Use of Health Services Among Post-9/11 Veterans With Mental Health Conditions Within 90 Days of Separation From the Military

There are more than 2.6 million post-9/11 veterans, and the population is projected to grow to 3.5 million by 2019 (1). The majority of veterans make a successful transition back to their communities (2). However, a significant minority have difficulty coping with the adjustments required for successful reintegration into civilian life (3). Mental health problems, which veterans frequently report, are one of several challenges to reintegration (3). Of post-9/11 veterans enrolled in health care at the U.S. Department of Veterans Affairs (VA), 57% received a provisional psychiatric diagnosis (4), a proportion significantly higher than that seen in non-VA and civilian samples (5). Post-9/11 veterans report that receiving care for psychological challenges (e.g., posttraumatic stress disorder [PTSD], depression, and anxiety) is a pressing need (6).

The VA serves 9 million veterans annually, but some veterans experience access-related challenges to VA care. Both VA and private health care networks face provider shortages (7, 8). There are also barriers to veterans’ receipt of health care (9), including lack of transportation, family or work obligations, and limited child care options (10). Help-seeking stigma has likely contributed to veterans underusing mental health care (10). Some veterans also report being confused about their eligibility as a reason for not using VA programs (6).

Providing high-quality health care to veterans with mental health issues is complex. A number of factors must be addressed (3), including predicting staffing needs, identifying and lowering barriers to care, and decreasing help-seeking stigma. One of the most important factors is anticipation of future health care use patterns. Thus, more information is needed about the types of health care that veterans actually use (8), the timing of use (11), and the characteristics of underserved veterans (12).

This study examined the use and predictors of use of health services and programs within the first 90 days of military separation by post-9/11 veterans who screened positive for a probable mental health problem. Health services were defined as all services dealing with the diagnosis and treatment of disease or the promotion, maintenance, and restoration of health (13). Health programs were defined as any activity designed to meet a veteran’s specific health needs (13). (Further explanation of services and programs is provided in an online supplement to this article.) Focusing on health care utilization in the early stages of transition to civilian status is especially important among veterans with probable mental health problems because delayed treatment may exacerbate symptoms and lead to poorer outcomes.

Methods

Participants

Participants were identified by using the VA/Department of Defense Identity Repository (VADIR). The total population of post-9/11 veterans who separated from active duty service in the prior 90 days (N=48,965) was identified and invited to participate in the study. Wave 1 data from the the Veterans Metrics Initiative were collected between September 13 and November 20, 2016. Complete data were provided by 9,566 veterans. Thirty-four percent of the sample (N=3,295) were discharged from the active component (Army, Navy, Air Force, Marine Corps) and screened positive for one or more probable mental health problems.

Measures

Veterans were asked, “How often have you used the following services (e.g., VA hospitals or clinics, VA Vet Center, non-VA hospital or clinic, Veterans Choice program, alternative medicine such as acupuncture or chiropractic care, and counseling services) over the last 3 months?” Response options were never, less than once a month, once or twice per month, three or four times per month, two to three times per week, and four or more times per week. Next, to assess the use of programs that lower barriers to care, veterans were asked, “Aside from VA options and your health insurance, what program(s) have you used in order to increase your access to care, such as Vets Prevail, Warrior Care Network, or Disabled American Veterans Transportation Network?” Veterans were also asked, “What program(s) have you used for a brain injury or PTSD, such as the Soldiers Project or the Defense and Veterans Brain Injury Center?” Last, respondents were asked to identify any programs they had used to increase their physical activity or assist with weight management.

Several covariates related to mental health were included in the analyses. Combat exposure was measured via a nine-item scale that asks how often the veteran encountered a variety of combat-related events, such as encountering booby traps or roadside bombs. Response options included 0 (never), 1 (once or twice), 2 (several times), and 3 (many times). Because of a nonnormal distribution, response options were recoded into a dichotomized variable (1, at least once; and 0, never). Second, respondents were asked, “Do you have an ongoing physical health condition, illness, or disability (for example, high blood pressure, chronic pain)?” and “Do you have an ongoing mental/emotional health condition, illness, or disability (for example, depression, anxiety)?”

Participants completed the five-item version of the Primary Care PTSD Screen, with a cutoff score of 3 or more indicating probable PTSD (14). In this study, the screen demonstrated good internal consistency reliability (α=0.82). To assess probable depression or anxiety, veterans were asked to complete the Patient Health Questionnaire–4 (PHQ-4), with a cutoff score of 3 or more indicating probable depression and anxiety (15). The PHQ-4 demonstrated good internal consistency reliability for anxiety (α=0.82) and depression (α=0.77) in this study. The Alcohol Use Disorders Identification Test for Consumption (AUDIT-C) (16) was used to identify veterans with drinking problems, with a cutoff score of 3 for women and 4 for men. In this study, the AUDIT-C demonstrated good internal consistency reliability (α=0.82). (The validity of these measures is reported in the online supplement.)

Procedures

Institutional review board approval was obtained from ICF International. Veterans who recently separated from active duty service were identified from VADIR. Each veteran was sent notification of eligibility to participate in the study by using a reminder and incentive strategy (11). Informed consent was obtained, and each veteran received a $20 electronic gift card from Amazon.com upon survey completion.

Data Analytic Approach

In survey studies, there is always a question regarding the degree to which sample respondents are representative of the full population. A commonly applied correction technique is weighting (see online supplement for more details). Weighted proportion estimates were computed by using STATA svy: proportion (17). (Unweighted proportion estimates are provided in the online supplement.)

Results

Demographic Characteristics

Table 1 provides the demographic characteristics of the sample. In brief, the sample was predominantly male and white and from the enlisted ranks. The average age was 34 years. Detailed information about the sample has been published previously (11).

Self-Reported Health Status and Use of Health Care

Table 2 describes the health status and use of health care among the 3,295 (N=16,204 weighted) veterans who screened positive for a mental health problem. Seventy-three percent screened positive for a probable anxiety disorder, 51% for probable depression, 69% for probable PTSD, and 57% for probable alcohol misuse. A significant majority (77%) of the veterans reported having an ongoing general medical condition, and 73% self-reported having an ongoing mental health condition.

Twenty percent of the sample screened positive for a single probable psychological problem, 34% screened positive for two probable psychological problems, 30% screened positive for three, and 17% screened positive for all four.

Nearly half (47%) of those who screened positive for a probable mental health problem reported using a VA hospital or clinic, and 19% used a non-VA hospital or clinic. Nearly one-third received counseling for a mental health, relationship, or substance misuse problem, 11% used alternative medicine like acupuncture or chiropractic services, and 8% used various other programs or services, while 31% reported not using any health programs or services. More than one-third of veterans reported using one health service or program, 23% used two, and 10% used three or more.

Factors Associated With Use of Health Care Services and Programs

Results from multivariate logistic regression examining predictors of health service use among veterans who screened positive for a mental health condition are provided in Table 3.

Use of VA hospitals and health clinics.

Male and female veterans did not differ in use of VA hospitals or health clinics. Non-Hispanic blacks and Asian/Native Hawaiian and other Pacific Islanders were more likely than non-Hispanic whites to use the VA (odds ratios [ORs]=1.32 and 2.05, respectively). Veterans who were exposed to combat were more likely than noncombat veterans to use the VA (OR=1.37). Veterans with a medical discharge were nearly three times more likely than those without a medical discharge to use the VA, whereas those who reported an ongoing general medical problem were twice as likely as those who did not report a medical problem to use the VA. Those who reported an ongoing mental health problem were more likely to use the VA than those who did not report a mental health problem (OR=1.59). Finally, full-time employment significantly predicted higher use of non-VA health care and lower use of health care at the VA.

Use of non-VA hospitals and health clinics.

Compared with female veterans, male veterans were significantly less likely to use non-VA health care (Table 3). Asian/Native Hawaiian and other Pacific Islanders were significantly less likely than non-Hispanic whites to use non-VA hospitals and clinics. Compared with veterans from the most junior enlisted ranks (E1–E4), veterans from all ranks except some officers (ranks O1–O3) were more likely to use non-VA health care (OR=1.35–3.24). Veterans with a medical discharge were significantly less likely than those without a medical discharge to use non-VA health care.

Use of counseling.

As shown in Table 3, male and female veterans did not differ with respect to the use of any type of counseling services. Hispanic and Asian veterans were more likely than white veterans to use counseling (OR=129 and 1.70, respectively). Some members of the senior enlisted ranks (O1–O3) and junior officer ranks (E7–E9) were more likely than those from the junior enlisted ranks (E1–E4) to use counseling services (OR=1.44 and 1.62, respectively). Combat-exposed veterans were more likely than noncombat veterans to use counseling services (OR=1.30). Those who self-reported having an ongoing mental health condition were nearly six times more likely to report using counseling services. Veterans screening positive for alcohol misuse were significantly less likely to attend counseling relative to those who did not screen positive.

Use of alternative medicine.

Compared with veterans from the most junior enlisted ranks, those from the highest enlisted ranks, warrant officers, the most junior officers, and members of the highest officer ranks were more likely to use complementary and alternative medicine (CAM) (OR=1.52–2.30). Veterans with a medical discharge and those reporting an ongoing general medical problem were both more than twice as likely to use CAM than those without a medical discharge or who did not report an ongoing medical problem. Veterans who screened positive for alcohol misuse were significantly less likely than those who screened negative to use CAM.

Use of programs that enhance access to health care.

As shown in Table 4, veterans’ use of programs that enhance access to health care did not differ as a function of gender or race-ethnicity. Compared with veterans from the most junior enlisted ranks, those from the highest enlisted ranks were nearly three times more likely to use access programs, and veterans from the highest officer ranks were nearly five times more likely to use these programs. Veterans who saw combat were twice as likely as noncombat veterans to use programs that assist with access to care.

Use of PTSD/brain injury–specific programs.

As shown in Table 4, compared with veterans from the most junior enlisted ranks, those from more senior enlisted ranks were more than twice as likely to use services for PTSD or brain injury, and those from the highest officer ranks were nearly five times more likely to use such services. Veterans with combat exposure were almost five time more likely than those without combat exposure to use PTSD/brain injury services. Veterans who reported an ongoing mental health or general medical problem were nearly six and four times more likely, respectively, to use these programs, compared with veterans who did not report such problems. Veterans who screened positive for PTSD were five-and-a-half times more likely than those who screened negative to use PTSD/brain injury programs. However, only 5% of all veterans with a probable PTSD diagnosis reported using a program that specifically targets PTSD or brain injury (data not shown in table).

Use of physical activity or weight management programs.

There were no differences in the use of physical activity or weight management programs as a function of gender or race-ethnicity (Table 4). Compared with veterans from the most junior enlisted ranks (E1–E4), veterans in the highest enlisted ranks were approximately twice as likely to use these programs. Combat-exposed veterans were more than two times as likely to use these programs.

Discussion

Understanding patterns of health care utilization among new post-9/11 veterans with mental conditions who are transitioning to civilian life can inform efforts to enhance access to care, reduce barriers, and improve treatment options for those who are underserved (4). Similar to findings from other studies (18, 19), approximately one-third of post-9/11 veterans screened positive for a mental health condition. Approximately one-quarter of veterans who screened positive for a mental health problem did not self-identify as having any mental health problems. Thus, some veterans may underestimate their own likelihood of having a problem or are not willing to acknowledge problems they are experiencing (20). However, it is important to note that screening measures typically overestimate prevalence and that veterans may not always interpret symptoms as indicative of a mental disorder. Veterans with symptoms may continue to function well and see themselves as psychosocially sound.

Understanding where veterans with mental problems receive care has become increasingly important, given long wait times at VA facilities and the wider options for receiving non-VA health care under the Veterans Choice Act/Veterans Community Care program. Nearly half of veterans in this sample reported receiving care in VA hospitals or clinics, whereas 19% used non-VA facilities. The VA is concerned about the extent to which the VA treats patients with both mental and general medical problems. In this study, the vast majority of veterans who screened positive for a mental health condition also reported at least one ongoing general medical problem (6, 7). Many post-9/11 veterans are likely to need treatment for both mental health and general medical problems, and as a result there may be a significant rise in veteran health care costs over the next 15 years (21). Also, demand for services will likely increase, given that veterans’ health problems tend to emerge over time (22) and veterans with comorbid conditions may require more team-based interdisciplinary care (23).

With respect to predictors of service and program use, we reported a consistent finding. Among veterans who screened positive for a mental health problem, those from the junior enlisted ranks (E1–E4) were significantly less likely than veterans from higher ranks to use a number of health care services, suggesting that the more junior group is underserved (24). This finding is concerning, given that veterans from the junior enlisted ranks encounter more challenges as they transition to civilian life, including lower earnings, higher risk of homelessness, and higher unemployment (25). Veterans from the junior enlisted ranks should be targeted for enhanced transition support (e.g., development of transition plans, access to mentors) to reduce the likelihood of facing these challenges.

The VA is further concerned about the degree to which veterans from racial-ethnic minority groups access care. Black and Asian veterans were significantly more likely than their white peers to use VA services, although there were few differences in their use of non-VA services and programs. The higher use of VA services by veterans from racial-ethnic minority groups may be the result of efforts undertaken by the VA to increase access to care. The VA established the Center for Minority Veterans, which works to ensure that all veterans receive equal services and helps steer veterans from racial-ethnic minority groups to resources and programs offered by VA. This suggests that the VA is making attempts to be sensitive to cultural issues (26).

Veterans from racial-ethnic minority groups may be more likely to use VA services because of other factors as well, such as higher rates of unemployment and a lower likelihood of having private insurance among veterans from some groups. Non-Hispanic white veterans were significantly less likely than nonwhite veterans to use VA services or counseling, replicating some prior research (12). This is a concern, given that non-Hispanic white females screen positive for mental health challenges more frequently than their nonwhite counterparts (27).

Use of health care services and programs was similar among male and female veterans who screened positive for mental health problems. This is consistent with recent research indicating that the newest generation of post-9/11 female veterans seek out VA services more than previous cohorts, likely because of enhanced programmatic offerings to women, greater outreach efforts, and increasing satisfaction with services (20). Men and women also did not differ in their use of CAM or in any of the three program types examined, although female veterans were more likely to use non-VA services. Male veterans may need to seriously consider using more non-VA services, particularly because the wait times may be shorter than at the VA.

Among veterans who screened positive for a mental health condition, those exposed to combat, those with a medical discharge, and those with a chronic general medical or mental health problem were all more likely to report using VA health care services. Given that prior health care utilization is the best predictor of future utilization (28), it is likely that at least some of these veterans had received treatment while on active duty. Therefore, increasing access to health care (and decreasing barriers) during active duty as well as providing consistent and accurate information about VA services prior to separation would likely smooth the military-to-civilian transition. On the other hand, veterans who screened positive for alcohol misuse demonstrated similar or lower health care use (i.e., counseling, alternative medicine) than veterans without mental health problems. Alcohol and substance abuse remains a difficult problem among veterans (3). Individuals known to have substance abuse or alcohol problems may require patient-centered multidisciplinary care. It was also concerning that only 5% of veterans who screened positive for PTSD reported using a PTSD program. Targeted approaches to increasing the health care engagement of veterans with PTSD are needed.

This study had several limitations. First, it reported health care use at one point in time. Usage likely changes significantly over time. Second, limitations are inherent in the use of mental health screeners to estimate the prevalence of mental health problems within a population. The screeners used in this study were extremely brief and were not validated by subsequent diagnoses. Third, screening instruments typically overestimate the number of individuals with mental health problems (i.e., high false positives). Fourth, the questions asked of veterans in this study were not standardized, nor were they validated against actual VA clinical records. Therefore, the results may be skewed. To the extent possible, future studies should compare veteran reports of health care use against objective indices of use. Fifth, this study examined the health care utilization of post-9/11 veterans who were separated only recently from active duty service. Therefore, these findings should not be generalized to other veteran groups. (Additional limitations are described in the online supplement.)

Conclusions

Post-9/11 veterans with a probable mental health problem primarily used the VA for their health care. Approximately 19% sought services from non-VA providers, possibly reflecting the wider health care choices veterans now have. It is not clear, however, whether this proportion represents an increase over prior use of civilian and community health care. There were more similarities than differences between female and male veterans and among veterans from various racial-ethnic backgrounds, perhaps reflecting enhanced efforts by the VA to provide outreach and services targeted to women and veterans from racial-ethnic minority groups. Veterans from the most junior enlisted ranks were significantly less likely to use health care services or programs; a concerning finding given that they are most at risk of difficult transitions from military to civilian life. It may be true that veterans from the junior enlisted ranks who screen positive for probable mental health problems function well in the early civilian transition. Later, however, they may require more encouragement to access both VA and non-VA health care services and programs.