Despite the high prevalence of psychiatric symptoms among male and female veterans returning from Afghanistan (Operation Enduring Freedom [OEF]) and Iraq (Operation Iraqi Freedom [OIF] and Operation New Dawn [OND]) (1–4), many affected veterans do not seek mental health services (5–7). OEF/OIF veterans with mental disorders favor the use of primary care services (8). According to Hoge and colleagues (1), OEF/OIF veterans with a mental health condition were twice as likely as their counterparts without mental conditions to report barriers to seeking mental health care.
Although several studies have examined rates of utilization of mental health and primary care services among OEF/OIF/OND veterans (7,8), to date no study is known to have investigated correlates of length of time to engagement in primary care and mental health outpatient treatment, including minimally adequate mental health treatment. Studies of several mental disorders have found eight sessions to constitute minimally adequate mental health treatment (6,9). Receipt of minimally adequate treatment by OEF/OIF/OND veterans is of particular concern because these veterans have been shown to be highly ambivalent about mental health outpatient treatment. They may attend only one or two mental health outpatient sessions and then drop out of treatment (7,10). More specifically, only 9.5% of OEF/OIF veterans with PTSD received minimally adequate mental health care in the year after the diagnosis (7).
Decreasing the length of time to treatment may reduce chronicity of symptoms and unremitting functional impairment among veterans who experience mental symptoms in response to war zone exposure (11,12). We sought to further delineate time to health care utilization among OEF/OIF/OND veterans with psychiatric diagnoses who sought health care from the U.S. Department of Veterans Affairs (VA). We used a retrospective design to describe time to first primary care visit, first mental health outpatient visit, and first mental health outpatient visit in a series of eight or more visits completed within 12 months (minimally adequate mental health care).
Furthermore, we examined demographic characteristics, military service, and psychiatric diagnoses that predicted each of the three outcomes, with particular attention to gender differences because of prior studies highlighting differences in utilization by gender (13). We hypothesized that veterans would seek primary care before seeking mental health care, that there would be a delay between the first mental health outpatient visit and initiation of minimally adequate mental health care (eight or more visits within a 12-month period), and that time to utilization would be shorter among female veterans.
Data source and extraction
We conducted a retrospective cohort analysis by using existing medical records from the VA OEF/OIF/OND Roster, a national database of veterans who had separated from OEF/OIF/OND military service and who enrolled in VA health care. We linked the OEF/OIF/OND Roster database (current as of March 2011), which contains veterans’ demographic and military service information, to the VA National Patient Care Database, which provides VA visit dates and associated ICD-9-CM diagnostic codes.
All analyses were restricted to OEF/OIF/OND veterans who had received at least one of six common psychiatric diagnoses, who were new users of VA care, and who had utilized either primary or mental health outpatient care between October 7, 2001, the start of OEF, and September 30, 2011 (N=314,717). Psychiatric diagnoses included depression (293.83, 296.20–296.25, 296.30–296.35, 300.4, and 311), posttraumatic stress disorder (PTSD) (309.81), other anxiety disorders (300.00–300.09, 300.20–300.29, and 300.3), adjustment disorders (308, 309.0–309.9, excluding 309.81), alcohol use disorders (305.00–305.03 and 303), and drug use disorders (305.20–305.93 and 304). Veterans were also categorized by number of comorbid psychiatric diagnoses (0 to ≥4).
These data were derived from electronic medical records generated during clinical visits. Visits to primary care or mental health services were categorized by using clinic stop codes (7,8). Fee basis codes designated care reimbursed by VA but rendered at non-VA facilities but did not capture all non-VA care (for example, care reimbursed by private insurance).
Time until initiation of treatment was defined as the number of years from the end of the last deployment until the first primary care or mental health outpatient visit. Time until initiation of minimally adequate mental health outpatient care was defined as the number of years until the first visit in a series of eight visits completed within any 12-month period. We were also interested in determining whether veterans experienced delays in receiving minimally adequate treatment once they had accessed mental health care. Therefore, time until initiation of minimally adequate mental health outpatient care was also defined as the number of years from the first mental health outpatient visit until the first mental health outpatient visit that was part of a series of at least eight visits within any 12-month period. Veterans who did not have any visits for primary or mental health care by the time of death or the end of the study, whichever came first, were censored.
Analyses of time until initiation of minimally adequate mental health outpatient care were restricted to veterans whose first clinic visit occurred at least one year prior to the end of the study (N=290,104) to allow adequate time for follow-up. Analyses of time from first mental health outpatient visit to initiation of minimally adequate care were restricted to veterans who had utilized mental health outpatient care at least once and who had been in the VA system for at least one year (N=254,541). In multivariable analyses the samples were slightly reduced because of missing data for some covariates.
The Kaplan-Meier (product-limit) method was used to calculate the nonparametric estimate of the survival function for initiation of primary care after last deployment, mental health outpatient care after last deployment, minimally adequate mental health outpatient care after last deployment, and minimally adequate care after first mental health outpatient visit. Median time to initiation—when 50% of the sample had initiated care—was calculated. The 95% confidence interval (CI) of the median was calculated on the basis of a log-transformed CI for the survivor function (14). Gender-specific survival functions for each time-to-initiation outcome were calculated and compared with the log-rank test of homogeneity. Therapeutic intensity among those who engaged in minimally adequate care was examined in two ways: the median time between the first and the last of the eight mental health clinic visits and the median of the patient-level median time intervals between successive visits.
We used Cox proportional hazards analysis to examine the association of independent predictor variables and time to initiation of primary care, mental health outpatient care, and minimally adequate mental health outpatient care. For each of the four initiation-of-care outcomes, a separate Cox proportional hazards regression model was built to consider each of the six comorbid psychiatric diagnosis predictors. Because PTSD and the other mental disorders are highly overlapping, it was not possible to include these six diagnoses simultaneously in a single model. All models were adjusted for age, race, military component (active duty versus National Guard or reserves), rank, branch, history of multiple deployments, other comorbid psychiatric diagnoses (ICD-9 codes 290–319 that are not among the six disorders listed above), distance from and type of nearest VA facility, and time in the VA system (time between first clinical encounter of any kind and initiation of primary care, mental health outpatient care, or minimally adequate mental health outpatient care, death, or end of the study period).
All tests were two tailed, and because of the large sample, statistics were considered significant if the associated p value was <.001. Analyses were performed by using SAS, version 9.2. The study was approved by the Committee on Human Research of the University of California, San Francisco, and the San Francisco VA Medical Center.
The sample was 88% male, with a median age of 26 years (interquartile range 22–35 years); its racial-ethnic makeup was 51% white, 11% black, 11% Hispanic, 2% Asian or Pacific Islander, and 24% other or unknown. The cumulative prevalence of psychiatric diagnoses was 63% PTSD, 57% depression, 38% anxiety disorder, 34% adjustment disorder, 26% alcohol use disorder, and 12% drug use disorder. A total of 32% of the sample had one psychiatric diagnosis; 30%, two; 22%, three; and 17%, four or more.
The median time from the end of last deployment until initiation of care was 1.47 years (CI=1.47–1.48) for primary care, 2.13 years (CI=2.12–2.14) for mental health outpatient care, and 4.1 years (CI=2.2–6.3) for minimally adequate mental health care (Figure 1). By the end of the study, less than 35% of veterans who had been patients of the VA for at least one year had initiated minimally adequate mental health care (Figure 1). Among veterans attending mental health outpatient care at least once, nearly 30% received minimally adequate care within a year of their first visit for mental health care (Figure 2).
Figure 1Time since last deployment to initiation of primary care, mental health outpatient care, and minimally adequate mental health outpatient care among 314,717 veterans with one or more psychiatric diagnoses, in yearsa
aKaplan-Meier estimates. Analyses of time until initiation of minimally adequate mental health outpatient care were restricted to 290,104 veterans whose first clinic visit occurred at least one year prior to the end of the study.
Figure 2Time from first mental health outpatient visit to initiation of minimally adequate mental health care among 254,541 veterans with one or more psychiatric diagnoses, in yearsa
aKaplan-Meier estimates. Analyses were restricted to veterans who had utilized mental health outpatient care at least once and who had been in the Veterans Affairs system for at least one year.
The median time in years to initiation of care for men and women is reported in Table 1. The time between last deployment and initiation of primary care was slightly longer for male versus female veterans (1.48 and 1.44 years, respectively, p=.004). Male veterans initiated mental health care slightly sooner than female veterans (2.10 and 2.39 years, respectively, p<.001). The time between first mental health outpatient clinic and initiation of minimally adequate care was two years longer for male veterans than for female veterans (8.02 and 5.98 years, respectively, p<.001).
Table 1Time to initiation of VA care among veterans with a psychiatric diagnosis, by genderaa
| Add to My POL
|Time to initiation||Log-rank|
|Variable||N||%||(M years)||95% CI||p|
|Last deployment to initiation of primary care||.004|
|Last deployment to initiation of mental health care||<.001|
|Last deployment to initiation of minimally adequate mental health carebb||.16|
|First mental health visit to initiation of minimally adequate mental health care||<.001|
Among veterans who engaged in minimally adequate mental health outpatient care, the median time between the first and eighth mental health clinic visit was 157 days (range 87–276 days). The median gap between each visit was 14 days (range seven to 23 days). Thus a majority of veterans who engaged in minimally adequate care did so within a 23-week (or six-month) period with a two-week gap between most visits.
Cox proportional hazards regression of time until initiation of care showed that some demographic and military service characteristics (Table 2) and psychiatric diagnoses (Table 3) were independently associated with the likelihood of earlier initiation of care. The hazard ratios (HRs) represent the ratio of the rates of initiating care by veterans with various characteristics and by members of a reference group, with HRs >1 representing more rapid initiation of care compared with the reference group.
Table 2Time to initiation of VA care among veterans with a psychiatric diagnosis, by demographic and military service predictoraa
| Add to My POL
|Last deployment to primary care||Last deployment to mental health care||Last deployment to minimally adequate care||First mental health care visit to minimally adequate care|
|Characteristic||Hazard ratio||95% CI||Hazard ratio||95% CI||Hazard ratio||95% CI||Hazard ratio||95% CI|
|Female (reference: male)||1.10*||1.09–1.11||1.01||1.00–1.02||1.12*||1.10–1.14||1.12*||1.09–1.14|
|Age at last deployment (reference: 18–24)|
|Race or ethnicity (reference: white)|
| Native American||.79*||.77–.82||.81*||.78–.84||.92||.86–.98||1.07||1.00–1.13|
| Asian/Pacific Islander||.84*||.82–.86||.84*||.81–.86||.94||.89–.98||1.05||1.00–1.09|
| Other or unknown||.63*||.62–.63||.60*||.59–.61||.76*||.75–.78)||1.18*||1.16–1.20|
|Marital status (reference: never married)|
| Divorced, separated, or widowed||.93*||.91–.94||.93*||.92–.95||.92*||.90–.95||1.00||.97–1.03|
|Military reserves (reference: active duty)||1.68*||1.67–1.70||1.63*||1.62–1.65||1.67*||1.64–1.69||1.20*||1.18–1.22|
|Officer (reference: enlisted)||.90*||.88–.92||.92*||.90–.94||.88*||.85–.91||.96||.93–.99|
|Branch of service (reference: Army)|
| Air Force||.63*||.62–.64||.61*||.60–.62||.64*||.62–.65||.96*||.93–.98|
| Navy and Coast Guard||.71*||.70–.72||.68*||.67–.69||.68*||.67–.70||.94*||.92–.97|
|≥2 deployments (reference: 1)||1.33*||1.32–1.34||1.35*||1.34–1.36||1.24*||1.22–1.25||1.01||1.00–1.02|
Table 3Time to initiation of VA care among veterans with psychiatric diagnoses, by psychiatric diagnostic predictoraa
| Add to My POL
|Last deployment to primary care||Last deployment to mental health care||Last deployment to minimally adequate care||First mental health visit to minimally adequate care|
|Predictor||Hazard ratio||95% CI||Hazard ratio||95% CI||Hazard ratio||95% CI||Hazard ratio||95% CI|
| Adjustment disorder||1.22*||1.21–1.23||1.40*||1.39–1.41||1.36*||1.34–1.38||1.07*||1.05–1.08|
| Alcohol use disorder||.99||.98–1.00||1.05*||1.04–1.06||1.49*||1.47–1.51||1.38*||1.36–1.40|
| Drug use disorder||.93*||.92–.95||1.09*||1.08–1.10||1.62*||1.59–1.64||1.46*||1.44–1.49|
|Comorbid psychiatric diagnosesdd|
Female veterans were more likely than male veterans to initiate primary care sooner (HR=1.10, p<.001). With the exception of alcohol use disorders, each psychiatric disorder was associated with an increased likelihood of initiating primary care earlier. A greater number of comorbid psychiatric disorders was associated with increased likelihood of initiating primary care sooner postdeployment.
All of the psychiatric diagnoses and the number of comorbid psychiatric disorders were associated with increased chances of earlier initiation of mental health outpatient care postdeployment. Adjustment disorder was the psychiatric diagnosis most strongly associated with earlier initiation of mental health outpatient visits (HR=1.40, p<.001).
Women were more likely than men to initiate minimally adequate mental health outpatient care sooner after the last deployment (HR=1.12, p<.001). All of the psychiatric diagnoses and the number of comorbid psychiatric diagnoses were associated with an increased chance of initiating minimally adequate mental health outpatient care sooner, and the association was strongest for PTSD (HR=2.27, p<.001). Among veterans with more than one psychiatric diagnosis, each additional diagnosis increased by one-and-a-half the likelihood of initiation of repeated mental health outpatient use after last deployment.
Other factors associated with greater delay between last deployment and initiation of primary care, mental health outpatient care, and minimally adequate mental health outpatient care were age younger than 25 years; being married; active duty; officer rank; all other branches of service besides Army; and single deployment. Being black and Hispanic was also associated with a greater delay between last deployment and initiation of care, except after the first mental health outpatient visit, when black and Hispanic veterans were more likely than whites to initiate minimally adequate care.
Delay in initiation of care was also associated with greater distance from the nearest VA facility, type of VA facility nearest to the veteran (community-based clinic versus VA medical center), and longer time in the VA system (data not shown).
Among veterans of the conflicts in Iraq and Afghanistan, the median time between the last deployment and engagement in mental health care was over two years. This period was longer by more than one-half year than the median period between the last deployment and engagement in primary care, where screening for and brief evaluation of mental health problems occur. Length of time to minimally adequate mental health treatment was even longer.
After over three years postdeployment, 75% of veterans with a psychiatric diagnosis who were in the VA system for at least one year had not engaged in minimally adequate mental health care. These results demonstrate that veterans who may benefit from mental health care are waiting several years to access care. This is particularly problematic given the benefits of early intervention for mental health problems and for PTSD in particular (15,16). Given that the largest proportion of veterans who received minimally adequate care did so within one year of a first visit, motivational and other interventions could be targeted at veterans who do not engage in minimally adequate care in the first year.
There was also a median lag time of nearly 7.5 years between presenting for an initial mental health treatment session and initiating a course of minimally adequate mental health care. Furthermore, male veterans waited nearly two years longer than female veterans to initiate minimally adequate care, highlighting gender differences in accessing care. One possibility is that younger men, and specifically those in the military, may be particularly susceptible to negative beliefs about mental health and treatment seeking (17). Placed in a broader context of gender differences, these findings suggest that men may subscribe to ideologies, norms, and gender roles related to masculinity that are associated with barriers to care and inversely associated with treatment seeking (18). However, given a median lag time of 7.5 years among all veterans, understanding why veterans do not return for ongoing care is a priority.
Prior studies have suggested that some of the most common barriers to care include fear of being labeled as weak or “crazy,” potential negative impact on career, pride, and a belief that one should be able to handle these issues on one’s own (1,10,19). One study suggested that social network support can be an important facilitator of care, despite individual-level barriers (20). One framework that can be used to conceptualize these findings at a systemic level suggests that optimal outcomes of care require addressing not only engagement gaps, such as individual- and social-level barriers, but also gaps in access and quality. An example of a gap in access is wait time, and an example of a gap in quality is availability of evidence-based care or technologically advanced treatments (21). Thus compared with VA medical centers, community-based clinics may have slightly longer wait times and decreased availability of evidence-based care because of fewer staff. Such gaps should be examined in future investigations.
Even among veterans who engaged in minimally adequate care, completion time was about six months, with a two-week lag time between sessions, potentially diluting therapeutic impact. This is particularly important, given that delivery of evidence-based PTSD treatment that involves trauma exposure (prolonged exposure therapy and cognitive-processing therapy) relies on the consistency and intensity of session delivery. Furthermore, although barriers to initially seeking mental health care have been explored to some extent (1,10,17,19,22), the barriers to staying engaged in mental health care—and at consistent time intervals—may be different. To date, no studies have examined psychosocial barriers to retention in mental health outpatient care.
Other demographic variables were also associated with delays in initiation of mental health treatment. For example, younger veterans took longer to initiate mental health outpatient care and to seek minimally adequate care, which is consistent with prior studies (23). Members of racial and ethnic minority groups took longer than Caucasians to initiate mental health outpatient care and to seek minimally adequate care; lower rates of utilization of mental health treatment by and ethnic minority groups have been noted in prior studies (24–26). In certain communities, a sense of pride or a lack of social network facilitators may be particularly important and even function as barriers to treatment. Factors such as male gender, younger age, and racial and ethnic minority status will be important to consider when designing and targeting interventions to decrease barriers to care.
Reduced time to initiation of mental health outpatient care and a course of minimally adequate mental health outpatient care was strongly associated with PTSD. Although this finding is encouraging, given that PTSD has been shown to be a barrier to seeking care (1), it is also surprising because prior studies have found that veterans with PTSD are more likely to seek primary care and underutilize mental health care (27,28). In fact, the National Comorbidity Survey Replication found that among individuals with PTSD, the delay between onset of diagnosis and treatment engagement was 12 years (9). Members of a Canadian military sample with PTSD waited a median of seven years from diagnosis to first mental health treatment contact (29).
One possible explanation for our finding is that PTSD is highly comorbid with other psychiatric diagnoses. We found that having more psychiatric diagnoses signaled greater need and, consequently, a shorter delay in initiation and maintenance of mental health outpatient care. Reduced time to initiation of adequate mental health outpatient care was positively associated with substance use disorders, demonstrating that initiation of treatment may take longer, but once engaged, individuals with substance use disorders are likely to complete a course of treatment.
A number of limitations should be considered when interpreting these findings. First, this study was conducted with a population of treatment-seeking veterans who had at least one visit to a VA health care facility. Therefore, our results should not be generalized to all OEF/OIF/OND military personnel or veterans. Second, we selected a population of veterans who served in support of OEF/OIF/OND. Therefore, these results should not be generalized to veterans of other eras or to veterans from other countries. Third, active-duty military personnel might not have the opportunity to utilize VA services between their last deployment and separation from the military, and we were not able to access military health data.
Fourth, ICD-9-CM diagnostic codes were acquired from administrative health records and were not verified with standardized diagnostic measures. There also may be important variables that were not measured but that may account for some of these differences. For example, we utilized number of deployments to index combat exposure in lieu of detailed information on level of exposure. We also were not able to measure variables such as employment status, social support, and other variables that index functioning. It may be that the emergence of these problems later, rather than immediately postdeployment, may influence the timing of care seeking.
Finally, different types of mental health outpatient visits (for example, psychiatry or psychotherapy) were grouped together. Although there is a precedent for doing so, future research could look at types of visits separately to better understand time to particular types of outpatient care.
A majority of OEF/OIF/OND veterans with psychiatric diagnoses waited over two years to initiate mental health care. After more than three years, 75% of veterans with psychiatric diagnoses had not yet engaged in minimally adequate mental health treatment at the VA. These findings highlight the importance of attending to timing of care, particularly among newly returning veterans, so that the VA and other health care systems serving veterans can begin to address individual- and system-level barriers that delay care. In turn, this information can help to design early interventions to reduce chronic mental health problems and improve functioning among recently returned veterans.
This research was supported by U.S. Department of Defense Concept Award Grant W81XWH-08-2-0077 and by VA Health Services Research and Development Career Development Award RCD 06-042 to Dr. Maguen.
The authors report no competing interests.