PTSD Treatment for Soldiers After Combat Deployment: Low Utilization of Mental Health Care and Reasons for Dropout

Over two million service members have deployed to Iraq or Afghanistan since 2001, and these deployments have been strongly associated with an increased risk of mental health problems (1–4). A meta-analysis of studies found that the average postdeployment prevalence of posttraumatic stress disorder (PTSD) was 13.2% for personnel assigned to operational infantry units and 5.5% for representative samples of total deployed forces (including support personnel) (3). These and other studies indicate that there will be a significant ongoing need for mental health care in this population.

Although a relatively high percentage of military personnel and veterans access mental health services (for example, one analysis of Army personnel showed that 21% had one or more clinical encounters during a year) (5), underutilization for those most in need remains an ongoing concern. Studies have shown that only 13%−53% of U.S. and Canadian veterans who meet criteria for a mental health problem after deployment receive care (6–8).

An additional problem is that veterans who enter mental health treatment often do not receive adequate care. At least three studies have found that only a third of Iraq and Afghanistan veterans treated for PTSD received minimally adequate care (8–10), broadly defined by the number of treatment sessions received. Similar studies involving active duty service members close to returning from deployment have not been conducted.

A number of explanations have been proposed for underutilization of services, including stigma (1,6,7), lack of appointment accessibility and availability (1,6,11), perceptions of self-reliance (12), and distrust or negative perceptions of care (6,13–16). Several recent studies among U.S. and Canadian veterans have suggested that negative attitudes toward mental health care may be more important in initially seeking treatment than stigma and other traditional barriers (13–16).

With this descriptive study, we offer new findings on the adequacy of PTSD treatment received by active duty service members after returning from combat deployment. The principal study aims were to determine the percentage of soldiers in need of PTSD treatment after returning from deployment and the percentage receiving an adequate number of treatment sessions according to a standard definition of minimally adequate care. A secondary aim was to explore reasons for dropping out of care, including negative attitudes toward mental health treatment.

Methods

Study groups

Data came from two very different but complementary sources: a population-based Army cohort and a cross-sectional unit-based survey. The cohort involved all Army active-component service members returning from Afghanistan between January 1, 2010, and December 31, 2010, who completed the mandatory Post-Deployment Health Assessment (PDHA) (N=45,462). This study focused on service members returning from Afghanistan, rather than Iraq, because the Afghanistan war zone had the most significant combat engagements during the study time frame. The PDHA is conducted just before leaving the combat theater or just after service members return home. The PDHA involves a brief self-assessment using standardized screening tools (including PTSD screening) followed by a health care encounter with a primary care clinician who determines the need for further referral (17). Analysis focused on soldiers who received a clinician diagnosis of PTSD within 90 days of completing the PDHA. All subsequent health care utilization within the military medical system documented in these soldiers’ electronic medical records was measured for one year after return from deployment. A principal purpose of analysis of this cohort was to determine the proportion of soldiers who received a minimally adequate number of treatment encounters after PTSD diagnosis. Analysis was conducted under a protocol approved by the Walter Reed Army Institute of Research.

The second study group included soldiers from an infantry brigade surveyed confidentially in July 2011, four to five months after returning from an Afghanistan combat deployment. Soldiers were recruited by coordinating with unit commanders, who made soldiers available in their work areas for group recruitment briefings during which they could voluntarily consent to participate. The brigade census showed a total of 3,832 soldiers on duty at the time of recruitment, and 2,876 (75.1%) consented to participate. Of the 2,876 soldiers, 2,420 had deployed with the brigade to Afghanistan and were included in this study (the rest were new brigade members or transfers, who were excluded). The analysis of the survey data focused on characterizing health care utilization overall, and for soldiers who screened positive for PTSD. The principal purpose of the analysis was to identify how many high-risk infantry soldiers reported receiving mental health services, how many visits occurred, and, for those who reported dropping out of care, the key reasons for dropout. Informed consent was obtained under a protocol approved by the institutional review board of the Walter Reed Army Institute of Research.

Outcome measures

For the population cohort, all health care utilization during the year after the PDHA was measured with the electronic military Defense Medical Surveillance System, which includes records of all health care visits and associated ICD-9-CM diagnoses at all military health care facilities within or outside the continental United States (2,17). The cohort with PTSD was defined as all soldiers who received ICD-9-CM code 309.81 (any diagnostic position) from any health care encounter within 90 days of the PDHA. One encounter with this diagnosis was considered diagnostic for this descriptive study because of research showing that requiring two or more encounters produces minimal gains in predictive value but major reductions in sample size (18,19). In addition, the highest dropout rates tend to occur after the first visit (17), and there are unique diagnostic considerations in the military health care system (stemming from efforts to reduce stigma and high clinical focus on PTSD) that add to the likelihood that one ICD-9-CM PTSD code is sufficiently accurate (2). Minimally adequate care for PTSD was defined per previous research as receiving eight or more health care encounters involving this diagnosis in the 12-month follow-up period (8–10).

For the infantry survey group, PTSD symptoms were measured with the 17-item PTSD checklist (20,21). To meet PTSD screening criteria of DSM-IV-TR, soldiers had to report at least one intrusion, three avoidance, and two hyperarousal symptoms at the moderate or higher level and have a total score of ≥50 on a scale of 17–85. This well-established stringent cutoff is used widely in the military and found optimal for population studies (1,3,21).

Health care utilization for the infantry sample was measured with a series of questions that asked soldiers whether they had received mental health services for a stress, emotional, alcohol, or family problem in the past six months from a mental health, primary care, or military OneSource provider at any military, civilian, or Veterans Affairs (VA) health facility or vet center. Soldiers were also asked how many total visits they had with a mental health professional in the past six months and whether they were currently in mental health treatment. The six-month period was selected to include the four to five months since return from deployment as well as the final one to two months of deployment, when medical screenings, including the PDHA, are first initiated in preparation for returning home. Extensive mental health services are available in the combat theater and are used in conjunction with the PDHA to ensure coordinated care during transition from deployment to home. A minimally adequate number of mental health visits was defined as four or more visits in the past six months (8–10). Participants were also asked whether they had stopped treatment or dropped out before completing treatment. Those who reported dropping out of care were asked a series of additional yes-no questions concerning their reasons. These questions were informed by clinical experience and built on previous research on stigma, treatment barriers, and negative perceptions of mental health care (1,7,12–16).

Surveys were scanned with ScanTools (National Computer Systems), and quality control processes verified error rates below .25%. Analysis, which was largely descriptive, was conducted with SPSS version 12.0 for the surveys and with SAS version 9.1 for the cohort.

Results

Overall, the two study groups had comparable demographic characteristics, with the largest proportion being young, male, and junior enlisted rank (Table 1). The demographic characteristics closely matched those of other studies of deployed active duty personnel (1–4), and PTSD prevalence was also consistent with previous reports (3,4).

Mental health outcomes of the PDHA population cohort

Of the 45,462 soldiers completing the PDHA on return from Afghanistan, 4,674 (10.3%) were referred for further evaluation or treatment for any mental health problem (of whom 3,514, or 75%, followed up with this referral within 90 days), 15,094 (33.2%) had one or more mental health encounters through another referral mechanism (including primary care, self-referrals, and command-directed referrals) within 90 days, and 2,230 (5.0%) (from all referral sources) received a PTSD diagnosis from a clinician. Of the 2,230 soldiers who received a PTSD diagnosis within 90 days of their PDHA, 1,962 (88.0%) were able to be followed for a complete 12-month period with measurement of the total number of mental health or primary care encounters in which PTSD was listed as a diagnosis (the other 12% left military service before 12 months had lapsed). Table 2 shows the distribution of the total number of visits in which the PTSD diagnosis was recorded; 22% of soldiers received only one visit, 59% received four or more, and 41% received eight or more encounters over 12 months.

Mental health outcomes of the infantry survey group

Table 3 shows self-reports of mental health care utilization for all 2,420 soldiers as well as the 229 who met strict screening criteria for PTSD. Overall, 21% of soldiers reported receiving mental health services for any stress, emotional, alcohol, or family problem through any type of provider; 6% reported receiving a psychiatric medication, most commonly an antidepressant; and 17% received at least one visit with a mental health professional in the past six months (median of two visits). Of soldiers who accessed care through a mental health professional, 42% had only one visit. Of the 229 soldiers who met strict criteria for PTSD, 48% reported receiving any mental health services in the past six months, and 42% received care from a mental health professional in the past six months; 22% had only one visit, and 52% had four or more visits in the past six months (median of four visits). Soldiers with PTSD who reported being prescribed a psychiatric medication had a significantly greater number of mental health care visits than soldiers with PTSD who were not prescribed medication (median of six visits, interquartile range [IQR]=3–12, versus median of three visits, IQR=1–5; p<.001). Overall satisfaction with care was moderate and was somewhat higher for all soldiers in current treatment compared with those who screened positive for PTSD, with 79% versus 67%, respectively, responding that they were somewhat or very satisfied.

Of the 507 total soldiers and the 106 with PTSD who reported receiving any mental health services, 53 (11%) and 25 (24%), respectively, answered yes to the question, “Did you start receiving mental health treatment anytime in the past six months, but stopped or dropped out before completing the treatment?” Of these soldiers, 50 and 23, respectively, endorsed one or more reasons for dropping out (Table 4). A majority of these soldiers reported multiple reasons (median=4.5, IQR=2.75–8.00, for the 50 total soldiers and median=7, IQR=4–10, for the 23 soldiers who screened positive for PTSD). The distribution of responses for each of these 23 soldiers is shown in Table 5. The most common individual reasons included perceptions of self-sufficiency, not having sufficient time with the professional, lack of appointment availability, being too busy with work, and concerns about stigma. However, negative perceptions of the interaction with the clinician were also common; of the 23 soldiers, 15 (65%) endorsed one or more of five concerns related to how the professional communicated or interacted with them.

Discussion

Fostering engagement and willingness to remain in mental health treatment is critical to ensure the provision of evidence-based treatment to service members and veterans. This study provided important new findings, based on both cohort and cross-sectional methods, on the willingness of active duty soldiers to engage in and continue with needed treatment after combat deployment and provides additional qualitative data on reasons for dropping out of care. Despite the very different data collection methods, the two study groups had similar demographic characteristics (Table 1) and provided remarkably complementary findings on health care utilization and adequacy of treatment.

Among the large cohort of soldiers who completed the clinical PDHA process, 10% were referred for further mental health evaluation, and 75% of these had documentation in their electronic health records of following up with this referral. This 75% rate is significantly higher than has been reported previously (42% was reported in 2007 [17]), suggesting that efforts to improve postdeployment screening have been successful. However, despite this finding, the overall treatment reach for those most in need is estimated to remain low. Of 2,230 soldiers who received a PTSD diagnosis within 90 days of the PDHA, most did not have an adequate opportunity for evidence-based care; 22% had only one mental health visit (the one in which the diagnosis was made), and 41% received minimally adequate care, defined as eight or more visits involving this diagnosis within the ensuing year. Furthermore, previous research has documented that many soldiers returning from deployment are not willing to disclose concerns during the initial clinical PDHA process (22).

Data from the cross-sectional infantry sample complemented the cohort findings. Among the 229 soldiers who screened positive for PTSD under strict case criteria, only 106 (48%) reported receiving any mental health care, and 25 (24%) soldiers reported dropping out of care. Overall satisfaction with care was moderate, with nearly a third reporting dissatisfaction. Among the 95 infantry soldiers who met criteria for PTSD and accessed care with a mental health professional, the total number of visits reported was strikingly similar to the much larger PDHA cohort; 22% of these soldiers reported receiving only one visit, which was identical to the percentage in the cohort based on documented encounters; 52% met the study definition for minimally adequate care, compared with 41% in the cohort. The overall reach of treatment was very low. Of all 229 infantry soldiers who screened positive for PTSD, only 49 (17%) received treatment that would be considered adequate, with the remainder either not receiving any care or receiving an insufficient number of sessions.

The definition of minimally adequate mental health treatment used in this study was a composite of definitions from the literature, which have included the criterion of four or more pharmacotherapy encounters in any clinic over either a six-month (10) or 12-month (8,23) period, eight or more psychotherapy encounters over six (10) or 12 (8,23) months, or nine or more encounters (either psychotherapy or pharmacotherapy) in a PTSD-specific Veterans Health Administration (VHA) clinic over 12 months (9). However, all of these definitions have to do with only a minimally acceptable dose of care, not treatment adequacy, especially with the chronicity and comorbidities associated with PTSD and changing standards of clinical practice (24). These definitions are crude estimates of what should be considered a minimal number of sessions necessary for provision of evidence-based care. Nevertheless, they have produced comparable results in veteran studies, with estimates ranging from 30% to 33% (8–10).

PTSD psychotherapy modalities typically involve weekly treatment sessions spanning 12 weeks. Pharmacotherapy treatment usually involves a number of sessions in the initial higher risk period (for example, four to six visits over the first 12 weeks) to ensure appropriate titration of medication dose and monitoring for suicidal ideation (due to FDA black-box warnings); thereafter, follow-ups typically occur every one to three months. Contrary to assumptions used in some definitions of “minimally adequate treatment,” we found that individuals with PTSD who were prescribed medications reported a significantly greater number of encounters than those not prescribed medications; this may reflect greater disease severity or more intensive follow-up for medication management according to revised standards. Recovery from PTSD in both psychotherapy and pharmacotherapy randomized clinical trials can reach 70%−80% among individuals who complete treatment (which usually involves at least eight visits, even for pharmacotherapy trials). However, dropout plagues virtually every treatment trial, leading to average recovery rates in intent-to-treat analyses of only around 40% (24,25). Furthermore, the total number of encounters alone says nothing about the provision of evidence-based strategies, the quality of treatment, or whether the focus of visits was even related to the index diagnosis instead of other comorbid conditions. Thus it is unlikely that either four sessions in six months or eight in 12 months truly reflects an adequate opportunity to receive evidence-based care with either psychotherapy or pharmacotherapy. The actual percentage of patients who receive adequate evidence-based care is therefore unknown but is likely to be lower than the 30%−33% estimate in previous veteran studies or the 41%−52% estimate in this study of active duty personnel.

To add to the above concerns of low treatment reach, this study provides new data on the specific reasons soldiers report for dropping out of mental health care. It is possible that the survey did not identify all who dropped out of treatment. Soldiers who missed follow-up appointments but intended to eventually return to treatment may not have endorsed the question on dropping out of care. However, despite this limitation, the 24% dropout rate among soldiers who met criteria for PTSD was comparable to rates found in clinical trials and civilian studies (24,26,27). The wide range and high number of responses endorsed by each participant was impressive, spanning a variety of domains.

Concerns most commonly reported by soldiers included feeling like they could take care of problems on their own, not having sufficient time with the professional, work interference, stigma, confidentiality concerns, and the belief that care was ineffective. Two-thirds of soldiers also expressed discomfort with the interpersonal interaction with the mental health professional, including the perception that the professional was not suitably caring, communicative, or competent; soldiers sometimes felt judged or misunderstood. These data add to studies of the predictors of initial treatment access that have shown that negative attitudes may be more important than stigma perceptions (13–16), including civilian data showing that an important predictor of dropping out of treatment is the belief that treatment will not be effective (26).

Limitations of this study include the reliance on administrative data for the cohort and self-report data for the cross-sectional survey. However, the large sizes of both study groups, high survey response rate, and especially the consistency in findings between the cross-sectional surveys and longitudinal clinical records strongly support the methodology and conclusions. The study provides a unique integration of findings from different sources related to mental health care utilization, adequacy, and satisfaction, as well as soldiers’ perceptions of care. Although there are some unique aspects of treatment immediately after exposure to traumatic events in the war zone, by the time service members return home, access care, and receive a diagnosis of PTSD, the standard of treatment remains eight or more encounters (24). This study did not address potential benefits of brief or stepped-care interventions, which are being studied in primary care settings and which leverage strategies, such as motivational interviewing, behavioral activation, care management, phone follow-up, initiation of antidepressants, and treatment of comorbid sleep disturbance (28).

This study represents a call to action to develop and test interventions to improve perceptions of mental health care and treatment engagement and retention in military, VHA, and civilian treatment settings. Dropping out of care is clearly the most important predictor of treatment failure; therefore the most promising strategies to improve efficacy of evidence-based treatments will be those that address engagement, therapeutic rapport, and retention. Particular attention is needed to better understand the modifiable organizational, patient, and clinician factors and specific actions that clinicians and health care systems can take. Interventions related to organizational barriers include ensuring adequate appointment availability and duration at convenient times and locations, as well as peer-to-peer outreach. Strategies to address patients’ beliefs about treatment should consider perceptions of self-reliance (for example through motivational interviewing techniques) (29–31). Policies concerning confidentiality, especially for treatment of comorbid substance use disorders, remain an ongoing issue in the military (32). Clinician factors that warrant close examination concern the skills and training needed to optimally foster patient-centered care.

In a comprehensive review, Swift and colleagues (27) suggested six strategies to minimize client dropout in civilian psychotherapy settings. These strategies include providing information to clients about therapy duration and expected patterns of change, educating clients about the roles and behaviors of the therapist and client, incorporating the client’s preferences for treatment, strengthening the client’s early hopes for therapy, fostering the therapeutic alliance and sustaining rapport, and assessing and discussing treatment progress at intervals. Many of these principles are inherent in patient-centered care. Factors likely to be particularly important in military and veteran populations include the ability of the clinician to communicate in way that is sensitive to the military occupational context and providing as wide a range of treatment options as possible (24,25,33). Establishing ongoing simple measures of patient feedback is likely to be helpful (34). Military health care systems should reevaluate how mental health treatment programs are structured and marketed. Embedding mental health treatment within primary care settings and coordinating care between primary and specialty care are also important strategies (35). It is particularly important to consider integrated stepped approaches that enhance engagement through brief low-intensity treatments in primary care settings before stepping up to a specialty setting (28,36,37).

Conclusions

This research showed that the overall reach of mental health services for deployment-related PTSD remains low to moderate, despite the availability of extensive screening and treatment services, as well as measurable increases in mental health care utilization since the beginning of the conflicts in Iraq and Afghanistan. The study highlights important priorities for clinical interventions research. Improving perceptions of mental health care and fostering therapeutic rapport, engagement, and retention offer the greatest potential for improving overall treatment effectiveness.