Abstract
Objective:
Most veterans with PTSD do not receive individual psychotherapy. The purpose of this study was to explore gender differences in individual psychotherapy initiation and completion (defined as attending 8 or more sessions) among male and female Veterans Health Administration (VHA) users recently diagnosed with PTSD.
Method:
Participants (N=7,218) were veterans in a prospective national cohort survey of VHA users diagnosed with PTSD; oversampling was used to increase representation of women and minority veterans.
Results:
Forty-two percent of the sample (40.1% of men, 52.3% of women) initiated individual psychotherapy within six months of their index PTSD diagnosis. Of those who initiated, 12.1% (10.8% of men, 17.7% of women) completed a sufficient course of individual psychotherapy. Women were generally more likely than men to initiate individual psychotherapy. However, we found an interaction between gender and age, such that younger men were more likely to initiate psychotherapy than older men; age was not significantly associated with initiation among women. Regarding completion of individual psychotherapy, an interaction between gender and beliefs about psychotherapy was found, such that men were less likely to complete individual psychotherapy when they held more negative beliefs about psychotherapy; these beliefs did not significantly impact female veterans’ likelihood of completing psychotherapy.
Conclusions:
Overall, while female veterans are more likely than male veterans with PTSD to initiate individual psychotherapy, rates of initiation and completion of individual psychotherapy for both genders remain relatively low. Interventions are needed to increase engagement in individual psychotherapy, particularly for male veterans with PTSD.
Keywords: PTSD, Individual Psychotherapy, Veterans, Gender Differences
Posttraumatic Stress Disorder (PTSD) is a potentially disabling mental health condition with far ranging consequences including physical health problems, social impairment, and psychological sequelae such as increased suicidality (Spitzer et al., 2009; Nepon, Belik, Bolton, & Sareen, 2010). Up to a quarter of veterans have PTSD (10%-23%; Fulton et al., 2015; Kang, Natelson, Mahan, Lee, & Murphy, 2003), a rate substantially higher than the general US population (lifetime prevalence rate of 6.8%, Kessler et al., 2005). Furthermore, for many veterans PTSD symptoms do not naturally remit over time; longitudinal studies of veterans have found that PTSD symptoms can persist, and even worsen over time (Orcutt, Erickson, & Wolfe, 2004; Marmar et al., 2015). Fortunately, there are effective treatments for PTSD. Current practice guidelines for the treatment of PTSD recommend manualized individual trauma-focused psychotherapy (VA/DoD, 2017), which has been found to be more effective than other PTSD treatments such as group psychotherapy or pharmacotherapy (Bisson, Roberts, Andrew, Cooper, & Lewis, 2013; Lee et al., 2016). However, among veterans with PTSD receiving health care in the Veterans Health Administration (VHA), only one-third (34%) initiate individual psychotherapy (Doran, Pietrzak, Hoff, & Harpaz-Rotem, 2017). A much smaller subset of these individuals receive an adequate dose (~10%, Cully et al., 2008), defined as receiving eight or more psychotherapy sessions, a threshold used in previous research to signify minimally adequate treatment (Maguen, Madden, Cohen, Bertenthal, & Seal, 2012). Thus, given the high rates of PTSD among veterans, the potentially chronic nature of the disorder, and the availability of effective treatments, improving rates of treatment initiation and completion is of considerable importance for this population.
Social Cognitive Theory (Bandura, 1977; Bandura, 2004) is a useful framework for understanding factors that may influence male and female veterans’ decisions to initiate and complete PTSD treatment. Social Cognitive Theory posits that behavior (e.g., seeking treatment) is influenced by both the environment, such as access and barriers to treatment, as well as personal factors, such as an individual’s attitudes related to the behavior and perceived behavioral control (e.g. self-efficacy). Prior research has mapped relationships between environmental factors as well as beliefs and self-efficacy related to mental health and mental health treatments on treatment-seeking behavior in a veteran population. For example, a longer drive time to the nearest VHA facility and receiving a PTSD diagnosis in a general, non-mental health clinic are associated with a lower likelihood of receiving psychotherapy among veterans with PTSD (Spoont et al., 2014). Additionally, veterans’ beliefs about mental health treatment impact treatment initiation and completion, with negative beliefs about mental health care and treatment-seeking associated with a decreased likelihood of service use (Vogt, Fox, & Di Leone, 2014; Pietrzak et al., 2009), and positive beliefs about psychotherapy and greater self-efficacy to follow treatment recommendations associated with increased odds of receiving psychotherapy (Spoont et al., 2014).
Both environmental and personal factors may influence PTSD treatment seeking differently for male and female veterans. For example, among female VHA primary care users, negative gender-related experiences such as feeling uncomfortable or out of place due to being a woman were associated with lower perceived access to mental health care (Kimerling et al., 2015). Moreover, beliefs about treatment are associated with gender differences in accessing mental health services. In a sample of Operation Enduring Freedom/Operation Iraqi Freedom (OEF/OIF) veterans, women’s, but not men’s, positive perceptions of VHA care were associated with an increased likelihood of seeking mental health treatment (Fox, Meyer, & Vogt, 2015). For male, but not female, veterans, negative beliefs about mental illness were associated with a lower likelihood of mental health service use (Fox et al., 2015). Additionally, there is evidence that female veterans with PTSD may be more responsive to cognitive behavioral interventions focused on increasing treatment seeking and engagement via the modification of treatment-related beliefs and barriers than their male peers (Gallegos et al., 2015). Thus, male and female veterans may require different intervention targets and strategies to increase initiation and engagement in PTSD treatment.
While previous research has used electronic health records to identify important predictors of male and female veterans’ mental health treatment utilization (e.g., Doran et al., 2017; Cully et al., 2008), few studies have combined electronic health record utilization data with survey data that assessed male and female veterans’ beliefs and attitudes about mental health treatment (Spoont et al., 2014, Spoont et al., 2015). Using only electronic health record data does not allow for an understanding of important contextual factors (e.g., attitudes) related to mental health treatment, while using only survey data limits the ability to precisely track service usage. Linking data from these two sources can yield important insight into the treatment beliefs underlying gender differences in rates of individual psychotherapy initiation and completion among veterans. Research has found that beliefs and attitudes towards treatment may be modified through a brief cognitive behavioral intervention designed to address logistical barriers and challenge treatment-interfering beliefs (e.g., “going to treatment means I cannot handle my problems”), and can increase individuals’ intentions to seek treatment (Stecker, Fortney, & Sherbourne, 2011). Thus, it is particularly important to understand how they impact treatment utilization among male and female veterans with PTSD.
The current study is a secondary analysis of Spoont and colleagues’ (2014; 2015) research on factors influencing treatment utilization in veterans recently diagnosed with PTSD, which found that VA users’ race/ethnicity, beliefs about their need for help, and their social networks’ encouragement to seek treatment were associated with initiation and retention in mental healthcare. The aim of the current study was to explore gender differences in individual psychotherapy initiation and completion of at least eight sessions among veterans recently diagnosed with PTSD. To do so, we examined environmental and personal predictors of treatment initiation and completion drawn from Social Cognitive Theory and found to be important in previous research in this population (e.g., Spoont et al., 2014). We also examined if veteran gender moderated the relationship between these predictors and individual psychotherapy initiation and completion.
Methods
Participants and Procedure
Participants (N=7,218) were veterans in a prospective national cohort survey of VHA users recently diagnosed with PTSD in a VHA outpatient visit between June 2008 and July 2009 (Spoont et al., 2014; Spoont et al., 2015). To sample individuals who were initiating a new course of mental health treatment, veterans were excluded if they had been prescribed antidepressants or antipsychotics, or had attended a mental health appointment other than for chemical dependency treatment in the past year. Veterans with a diagnosis of a moderate to severe cognitive or schizophrenia spectrum disorder and those without a mailing address were also excluded (Spoont et al., 2014). Oversampling was used to increase representation of women and minority veterans; all women, all Latino men, and men of any non-African American minority race were sampled. White, African American, and men of unknown race were randomly sampled according to the sampling fractions of 0.1, 0.19, and 0.51, respectively (Spoont et al., 2015).
VHA administrative data were used to assess participant age, gender, OEF/OIF status, drive time to clinic, and initiation and completion of individual psychotherapy in the six months following the recent PTSD diagnosis. Survey data were collected to assess race/ethnicity, trauma type and symptomatology, and other environmental and personal variables. In an effort to survey participants soon after this index PTSD diagnosis and prior to treatment initiation, an electronic query of veterans with a new PTSD diagnosis was conducted every two weeks via the National Patient Care Database (NPCD) and surveys were sent to these individuals. Survey data collection was conducted using a modified Dillman approach: an introductory letter explaining the study sent within two weeks of the index PTSD diagnostic visit, followed two days later by a cover letter with a study description and informed consent information, the survey, and a $10 cash payment. Non-respondents were sent a second cover letter and survey 10 days later and, if they did not respond to the second survey, a third cover letter via Federal Express 10 days after that (Spoont et. al., 2014). Baseline surveys were sent to 12,952 veterans, of which 8,492 surveys were returned (response rate = 66%). Of those, 847 were excluded because they no longer met study inclusion criteria due to updated appointment/pharmacy administrative information. An additional 427 veterans are excluded in the analysis presented here because they had received an initial PTSD diagnosis outside of a mental health or primary care clinic, resulting in a final sample of 7,218 veterans. There were no differences between survey responders and non-responders in the odds of initiating individual psychotherapy or receiving a sufficient amount of treatment. Survey responders were more likely to be older and to vary from non-responders in characteristics associated with older veterans including being non-Hispanic and less likely to have served in Operation Enduring Freedom or Operation Iraqi Freedom.
Measures
The study outcomes – initiation of individual psychotherapy and completion of at least eight sessions – were abstracted from National Patient Care Database (NPCD). Individual psychotherapy sessions were identified by therapy-related procedure codes (CPT) for which the provider was a mental health specialist. Initiation of psychotherapy was defined as at least one psychotherapy visit in the six-month follow-up period. Predictors of individual psychotherapy initiation and persistence were organized into three conceptual groupings of variables: demographics and symptomatology, environmental, and social cognitive variables.
Demographics and symptomatology variables.
Participants’ age and gender were abstracted from NPCD data. Veterans indicated race and ethnicity on the survey, which allowed for multiple endorsements. Survey participants also provided information on the type of trauma experienced that was associated with their current PTSD symptomatology. Trauma type was categorized as combat-related, unwanted sexual experience, both combat and unwanted sexual experience, or other type of trauma. Level of PTSD symptoms was determined by summing the 17-item PTSD Checklist – Military Version (PCL-M; Weathers, Litz, Herman, Huska, & Keane, 1993). The PCL-M has demonstrated good internal consistency (Cronbach’s α = .96 and .97 for full scale) and convergent validity (r = .85 and .93) in samples of Vietnam and Persian Gulf veterans (Weathers et al., 1993). The PCL-M demonstrated good internal consistency in the current sample as well (Cronbach’s α = .94). Mental health functioning and physical health functioning were measured by the standardized scores from the Short Form Health Survey’s mental health and physical health components, respectively (SF-12; Ware, Kosinski, & Keller, 1996). The SF-12 has demonstrated good test-retest reliability (r = .89) and accounts for more than 90% of variance explained by a longer (36-item) version of the measure (Ware et al., 1996).
Environmental variables.
The facility and clinic type in which participants were seen at the time of the index diagnosis were extracted from NPCD data. Clinic type was coded as Primary Care Clinic (reference group), PTSD Clinic, or Mental Health Clinic. Participants’ drive time (in minutes) from their home to the nearest VHA facility was derived from the Planning Systems Support Group database. Because OEF/OIF veterans were given enhanced eligibility for health care, OEF/OIF status was considered an environmental variable and was collected from administrative data.
Social cognitive variables.
Perceived need for mental health help was assessed by the agree/disagree statement “At this time I feel I need help to deal with emotional problems, PTSD, and/or stress in my life.” Perceived barriers to care were assessed via a count of six potential barriers (e.g., “it would be too expensive”); participants were categorized as endorsing no barriers, one barrier, or two or more barriers. Participants’ beliefs about psychotherapy (e.g., “counseling or therapy can help people overcome stress in their lives”) were assessed via the sum of the 5-item Beliefs About Psychotherapy Scale (Wagner et al., 2005), which has demonstrated strong internal consistency (Cronbach’s α = .82; Bystritsky et al., 2005). The survey added an additional statement, “I would be able to talk about what bothers me in therapy,” to the measure as avoidance is a core symptom of PTSD. The perceived benefit of treatment for PTSD or other emotional problems was assessed via the mean of seven items, which addressed how much participants expected to experience improvements in different life domains following treatment (e.g., “overall quality of my life,” “Ability to rely less on others”) on a 4-item Likert Scale from “definitely would not improve” to “definitely would improve”. Self-efficacy to follow treatment recommendations was assessed by a response to the statement, “it is (or would be) easy for me to follow doctors’ recommendations to treat PTSD symptoms or similar emotional problems” on a 4-item Likert Scale from “strongly agree” to “strongly disagree”. Social encouragement to seek mental health treatment was assessed by the question, “in the past year, have people in your life encouraged you to get treatment for PTSD or other emotional problems?” Responses were coded as no one; family members only; friends or other veterans only; family and friends or other veterans; or other social encouragement.
Analysis
We used hierarchical logistic regression analyses to model the odds of initiating psychotherapy and the odds of completing a sufficient course of psychotherapy (eight sessions or more), respectively. To account for potential correlation among participants seen in the same VHA facility, facility of care was included in the models as a random effect. The remaining predictors were added to the model sequentially in the three conceptual blocks discussed above: demographics and symptomatology variables, environmental variables, and social cognitive variables. To explore variations in gender differences in the initiation and completion of individual psychotherapy, interactions between gender and the other predictors were examined for each block of variables. Significant interactions (p-value < .05) were retained in the model prior to considering the addition of any subsequent block of variables. We estimated the mediation of the gender differences (including any interactions with gender) stemming from the inclusion of a given block of variables and used bootstrap methods to construct confidence intervals for this mediation of the effects. To identify specific mediating variables within a block of variables leading to a significant change in the estimated gender differences, we subsequently examined a) the change in effects stemming from the addition of individual predictors in the block to the regression model, b) the association between an individual predictor and the outcome, and c) using a generalized linear mixed model, the association between gender and the individual predictor.
Results
Participant characteristics by gender are presented in Table 1. Overall, the largest proportion of participants were White (43.4%; n = 3,133), followed by African American (20.0%; n = 1,444). Participants were middle-aged (M years = 51.5, SD = 16.2). Female veterans (n = 1,135) comprised 15.7% of the sample. Forty-two percent of the sample (n = 3,029; 40.1% of men, 52.3% of women) initiated individual psychotherapy within six months of their index PTSD diagnosis. Of those who initiated, 12.1% (n = 367, 10.8% of men, 17.7% of women; 5.1% of the total sample) completed a sufficient course of individual psychotherapy. The hierarchical logistic regression models for initiation of individual psychotherapy and a sufficient course of individual psychotherapy are summarized in Tables 2 and 3, respectively.
Table 1.
Participant Characteristics by Gender (N=7218)
| Variable | Female veterans (N=1135) | Male veterans (N=6083) |
|---|---|---|
| Race % (n) | ||
| Latino, Hispanic | 10.5 (119) | 24.8 (1510) |
| Native American, Alaskan | 5.4 (61) | 9.8 (597) |
| African-American | 29.4 (334) | 18.3 (1110) |
| Native Hawaiian, Asian-American, | 1.6 (18) | 5.5 (336) |
| Pacific Islander | ||
| White | 53.1 (603) | 41.6 (2530) |
| Age M (SD) | 38.8 (12.4) | 53.8 (15.8) |
| PCL Total Symptom Score M (SD) | 56.8 (15.8) | 57.8 (14.8) |
| SF-12 Mental Component Score M (SD) | 32.1 (12.4) | 33.5 (11.6) |
| SF-12 Physical Component Score M (SD) | 56.8 (15.8) | 57.8 (14.8) |
| Perceived Need for Help % (n) | ||
| No | 15.7 (174) | 20.6 (1219) |
| Yes | 84.4 (938) | 79.4 (4704) |
| Perceived Barriers % (n) | ||
| No Perceived Barriers | 37.8 (420) | 46.7 (2732) |
| One Perceived Barrier | 29.7 (330) | 26.6 (1553) |
| Two Perceived Barriers | 32.4 (360) | 26.7 (1653) |
| Beliefs about Psychotherapy M (SD) | 12.8 (1.9) | 12.1 (1.8) |
| Perceived Benefit of Therapy M (SD) | 3.1 (0.6) | 2.9 (0.6) |
| Self-Efficacy to Follow Treatment | ||
| Recommendations % (n) | ||
| Strongly Agree | 16.2 (176) | 13.5 (781) |
| Agree | 62.3 (679) | 63.2 (3661) |
| Disagree | 19.3 (210) | 20.9 (1208) |
| Strongly Disagree | 2.3 (25) | 2.4 (141) |
| Social Encouragement % (n) | ||
| No one | 20.8 (233) | 21.4 (1280) |
| Family, Friends, & Other Vets | 35.1 (394) | 37.5 (2239) |
| Family Only | 25.4 (285) | 26.8 (1598) |
| Friends & Other Vets Only | 11.8 (133) | 10.3 (615) |
| Provider or Employer | 10.0 (78) | 4.0 (238) |
Table 2.
Hierarchical Logistic Regression Predicting Initiation of Individual Psychotherapy (N=6,335)
| Predictor | OR | 95% CI |
|---|---|---|
| Block 1: Demographics & Symptomology | ||
| Female (Male reference) | 1.37* | [1.06, 1.76] |
| Age | 0.87 | [0.79, 0.95] |
| Age x Female | 1.26* | [1.03, 1.53] |
| Hispanic, Latino (White reference) | 0.91 | [0.78, 1.07] |
| Native American/Alaskan | 1.00 | [0.81, 1.23] |
| African-American | 0.86 | [0.73, 1.01] |
| Native Hawaiian/Pacific Islander/Asian-American | 0.71* | [0.53, 0.95] |
| Trauma Type (Other type of trauma - reference) | ||
| Combat and Unwanted Sexual Experience | 1.16 | [0.87, 1.55] |
| Combat | 0.94 | [0.80, 1.10] |
| Unwanted Sexual Experience | 1.25 | [0.94, 1.64] |
| PTSD Symptoms (PCL) | 1.15** | [1.06, 1.26] |
| SF-12 Mental Component Score | 0.86** | [0.79, 0.93] |
| SF-12 Physical Component Score | 1.02 | [0.96, 1.09] |
| Block 2: Environmental Variables | ||
| OEF/OIF Status | 1.45** | [1.20, 1.75] |
| Drive Time | 0.93* | [0.87, 0.99] |
| Clinic Location of Diagnosis (PC reference) | ||
| PTSD Clinic | 3.43** | [2.81, 4.18] |
| Mental Health Clinic | 2.68** | [2.37, 3.02] |
| Block 3: Social Cognitive Variables | ||
| Perceived Need for Help | 2.18** | [1.80, 2.62] |
| Perceived Barriers (2+ barriers reference) | ||
| One Perceived Barrier | 1.13 | [0.97, 1.31] |
| No Perceived Barriers | 1.36** | [1.18, 1.56] |
| Beliefs about Psychotherapy | 1.10** | [1.03, 1.16] |
| Perceived Benefit of Therapy | 1.07* | [1.00, 1.14] |
| Self-Efficacy to Follow Treatment Recommendations | ||
| Strongly Agree (Strongly Disagree Reference) | 1.07 | [0.70, 1.63] |
| Agree | 1.33 | [0.89, 1.98] |
| Disagree | 1.23 | [0.82, 1.85] |
| Social Encouragement (No Encouragement reference) | ||
| Family, Friends & Other Vets | 1.51** | [1.26, 1.81] |
| Family Only | 1.30** | [1.09, 1.56] |
| Friends & Other Vets Only | 1.50** | [1.20, 1.87] |
| Provider or Employer | 1.13 | [0.82, 1.53] |
Note.
p < .05
p < .01.
Table 3.
Hierarchical Logistic Regression Predicting Completion (>=8 sessions) of Individual Psychotherapy for PTSD (N=2,706)
| Predictor | OR | 95% CI |
|---|---|---|
| Block 1: Demographics & Symptomology | ||
| Female (Male reference) | 1.41 | [0.96, 2.09] |
| Age | 0.89 | [0.73, 1.09] |
| Hispanic, Latino (White reference) | 0.83 | [0.59, 1.17] |
| Native American/Alaskan | 1.16 | [0.75, 1.83] |
| African-American | 0.74 | [0.52, 1.05] |
| Native Hawaiian/Pacific Islander/Asian-American Trauma Type (Other type of trauma - reference) | 1.19 | [0.65, 2.20] |
| Combat and Unwanted Sexual Experience | 0.97 | [0.52, 1.83] |
| Combat | 1.27 | [0.86, 1.88] |
| Unwanted Sexual Experience | 1.63 | [0.95, 2.82] |
| PTSD Symptoms (PCL) | 1.00 | [0.82, 1.22] |
| SF-12 Mental Component Score | 0.83* | [0.69, 0.99] |
| SF-12 Physical Component Score | 1.14 | [0.99, 1.30] |
| Block 2: Environment Variables | ||
| OEF/OIF Status | 0.95 | [0.65, 1.40] |
| Drive Time | 0.90 | [0.77, 1.06] |
| Clinic Location of Diagnosis (PC reference) | ||
| PTSD Clinic | 1.67** | [1.15, 2.41] |
| Mental Health Clinic | 1.53** | [1.17, 2.00] |
| Block 3: Social Cognitive Variables | ||
| Perceived Need for Help | 1.60 | [0.91, 2.81] |
| One Perceived Barrier (2+ perceived barriers reference) | 1.11 | [0.78, 1.59] |
| No Perceived Barriers | 1.90** | [1.39, 2.60] |
| Beliefs about Psychotherapy | 1.13 | [1.00, 1.29] |
| Female x Beliefs about Psychotherapy | 1.55* | [1.04, 2.30] |
| Perceived Benefit of Therapy | 1.01 | [0.88, 1.17] |
| Self-Efficacy to Follow Treatment Recommendations | ||
| Strongly Agree (Strongly Disagree Reference) | 0.70 | [0.26, 1.84] |
| Agree | 0.77 | [0.31, 1.91] |
| Disagree | 0.79 | [0.31, 2.01] |
| Social Encouragement (No Encouragement reference) | ||
| Family, Friends & Other Vets | 1.07 | [0.71, 1.60] |
| Family Only | 0.93 | [0.61, 1.43] |
| Friends & Other Vets Only | 1.10 | [0.66, 1.82] |
| Provider or Employer | 0.97 | [0.47, 1.99] |
Note.
p < .05
p < .01.
Initiation of psychotherapy
There was a significant amount of variation in rates of treatment initiation across facilities of care (p < .0001 in all fitted models). Among demographic and symptomatology predictors, a higher level of PTSD symptoms was associated with increased odds of initiating psychotherapy, and lower mental health functioning (via the SF-12 mental component score) was associated with increased odds of initiating psychotherapy. With respect to environmental variables, OEF/OIF status was associated with increased odds of initiating psychotherapy, as was receiving a diagnosis of PTSD in a mental health clinic (PTSD clinic or general mental health clinic). A longer drive time to a VHA facility was associated with decreased odds of initiating individual psychotherapy. Several social cognitive variables predicted increased odds of initiating psychotherapy, including a perceived need for help, not perceiving any barriers to treatment, having more positive beliefs about psychotherapy, perceiving a benefit of therapy, and social encouragement from friends, family, and other veterans.
A significant difference in odds of initiation of individual psychotherapy was found between men and women, with women generally more likely than men to initiate individual psychotherapy. However, this varied with age as we found an interaction between gender and age, such that younger men were more likely to initiate psychotherapy than older men; age was not significantly associated with initiation among women (Figure 1). Further, older women were more likely to initiate therapy than older men but there was no difference among younger men and women in adjusted rates of therapy initiation (Figure 1). For context, the model estimates an odds ratio for gender of 1.72 (95% CI 1.15, 2.57, p=.01) for veterans 67 years of age (approximately 1 SD above the sample mean), an odds ratio for gender of 1.37 (95% CI 1.06, 1.76, p=.06) for veterans 51 years of age (approximately the sample mean), and an odds ratio of 1.09 (95% CI 0.89, 1.33, p=.41) for veterans 35 years of age (approximately 1 SD below the sample mean), as indicated in Figure 1 and Table 2. While the first two of these estimated associations are substantively smaller than those for the clinic type where the index diagnosis was made and for the veteran perception of needing help, they are in line with the associations found for social encouragement and OEF/OIF status. As age increases, as seen in Figure 1, the difference in the estimated rates of treatment initiation becomes quite substantive. Compared to a model incorporating just the demographic measures, the interaction between age and gender was significantly attenuated in the final model incorporating all three blocks of predictors. Table 4 presents results from analyses examining mediation of the interaction between age and gender. For each gender, the table presents the estimated log odds ratio (or slope) for age from this base demographic model as well as from the models adding symptomatology, environmental, and social cognitive blocks of predictors, respectively. The log odds ratios for age change significantly with the addition of the environmental measures: the slope for female veterans increases but remains non-significant while the slope for male veterans is significantly reduced, indicating that much of the association between age and receipt of therapy among male veterans may stem from differences in access between younger and older male veterans. The environmental measures are predictive of receiving individual psychotherapy but subsequent analyses indicated this change in the age by gender interaction is attributable almost wholly to the inclusion of OEF/OIF status in the model.
Figure 1. Interaction between gender and age for initiation of individual psychotherapy.
Note. The figure plots model estimated probabilities of receiving psychotherapy calculated at the sample mean value for continuous predictors, white race, combat trauma, index diagnosis in a PTSD clinic, no reported barriers to treatment, agreement that veteran can follow treatment recommendations, and encouragement from spouse or family to seek treatment.
Table 4.
Stepwise Examination of Gender by Age Interaction for Receipt of Individual Psychotherapy (N=2,706)
| Fitted Model | β Age | SE β | Parameter Change Δβ | SE Δβ | Δβ 95% Confidence Interval |
|---|---|---|---|---|---|
| Simple Model (Race, Age, Gender, Age by Gender Interaction) | |||||
| Male | −0.34 | 0.03 | |||
| Female | −0.09 | 0.09 | |||
| Model adding Symptomatology | |||||
| Male | −0.31 | 0.04 | 0.03 | 0.01 | 0.01, 0.06 |
| Female | −0.05 | 0.10 | 0.04 | 0.03 | −0.01, 0.09 |
| Model adding Environmental Measures | |||||
| Male | −0.15 | 0.05 | 0.19 | 0.03 | 0.11, 0.27 |
| Female | 0.08 | 0.10 | 0.17 | 0.04 | 0.09, 0.25 |
| Model adding Social Cognitive Measures | |||||
| Male | −0.15 | 0.05 | 0.19 | 0.04 | 0.11, 0.27 |
| Female | 0.10 | 0.10 | 0.19 | 0.05 | 0.11, 0.26 |
Completion of eight psychotherapy visits
There was also a significant amount of variation in rates of completion of individual psychotherapy across facilities of care (p< .0003 in all fitted models). Within the block of demographic and symptomatology predictors, having fewer general mental health symptoms was associated with lower odds of completing psychotherapy. Only one environment variable, receiving a diagnosis of PTSD in a mental health clinic (PTSD clinic or general mental health clinic), was significantly associated with increased odds of completing a sufficient course of therapy. Two social cognitive variables, not perceiving any barriers to treatment and having more positive beliefs about psychotherapy, were generally associated with increased odds of completing individual psychotherapy. However, this latter association may vary with gender, as there may be an interaction between gender and beliefs about psychotherapy (p=.06 in final model, p<.05 in model with all interactions between gender and social cognitive measures) such that men were less likely to complete individual psychotherapy when they held more negative beliefs about psychotherapy; beliefs about psychotherapy did not significantly impact female veterans’ likelihood of completing a sufficient course of psychotherapy (Figure 2). The model indicated no difference in receiving a sufficient course of therapy among men and women when both had positive beliefs in therapy, but there were significantly higher rates of sufficient therapy completion for women than men among veterans with more negative therapy beliefs (Figure 2). For context, the model estimates an odds ratio for gender of 2.01 (95% CI 1.18, 3.40, p=.01) at a belief score of 10 (approximately 1 SD below the sample mean) and an odds ratio for gender of 1.55 (95% CI 1.04, 2.30, p=.03) at a belief score of 12 (near the sample mean for the belief scores). These are in line with the magnitudes of the other significant associations and as illustrated in Figure 3, the model estimates substantial differences in the rates of treatment completion among both male and female veterans with negative therapy beliefs.
Figure 2. Interaction between gender and beliefs about psychotherapy for completion of individual psychotherapy.
Note. The figure plots model estimated probabilities of receiving at least 8 psychotherapy sessions calculated at the sample mean value for continuous predictors, white race, combat trauma, index diagnosis in a PTSD clinic, no reported barriers to treatment, agreement that veteran can follow treatment recommendations, and encouragement from spouse or family to seek treatment.
Discussion
The current study examined gender differences in rates of individual psychotherapy initiation and completion among veterans recently diagnosed with PTSD. For both male and female veterans, we found modest rates of individual psychotherapy initiation within six months of PTSD diagnosis, and much lower rates of individual psychotherapy completion. Forty-two percent of the veterans in the study initiated psychotherapy within six months of receiving a PTSD diagnosis, and a smaller subset of those individuals (12.1%) completed a sufficient course of psychotherapy. Despite being higher than prior estimates (e.g., Mott, Hundt, Sansgiry, Mignogna, & Cully, 2014), these rates still suggest that the majority of veterans were not receiving individual therapy at the time these data were collected. Thus, there remains a need to increase veteran engagement in individual psychotherapy, particularly after they initiate treatment, to ensure all veterans with PTSD receive optimal treatment.
Although these broader trends were present irrespective of gender, female and male veterans’ rates of initiation and completion differed in several important ways. Female veterans had higher observed rates of both initiation (52.3% v. 40.1%) and completion (17.7%v. 10.7%) of individual psychotherapy compared to male veterans. Logistic regression analyses accounting for demographics, symptomatology, environmental, and social cognitive variables also indicated that female veterans were generally more likely to initiate and complete individual psychotherapy than male veterans. This is consistent with Doran and colleagues’ (2017) finding that female veterans were more likely to engage in psychotherapy than male veterans. It is notable that female veterans are accessing needed mental health treatment in the VHA at comparable or greater rates than their male counterparts, despite VHA’s history as a healthcare system designed for men. Female veterans’ access to treatment may reflect efforts to improve women’s mental healthcare within VHA (Oishi et al., 2011). A more in-depth understanding of the active elements that lead to female veterans’ higher rates of initiation and completion of individual psychotherapy should be pursued in order to continue improving mental health care for this population.
We also found an interaction between gender and age in initiation of psychotherapy, such that younger men were more likely to initiate psychotherapy than older men; age did not impact initiation for women. Further, younger men and women initiated therapy at comparable rates while older men were less likely to initiate therapy than older women. Importantly, environment variables, specifically OEF/OIF status, partially accounted for the lower odds of psychotherapy initiation among older men. Thus, one potential explanation for this age-related disparity is that the VHA and the Department of Defense have made substantial efforts to increase OEF/OIF veterans’ access and engagement in care, (Burnam, Meredith, Tanielian, & Jaycox, 2009). For example, VHA implemented an initiative to provide cost-free healthcare for five years after separation for conditions potentially related to combat service after November 11th 1998 to increase access to VHA care (Burnam et al., 2009). Such large-scale initiatives may influence treatment initiation directly by reducing barriers to care and conferring priority access, and indirectly by reducing mental health treatment-related stigma, which has been shown to disproportionately deter men from seeking treatment (Clement et al., 2015). Without these specialized supports, older male veterans may be in need of outreach strategies to increase their rates of psychotherapy initiation.
Additionally, beliefs about treatment were found to potentially moderate the relationship between gender and completion of psychotherapy, such that male veterans were less likely to complete individual psychotherapy when they held negative beliefs about psychotherapy; beliefs about psychotherapy did not impact female veterans’ likelihood of completing a sufficient course of psychotherapy. There is evidence that men have more stigmatizing beliefs about mental health treatment and treatment seeking than women more generally (Fox et al., 2015). Male veterans may be more vulnerable to traditionally masculine messages of self-reliance (Vogt, 2011), which in turn, may make them less likely to fully engage in psychotherapy. A broader, systematic challenge of traditional general role messaging, especially within the military and VHA communities, may be an important component of changing male veterans’ beliefs about psychotherapy. Creating outreach campaigns to target misperceptions about individual psychotherapy for PTSD featuring male veterans who have benefitted from treatment may help to address concerns about psychotherapy. For example, the National Institute of Mental Health ran a “Real Men. Real Depression” campaign between 2003 and 2005 that resulted in 14 million website hits and close to 5,000 email and phone calls to their information hotline (NIMH, 2018) and the VHA’s National Center for PTSD has developed the “AboutFace” campaign which includes videos of veterans describing their experiences with PTSD treatment how it helped them get their lives back (Hamblen et al., 2018). Additionally, role induction techniques, which educate patients on the therapy rationale, course of treatment, and patient and therapist roles in therapy, as well as address misperceptions of psychotherapy, have been shown to improve attendance rates and reduce dropout among prepared patients (Ogrodniczuk, Joyce, & Piper, 2005). Research focused on improving psychotherapy retention among veterans with PTSD is growing (Kehle-Forbes & Kimmerling, 2017), with initial results suggesting that motivational enhancement interventions and shared decision making in treatment planning may improve retention in psychotherapy (Seal et al., 2012; Murphy, Thompson, Murray, Rainey, & Uddo, 2009; Mott, Stanley, Street, Grady, & Teng, 2014). More work is needed to understand how best to address negative beliefs about psychotherapy, especially among male veterans with PTSD, in order to improve psychotherapy retention.
There were several limitations to this study, including the limited follow-up time frame (six months after index PTSD diagnosis). It is possible veterans sought PTSD treatment after the six-month assessment period or were continuing treatment at the end of this follow-up period; thus, actual rates of initiation and completion of individual psychotherapy may be higher over a longer period (e.g., two years) after initial diagnosis, especially as clinics may have waiting periods for beginning individual therapy. Nevertheless, this study captures veterans within the first six months after their PTSD diagnosis, an important time to engage veterans in mental health treatment, in order to reduce sequelae related to untreated PTSD symptomatology. Additionally, this study did not assess predictors of group therapy engagement or PTSD-specific individual psychotherapy, which have become more widely available since the current study’s data collection period. While the most efficacious PTSD treatments are delivered in an individual psychotherapy format, there are many individual psychotherapies that are not evidence-based and do not represent best practice. It is unknown what the rates of initiation and completion are for evidence-based individual PTSD psychotherapies in this population. We also did not examine rates of pharmacotherapy for PTSD; it is possible patients who did not initiate psychotherapy received an adequate dose of medication. In addition, the results could be affected by response biases. The respondents and non-respondents did not differ with respect to psychotherapy outcomes but the non-respondents were generally an older population with corresponding demographics and patterns of VHA use.
This study also had several strengths, including the use of longitudinal assessment of veterans’ engagement in mental health services. This study also used enrichment sampling for women, which allowed for meaningful comparisons between male and female veterans with PTSD. Additionally, both administrative and survey data were utilized, which allowed for the assessment of important social cognitive variables that provide needed context around veterans’ treatment utilization data.
In conclusion, our study revealed that while older male veterans were less likely to initiate treatment than older female veterans, rates of initiation and especially completion of individual psychotherapy for both genders remain relatively low. Creating interventions to increase veteran engagement in psychotherapy and shift negative beliefs about psychotherapy, particularly for male veterans, is an important next step in increasing engagement in individual psychotherapy for PTSD.
Acknowledgments
Authors Note: This material is based upon work supported by [VA HSR&D grant# IAC-06-266]. This material is the result of work supported with resources and the use of facilities at the Minneapolis VA Healthcare System. The views expressed in this article are those of the authors and do not necessarily reflect the position or policy of the Department of Veterans Affairs or the United States government.
Clinical Impact Statement
This study found that female veterans with PTSD are more likely than male veterans to initiate individual psychotherapy, although rates of initiation and completion of individual psychotherapy remain relatively low for both genders. Younger male veterans were more likely to initiate psychotherapy than older male veterans and male veterans with PTSD who held negative beliefs about psychotherapy were less likely to complete a sufficient course of psychotherapy than those who didn’t. Thus, interventions are needed to increase engagement in individual psychotherapy, particularly for male veterans with PTSD.
References
- Bandura A (1977). Social learning theory. Englewood Cliffs, NJ: Prentice Hall. [Google Scholar]
- Bandura A (2004). Health promotion by social cognitive means. Health Education and Behavior, 31, 143–164. [DOI] [PubMed] [Google Scholar]
- Bisson JI, Roberts NP, Andrew M, Cooper R, & Lewis C (2013). Psychological therapies for chronic posttraumatic stress disorder (PTSD) in adults. The Cochrane Library. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Burnam MA, Meredith LS, Tanielian T, & Jaycox LH (2009). Mental health care for Iraq and Afghanistan war veterans. Health Affairs, 28, 771–782. [DOI] [PubMed] [Google Scholar]
- Bystritsky A, Wagner AW, Russo JE, Stein MB, Sherbourne CD, Craske MG, & Roy-Byrne PP (2005). Assessment of beliefs about psychotropic medication and psychotherapy: development of a measure for patients with anxiety disorders. General Hospital Psychiatry, 27, 313–318. [DOI] [PubMed] [Google Scholar]
- Clement S, Schauman O, Graham T, Maggioni F, Evans-Lacko S, Bezborodovs N, … & Thornicroft G. (2015). What is the impact of mental health-related stigma on help-seeking? A systematic review of quantitative and qualitative studies. Psychological Medicine, 45, 11–27. [DOI] [PubMed] [Google Scholar]
- Cully JA, Tolpin L, Henderson L, Jimenez D, Kunik ME, & Petersen LA (2008). Psychotherapy in the Veterans Health Administration: Missed opportunities? Psychological Services, 5, 320–331. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Department of Veterans Affairs and Department of Defense (VA/DoD). (2017). Clinical practice guidelines for the management of posttraumatic stress disorder and acute stress disorder. https://www.healthquality.va.gov/guidelines/MH/ptsd/VADoDPTSDCPGFinal082917.pdf
- Doran JM, Pietrzak RH, Hoff R, & Harpaz-Rotem I (2017). Psychotherapy utilization and retention in a national sample of veterans with PTSD. Journal of Clinical Psychology, 73, 1259–1279. [DOI] [PubMed] [Google Scholar]
- Fox AB, Meyer EC, & Vogt D (2015). Attitudes about the VA health-care setting, mental illness, and mental health treatment and their relationship with VA mental health service use among female and male OEF/OIF veterans. Psychological Services, 12, 49–58. [DOI] [PubMed] [Google Scholar]
- Fulton JJ, Calhoun PS, Wagner HR, Schry AR, Hair LP, Feeling N, … & Beckham JC (2015). The prevalence of posttraumatic stress disorder in Operation Enduring Freedom/Operation Iraqi Freedom (OEF/OIF) veterans: A meta-analysis. Journal of Anxiety Disorders, 31, 98–107. [DOI] [PubMed] [Google Scholar]
- Gallegos AM, Wolff KB, Streltzov NA, Adams LB, Carpenter-Song E, Nicholson J, & Stecker T (2015). Gender differences in service utilization among OEF/OIF veterans with posttraumatic stress disorder after a brief cognitive–behavioral intervention to increase treatment engagement: A mixed methods study. Women’s Health Issues, 25, 542–547. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Hamblen JL, Grubaugh AL, Davidson TM, Borkman AL, Bunnell BE, & Ruggiero KJ (2018). An online peer educational campaign to reduce stigma and improve help seeking in veterans with posttraumatic stress disorder. Telemedicine and e-Health, epub ahead of print. [DOI] [PubMed] [Google Scholar]
- Kang HK, Natelson BH, Mahan CM, Lee KY, & Murphy FM (2003). Posttraumatic stress disorder and chronic fatigue syndrome-like illness among Gulf War veterans: A population-based survey of 30,000 veterans. American Journal of Epidemiology, 157, 141–148. [DOI] [PubMed] [Google Scholar]
- Kehle-Forbes S, & Kimmerling R (2017). Patient Engagement in PTSD Treatment. National Center for PTSD Research Quarterly, 28(3). Retrieved from: https://www.ptsd.va.gov/professional/newsletters/research-quarterly/V28N3.pdf [Google Scholar]
- Kessler RC, Berglund P, Demler O, Jin R, Merikangas KR, & Walters EE (2005). Lifetime prevalence and age-of-onset distributions of DSM-IV disorders in the National Comorbidity Survey Replication. Archives of General Psychiatry, 62(6), 593–602. [DOI] [PubMed] [Google Scholar]
- Kimerling R, Pavao J, Greene L, Karpenko J, Rodriguez A, Saweikis M, & Washington DL (2015). Access to mental health care among women veterans: Is VA meeting women’s needs? Medical Care, 53, 97–104. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Lee DJ, Schnitzlein CW, Wolf JP, Vythilingam M, Rasmusson AM, & Hoge CW (2016). Psychotherapy versus pharmacotherapy for posttraumatic stress disorder: Systemic review and meta-analyses to determine first-line treatments. Depression and Anxiety, 33(9), 792–806. [DOI] [PubMed] [Google Scholar]
- Marmar CR, Schlenger W, Henn-Haase C, Qian M, Purchia E, Li M, … & Karstoft KI. (2015). Course of posttraumatic stress disorder 40 years after the Vietnam War: Findings from the National Vietnam Veterans Longitudinal Study. JAMA Psychiatry, 72(9), 875–881. [DOI] [PubMed] [Google Scholar]
- Maguen S, Madden E, Cohen BE, Bertenthal D, & Seal KH (2012). Time to treatment among veterans of conflicts in Iraq and Afghanistan with psychiatric diagnoses. Psychiatric Services, 63, 1206–1212. [DOI] [PubMed] [Google Scholar]
- Mott JM, Hundt NE, Sansgiry S, Mignogna J, & Cully JA (2014). Changes in psychotherapy utilization among veterans with depression, anxiety, and PTSD. Psychiatric Services, 65, 106–112. [DOI] [PubMed] [Google Scholar]
- Mott JM, Stanley MA, Street RL, Grady RH, & Teng EJ (2014). Increasing engagement in evidence-based PTSD treatment through shared decision-making: A pilot study. Military Medicine, 179, 143–149. [DOI] [PubMed] [Google Scholar]
- Murphy RT, Thompson KE, Murray M, Rainey Q, & Uddo MM (2009). Effect of a motivation enhancement intervention on veterans’ engagement in PTSD treatment. Psychological Services, 6, 264–278. [Google Scholar]
- National Institute of Mental Health. “Real Men. Real Depression.” Campaign. Retrieved from: https://www.nimh.nih.gov/health/topics/men-and-mental-health/men-and-depression/nimhs-real-men-real-depression-campaign.shtml#part_152889
- Nepon J, Belik SL, Bolton J, & Sareen J (2010). The relationship between anxiety disorders and suicide attempts: Findings from the National Epidemiologic Survey on Alcohol and Related Conditions. Depression and Anxiety, 27, 791–798. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Ogrodniczuk JS, Joyce AS, & Piper WE (2005). Strategies for reducing patient-initiated premature termination of psychotherapy. Harvard Review of Psychiatry, 13, 57–70. [DOI] [PubMed] [Google Scholar]
- Oishi SM, Rose DE, Washington DL, MacGregor C, Bean-Mayberry B, & Yano EM (2011). National variations in VA mental health care for women veterans. Women’s Health Issues, 21, 130–137. [DOI] [PubMed] [Google Scholar]
- Orcutt HK, Erickson DJ, & Wolfe J (2004). The course of PTSD symptoms among gulf war veterans: A growth mixture modeling approach. Journal of Traumatic Stress, 17, 195–202. [DOI] [PubMed] [Google Scholar]
- Pietrzak RH, Johnson DC, Goldstein MB, Malley JC, & Southwick SM (2009). Perceived stigma and barriers to mental health care utilization among OEF-OIF veterans. Psychiatric Services, 60, 1118–1122. [DOI] [PubMed] [Google Scholar]
- Seal KH, Abadjian L, McCamish N, Shi Y, Tarasovsky G, & Weingardt K (2012). A randomized controlled trial of telephone motivational interviewing to enhance mental health treatment engagement in Iraq and Afghanistan veterans. General Hospital Psychiatry, 34, 450–459. [DOI] [PubMed] [Google Scholar]
- Spitzer C, Barnow S, Völzke H, John U, Freyberger HJ, & Grabe HJ (2009). Trauma, posttraumatic stress disorder, and physical illness: Findings from the general population. Psychosomatic Medicine, 71, 1012–1017. [DOI] [PubMed] [Google Scholar]
- Spoont MR, Nelson DB, Murdoch M, Rector T, Sayer NA, Nugent S, & Westermeyer J (2014). Impact of treatment beliefs and social network encouragement on initiation of care by VA service users with PTSD. Psychiatric Services, 65, 654–662. [DOI] [PubMed] [Google Scholar]
- Spoont MR, Nelson DB, Murdoch M, Sayer NA, Nugent S, Rector T, & Westermeyer J (2015). Are there racial/ethnic disparities in VA PTSD treatment retention? Depression and Anxiety, 32, 415–425. [DOI] [PubMed] [Google Scholar]
- Stecker T, Fortney JC, & Sherbourne CD (2011). An intervention to increase mental health treatment engagement among OIF veterans: A pilot trial. Military Medicine, 176, 613–619. [DOI] [PubMed] [Google Scholar]
- Vogt D, Fox AB, & Di Leone BA (2014). Mental health beliefs and their relationship with treatment seeking among US OEF/OIF veterans. Journal of Traumatic Stress, 27, 307–313. [DOI] [PubMed] [Google Scholar]
- Vogt D (2011). Mental health-related beliefs as a barrier to service use for military personnel and veterans: a review. Psychiatric Services, 62, 135–142. [DOI] [PubMed] [Google Scholar]
- Wagner AW, Bystritsky A, Russo JE, Craske MG, Sherbourne CD, Stein MB, & Roy-Byrne PP (2005). Beliefs about psychotropic medication and psychotherapy among primary care patients with anxiety disorders. Depression and Anxiety, 21, 99–105. [DOI] [PubMed] [Google Scholar]
- Ware JE, Kosinski M, & Keller SD (1996). A 12-Item Short-Form Health Survey: Construction of scales and preliminary tests of reliability and validity. Medical Care, 34, 220–233. [DOI] [PubMed] [Google Scholar]
- Weathers FW, Litz BT, Herman DS, Huska JA, & Keane TM (1993, October). The PTSD Checklist (PCL): Reliability, validity, and diagnostic utility. In the Annual Convention of the International Society for Traumatic Stress Studies, San Antonio, TX: (Vol. 462). [Google Scholar]