using SAS 9.4. Significance was defined as P-value < .05. Restrictions apply to the availability of these data, which were used under federal Data User Agreements for the current study, and thus are not publicly available.
This study was deemed exempt from human subjects review by the Institutional Review Board of the USU. Sociodemographic characteristics of ADSMs with SArelated health care encounters can be found in Table I. Of the 4,113 with an SA-related health encounter, the majority were young adults aged 18-25 years (72.3%), female (71.3%), White (59.0%), in an enlisted rank (93.8%), in the U.S. Army (54.2%), and received their care in the direct care setting (88.9%) versus private sector care.

RESULTS
Table II details the results from the multivariable logistic regression for SA-related care in ADSMs. Women were 12 times more likely to have received care related to SA than men (odds ratio [OR] = 12.02, 95% CI, 11.21-12.89, P < .0001). In addition, higher odds of SA-related care were observed among those identifying as American Indian/Alaskan Native (OR = 1.37, 95% CI, 1.04-1.80, P = .02) and "Other" race (e.g., multiracial) (OR = 4.60, 95% CI, 4.10-5.17, P < .0001). There was no significant difference in the odds of SArelated encounters for Black service members (OR = 0.98, 95% CI, 0.90-1.05, P = .53) compared to the reference group of White service members. Lower odds of SA-related care were observed in ADSMs aged 18-25 years (OR = 0.70, 95% CI, 0.65-0.75, P < .0001). Across the branches, service members in the Army (reference group) had the highest odds of SA-related care. Figure 1 details the distribution of provider types seen by ADSMs for SA-related care. The majority of the SA-related health encounters were with behavioral health and social service providers (41.4%). SA-related health encounters with medical professionals were as follows: 34.2% by physicians; 15.6% by PAs, NPs, and nurses; and the remaining 8.8% from the "other" category of providers. When examined by care setting, similar distributions were observed in ADSMs receiving direct care; however, those in the private sector (n = 458) were predominantly seen by providers in the "other" category (71.8%), followed by physicians (22.3%), PAs/NPs/nurses (2.0%), and behavioral health specialists (3.9%).

DISCUSSION
In contrast to previously published DoD sources that show an increase in both anonymous rates 4 and formal SA reports in recent years, 1-3 the MHS recorded SA-related health encounters decreased each FY from 2016 to 2018. As aforementioned, the Workplace Gender Relations Study saw that anonymous rates of SA increased from roughly 15,000 to  4 whereas only about a third of those who experienced SA made a formal report, increasing from 4.1 to 5.1 formal reports per 1,000 service members. 1-3 Considering our study data revealed SA-related care was delivered at the rate of 0.95-1.12 per 1,000 service members, these differences could potentially be indicative of a gap in care. Despite the difference in prevalence, the current results reflect similar findings in previously reported rates of SA, such that women and enlisted service members were the most likely to receive SA-related care. [14][15][16]20,[25][26][27] In addition, we found that ADSMs in the Army were most likely to use health care services for an SA, similar to previous research. 14,28 The Army is the largest military branch and soldiers are often overrepresented in studies compared to ADSMs from other branches; however, the present study utilized OR comparisons, indicating that the prevalence is not due to branch size alone. Despite previous research and DoD reports indicating those who are younger are at the highest risk for SA, 14 our study of MHS data indicated lower rates of SA-related health care among those aged 18-25 years compared to ADSMs aged 26-40 years, who had the highest odds of SA-related health encounters. Additional research is warranted to determine if there are barriers preventing younger service members from seeking SA-related health care, such as greater stigma or fear of retribution. 29 Regarding race, those who identified as American Indian/Alaskan Native or as part of an "Other" category of race (e.g., multiracial) were the most likely to have an SA-related health encounter, compared to those identifying as White. As aforementioned, the research on whether racial or ethnic minorities are at greater risk for MST is inconsistent; for instance, some studies have shown that those identifying as racial or ethnic minorities are at greater risk, whereas others show no significant difference at all. 16,26,27,30 However, a review of DoD Gender Issues Surveys completed by 22,372 female service members found that those of low sociocultural power status (i.e., younger age, lower education, minority racial group membership, and non-married) had a higher likelihood of experiencing SA and harassment. 5,31 Consistent with the findings from the current investigation, this suggests that there may be a compounding interaction between several sociodemographic factors (i.e., intersectionality) placing minorities at greater risk, which should be explored. 5,16,30,31 Behavioral health and social service providers were the category of providers most likely to be engaged in the initial SA-related health encounter. This is in contrast with some civilian studies, showing that the majority of SA reports are made to medical providers (e.g., ranging from 18% to 72%), whereas reporting to behavioral health and social service providers occurs at a lower rate (e.g., ranging from 19% to 27%). [32][33][34][35] The current findings may reflect greater accessibility of behavioral health services among some TRICARE beneficiaries, although additional studies are warranted.
To our knowledge, the current study is the first large-scale assessment of health care usage within the MHS setting by ADSMs who have been sexually assaulted. The strengths of the study are the inclusion of the entire active duty population and all reported health service utilization captured within the MDR. Furthermore, given the utilization of de-identified TRICARE claims data, the full population of ADSMs was included without the need for research recruitment or the reliance on self-reported SA data. The use of health care claims data and general SA International Classification of Diseases codes as a proxy for MST is a limitation of this study as the data only capture MHS encounters where SA was documented, and likewise may represent SAs that occurred before one's military service. Additionally, the MHS has begun transitioning from the MDR to a new electronic health record, MHS Genesis. Some of the initial sites began this transition in 2017; therefore, study data from FY 2017 to 2018 do not include information from sites enrolled in Genesis during that time. Furthermore, the MHS only includes information on race, but not ethnicity; therefore, it was not possible to isolate data for those identifying as Hispanic. This is a notable limitation of the study as Hispanic service members may have different experiences and outcomes. 5,31 Given that psychiatrists were included within the "physician" category for reasons cited above, the current study did not capture individuals who sought treatment from physicians within mental health specialty clinics, thereby potentially underestimating the proportion of mental health providers seen for SA-related care.
Future research utilizing MHS data should examine mental health outcomes following the documentation of SA. Similar to analyses in this current study, sociodemographic differences in mental health outcomes should be explored. Additionally, further delineation of the types of health care services sought could potentially reveal whether members from different services or sociodemographic categories have a greater likelihood of engaging in medical care versus mental health care. Research regarding trauma-informed care within the MHS should also be explored. Trauma-informed care includes universal precautions to make health care systems safer and more welcoming to all, as well as trauma-specific practices to promote treatment engagement, improve outcomes, and reduce the risk of re-traumatization for those who have experienced trauma. 36,37 Trauma-informed care is not limited to behavioral health environments but rather is applicable to all spaces (e.g., primary care, dentistry, and family health) where someone who has experienced trauma may ostensibly receive services. Comparing outcomes between patients receiving trauma-informed care and treatment as usual may be an important future direction to elucidate possible benefits of trauma-informed practices as well as the need for greater training and implementation within the MHS. 36,38 Service outreach programs and procedures should also be examined for differences that might encourage and support greater health care seeking behavior.

CONCLUSION
Although eliminating SA in the military remains a focus for military leaders, service members, and veterans, 1-3,19-21 to date there had been no large-scale study of health care utilization by ADSMs who have experienced SA. In addition, the majority of the currently available MST research has been conducted by the Veteran Health Administration using samples of veterans. The present study provides key insights about the prevalence of SA-related care within the MHS, not yet reported in previous literature. Future research should utilize the MDR to examine mental health outcomes following documentation of SA and MHS disruptions in SA-related care due to SARS-CoV-2.