Predictors of Mental Health Care Use Among Male ... - APA PsycNET

3 downloads 686 Views 78KB Size Report
Mar 12, 2012 - Deployed in Support of the Wars in Afghanistan and Iraq. Brooke A. L. Di Leone. VA Boston Healthcare System, Boston, Massachusetts.
Psychological Services 2013, Vol. 10, No. 2, 145–151

In the public domain DOI: 10.1037/a0032088

This document is copyrighted by the American Psychological Association or one of its allied publishers. This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

Predictors of Mental Health Care Use Among Male and Female Veterans Deployed in Support of the Wars in Afghanistan and Iraq Brooke A. L. Di Leone

Dawne Vogt

VA Boston Healthcare System, Boston, Massachusetts

VA Boston Healthcare System, Boston, Massachusetts and Boston University School of Medicine

Jaimie L. Gradus

Amy E. Street

VA Boston Healthcare System, Boston, Massachusetts, Boston University School of Medicine, and Boston University School of Public Health

VA Boston Healthcare System, Boston, Massachusetts and Boston University School of Medicine

Hannah L. Giasson

Patricia A. Resick

VA Boston Healthcare System, Boston, Massachusetts

VA Boston Healthcare System, Boston, Massachusetts and Boston University School of Medicine

What factors predict whether Operation Enduring Freedom and Operation Iraqi Freedom (OEF/OIF) veterans who need mental health care receive that care? The present research examined factors associated with a need for care, sociodemographic characteristics, deployment experiences, and perceptions of care as gender-specific predictors of overall mental health care use and Veterans Affairs (VA) mental health care use for male and female OEF/OIF veterans (N ⫽ 1,040). Only veterans with a probable need for mental health care, as determined by scores on self-report measures of mental health symptomatology, were included in the sample. Overall, predictors of service use were similar for women and men. A notable exception was the finding that lower income predicted use of both overall and VA mental health care for women, but not men. In addition, sexual harassment was a unique predictor of VA service use for women, whereas non-White race was predictive of VA service use for men only. Knowledge regarding the factors that are associated with use of mental health care (broadly and at VA) is critical to ensuring that veterans who need mental health care receive it. Keywords: veteran, mental health, gender, military, service use

Military personnel deployed in service of Operation Enduring Freedom in Afghanistan and Operation Iraqi Freedom in Iraq (OEF/OIF) face many potential challenges (Schell & Tanielian, 2011). For example, OEF/OIF veterans tend to have experienced multiple deployments that involved considerable combat exposure (Tanielian & Jaycox, 2008) and, thus, are at high risk for experiencing mental health problems following deployment (Hoge, Auchterlonie, & Milliken, 2006; Hoge et al., 2004). Fortunately,

the mental health conditions that this cohort is likely to experience are treatable, particularly if individuals receive adequate care at an early stage (Kehle et al., 2010; Tanielian & Jaycox, 2008). To date, there has been little research examining gender-specific barriers to and facilitators of mental health treatment seeking among this cohort. Such research is essential to inform efforts to ensure that these veterans have every opportunity to obtain needed mental health care.

Brooke A. L. Di Leone, National Center for PTSD, VA Boston Healthcare System, Boston, Massachusetts and Center for Organization, Leadership, and Management Research, VA Boston Healthcare System, Boston, Massachusetts; Dawne Vogt, National Center for PTSD, VA Boston Healthcare System, Boston, Massachusetts and Department of Psychiatry, Boston University School of Medicine; Jaimie L. Gradus, National Center for PTSD, VA Boston Healthcare System, Boston, Massachusetts, Department of Psychiatry, Boston University School of Medicine, and Department of Epidemiology, Boston University School of Public Health; Amy E. Street, National Center for PTSD, VA Boston Healthcare System, Boston, Massachusetts and Department of Psychiatry, Boston University School of Medicine; Hannah L. Giasson, National Center for PTSD, VA Boston Healthcare System, Boston, Massachusetts; Patricia A. Resick, National Center for PTSD, VA Boston

Healthcare System, Boston, Massachusetts and Department of Psychiatry, Boston University School of Medicine. This research was supported by the National Center for PTSD, Office of Mental Health Services, Department of Veterans Affairs. 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. The funding agency was not involved in the design and conduct of the study; collection, management, analysis, and interpretation of the data; or the preparation, review, or approval of the manuscript. Correspondence concerning this article should be addressed to Brooke A. L. Di Leone, Center for Organization, Leadership, and Management Research (152M), VA Boston HCS, 150 South Huntington Avenue, Boston, MA 02130. E-mail: [email protected] 145

This document is copyrighted by the American Psychological Association or one of its allied publishers. This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

146

DI LEONE ET AL.

A consistent finding from the broader literature on barriers to and facilitators of care is that veterans with a higher burden of mental health problems, and therefore greater need for mental health care, are more likely to seek treatment. This is true whether mental health problems are operationalized based on mental health diagnoses (Fikretoglu, Guay, Pedlar, & Brunet, 2008; Hoff & Rosenheck, 1998; Hoge et al., 2006; Katz et al., 1997; Kehle et al., 2010; Romeis, Gillespie, & Thorman, 1988) or self-identified need for treatment (Hoge, Auchterlonie, & Milliken, 2006; Hoge et al., 2004; Karlin, Duffy, & Gleaves, 2008; Katz et al., 1997; Ouimette, Wolfe, Daley, & Gima, 2003; Romeis, Gillespie, & Thorman, 1988). However, the finding that a sizable proportion of individuals who would benefit from mental health services do not receive treatment (e.g., Kehle et al., 2010; Schell & Tanielian, 2011) suggests that need for mental health care is not sufficient to explain service use. Consistent with this perspective, sociodemographic factors such as race and income are also associated with veterans’ use of care (e.g., Romeis, Gillespie, & Thorman, 1988). In addition, deployment experiences, such as sexual harassment and combat exposure (e.g., Hoge et al., 2006; Ouimette, Wolfe, Daley, & Gima, 2003) and opinions of Department of Veterans Affairs (VA) care (e.g., Vogt et al., 2006; Washington, Yano, Simon, & Sun, 2006) have demonstrated associations with use of VA care. Importantly, predictors of use of any mental health care and use of VA mental health care may differ. VA care is an option for many recent veterans, particularly given current VA policy providing for five years of free care for OEF/OIF/OND combat veterans for any condition related to service in Iraq or Afghanistan. However, access to mental health care outside the VA setting is likely to depend on many factors, including those related to cost of care and availability of health insurance coverage. Beyond demographics, factors such as deployment experiences and perceptions of quality of care may uniquely influence where a Veteran decides to seek care. For example, findings indicate that veterans may be more likely to turn to VA mental health care when seeking care for conditions associated with stressful military experiences (Rosenheck & Massari, 1993). Historically, a much larger proportion of patients at VA health care facilities have been men, which may lead women to be less comfortable in those settings and to have different perceptions of and experiences with care (e.g., Vogt et al., 2001; Vogt, Barry, & King, 2008). Given this difference, it is also important to take gender1 into account in any examination of veterans’ overall use of mental health care and use of VA mental health care. Indeed, research that has examined gender differences has found that male and female veterans exhibit differences in priorities related to satisfaction with and predictors of health care seeking (e.g., Kressin et al., 1999; Chatterjee et al., 2009). The vast majority of OEF/OIF veterans’ health care use research, however, examines men and women together, obscuring any gender differences that might exist. The extent to which findings on predictors of service use are relevant for both men and women in this cohort is a question that remains largely unanswered. Prior research has not explored the relative contribution of need for care, sociodemographic characteristics, deployment experiences, and perceptions of care as gender-specific predictors of use of mental health care among OEF/OIF veterans. The current study addresses this gap in the literature in three main ways: (a) by examining predictors of mental health care use above and beyond

factors related to probable need for care in a cohort of OEF/OIF veterans, (b) by separately examining predictors of overall mental health care use and VA mental health care use, and (c) by separately examining predictors of mental health care use for women and men. On the basis of prior research suggesting that sociodemographic characteristics play a role in receipt of care among those who need it, we expected that a number of different sociodemographic factors (i.e., minority race and higher income) would predict use of any mental health care beyond probable need for care. In addition, we expected that sociodemographic factors (i.e., minority race and lower income), deployment factors (i.e., greater warfare exposure and sexual harassment during deployment), and attitudinal factors (i.e., more positive perceptions of VA care) would additionally contribute to VA mental health care use. Because of the lack of prior literature on gender-specific predictors of use of mental health care in this cohort, we considered analyses of unique factors related to mental health care use for women and men to be exploratory.

Method Sample The study population consisted of female and male veterans who had completed at least one OEF or OIF deployment. Potential participants were drawn from the VA Environmental Epidemiology Service’s roster of all OEF/OIF veterans who have separated from military service. Out of 6,000 potential participants, a total of 2,348 participants returned a completed survey. After accounting for ineligible responders and correcting for estimated ineligibility among nonresponders, the response rate was 48.6%, which is comparable or larger than other large-scale OEF/OIF Veteran surveys (Smith, Smith, Gray, & Ryan, 2007; Tanielian & Jaycox, 2008). We compared survey responders to nonresponders on demographic and military characteristics drawn from administrative records data to explore nonresponse bias. Overall, findings revealed few differences, and those that were observed were generally small and unlikely to influence the specific associations under examination in this article. Specifically, differences between responders and nonresponders were small with regard to gender (␾ ⫽ ⫺0.021), race (Cramer’s V ⫽ 0.069), military rank (Cramer’s V ⫽ 0.146), military branch (Cramer’s V ⫽ 0.055), and duty status (Cramer’s V ⫽ 0.093). The age difference observed between responders and nonresponders was slightly larger but still fell below the commonly accepted criterion for a medium effect (Cohen’s d ⫽ ⫺0.445), such that responders were approximately 4 years older, on average, than nonresponders, t ⫽ ⫺16.66, p ⬍ .05. Given that the present research was concerned with factors that predict use of mental health care among those with a need for mental health care, we selected a subsample of participants based on their probable need for mental health care (in this case, defined by self-reported symptomatology). Specifically, participants were selected if they reported a level of symptomatology consistent with 1 We use the term gender, meaning the behaviors, attitudes, and personality traits associated with men and women within a particular context, rather than sex, which is often used to refer to biological processes.

PREDICTORS OF MENTAL HEALTH CARE USE

probable posttraumatic stress disorder (PTSD), anxiety, depression, or alcohol use, the most commonly reported mental health conditions among veterans (Tanielian & Jaycox, 2008). A total of 1,040 participants (44% of the sample) met at least one of the above criteria and were included in analyses. Women and men were approximately equally represented in the sample (535 women, 51% of sample included in analyses). Table 1 displays the sample’s self-reported sociodemographic and military service characteristics.

This document is copyrighted by the American Psychological Association or one of its allied publishers. This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

Procedure A multiple mailing strategy recommended by Dillman, Smyth, and Christian (2009) was used to encourage survey completion: an introductory letter alerting participants that they would be receiving a survey; 1 week later a survey packet including the survey, a fact sheet including the elements of informed consent, and a $5 cash incentive; 1 week later a thank you–reminder postcard; 1 week later an additional survey packet; 3 weeks later a final survey packet sent by priority mail. Returned packets for which the post office provided forwarding information were resent to the indicated address.

Table 1 Sociodemographic and Military Service Characteristics by Gender Variable Age M (SD) Racea White African American Other Income 25,000 or less 25,001–50,000 50,001–75,000 Over 75,001 Branch of military servicea Marines Army Navy Air Force Coast Guard Multiple branches Military service component Reserves/Guard Active duty Both Reserves/Guard and active duty Used any mental health carea Used VA mental health care Perceptions of VA care Warfare exposurea Sexual harassmenta PTSD symptomatology Anxiety symptomatology Depression symptomatologya Alcohol misusea

Females (n ⫽ 535)

Males (n ⫽ 505)

34 (8.9)

35 (9.5)

69.1% 22.5% 8.3%

80.6% 10.9% 8.5%

28.7% 34.0% 18.3% 19.1%

23.7% 34.9% 19.9% 21.5%

5.5% 63.2% 13.5% 17.1% 0.4% 0.4%

18.1% 58.8% 11.5% 11.1% 0.0% 0.4%

42.9% 56.1% 0.9% 55.8% 35.0% 11.8 (4.6) 46.8 (16.4) 10.3 (4.4) 45.5 (19.1) 9.2 (9.2) 14.1 (6.7) 1.1 (1.3)

45.0% 52.6% 2.4% 48.2% 33.9% 12.2 (4.3) 59.4 (22.6) 7.3 (1.2) 46.0 (18.3) 8.2 (8.1) 11.8 (6.5) 1.5 (1.4)

Note. VA ⫽ Veterans Affairs. a Denotes a significant difference between genders, p ⬍ .05.

147

Measures Sociodemographics. The sociodemographic items included questions about participants’ gender, race, ethnicity, age, annual income, and education. Race was coded dichotomously based on self-identification as either White or non-White. Income was a categorical variable, with the possible options being $25,000 or less, $25,001–$50,000, $50,001–$75,000, or over $75,001. Positive perceptions of VA health care. The attitudinal measure of perceptions of VA health care was adapted from an existing measure of health care perceptions (Vogt et al., 2006) and assessed individuals’ opinions of VHA care and facilities. Specifically, five items were included to assess perceptions of availability of services, privacy of medical records, staff skill, staff courtesy, and ease of seeking VA care. Participants responded to each item on a 5-point Likert-type scale that ranged from 1 (extremely negative) to 5 (extremely positive). Responses to each item were summed to create a total score. Coefficient alpha was .86 in the current sample. Use of overall and VA mental health care. Use of overall and VA mental health care was assessed via two questions. The first asked whether participants had seen a professional for help with an emotional problem since returning from their most recent deployment: “Have you seen a professional for help with an emotional problem? These professionals could be psychologists, therapists, counselors or psychiatrists and could include groups led by professional counselors and visits for medication for emotional problems.” The second question, building on the first, asked whether participants had used a VA facility for any of that care: “Did you go to a Department of veterans Affairs (VA) health care facility for any of this care?” Participants provided a “yes” or “no” response to each item, and each item was treated as a dichotomous variable. Warfare exposure. Consistent with several other recent studies (Vogt & Tanner, 2007; Vogt et al., 2011), warfare exposure was measured by summing the 15-item Combat Experiences and 15-item Aftermath of Battle scales from the Deployment Risk and Resilience Inventory (DRRI; King, King, Vogt, Knight, & Samper, 2006). As in other studies (e.g., Vogt, Proctor, King, King, & Vasterling, 2008), this measure was modified to include a 4-point Likert-type response format, ranging from 1 (never) to 4 (many times). The Combat Experiences scale assesses exposure to stereotypic warfare experiences such as firing a weapon, being fired on, and witnessing injury and death. The Aftermath of Battle scale assesses exposure to the consequences of combat, including experiences such as observing human remains and dealing with detainees. Items were rated on a scale, ranging from 1 (never) to 4 (many times), with higher scores indicating more warfare exposure. Responses to each item were summed to create a total score. Sexual harassment during deployment. Experiences of sexual harassment during deployment were measured using a 7-item scale taken from the DRRI. Items ask participants to rate how frequently they experienced particular sexual harassment events (including sexual assault) while deployed (e.g., made crude remarks or threatened retaliation for not engaging in sexual behavior). These items were rated on a scale, ranging from 1 (never) to 4 (many times), with higher scores indicating

148

DI LEONE ET AL.

more experiences with sexual harassment while deployed. Responses to each item were summed to create a total score.

This document is copyrighted by the American Psychological Association or one of its allied publishers. This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

Need for Mental Health Care The following measures were used in two ways. First, we used cutoffs indicative of probable diagnoses as a means of selecting our sample of individuals with a probable need for mental health care. Second, we included continuous scale scores in analyses to account for symptom severity. PTSD Checklist, Military Version (PCL-M; Weathers, Litz, Herman, Huska, & Keane, 1993). The PCL-M was used to assess posttraumatic stress symptomatology related to stressful deployment experiences. The 17 items are directly adapted from the Diagnostic and Statistical Manual of Mental Disorders (4th ed.; DSM–IV; American Psychiatric Association, 1994) to assess reexperiencing, avoidance and emotional numbing, and hyperarousal symptoms. Respondents are asked to rate how much they have been bothered by each symptom in the past month, ranging from 1 (not at all) to 5 (extremely). Responses to each item were summed to create a total score. On the basis of commonly used criteria (Tanielian & Jaycox, 2008), those who had a minimum score of 50 (440 participants, 20% of total sample) or those who met the symptom endorsement diagnostic criteria for PTSD (505 participants, 23%) were included in the sample. Coefficient alpha was .96 in the current sample. Anxiety Subscale of Depression, Anxiety, Stress Scale (DASS; Lovibond & Lovibond, 1995). This subscale of the DASS contains 14 items specifically related to anxiety, including autonomic arousal, skeletal musculature effects, situational anxiety, and subjective experience of anxious affect. Participants are asked to report on their symptomatology in the past week from 0 (did not apply to me at all) to 3 (applied to me very much, or most of the time). Responses to each item were summed to create a total score. Those who had a minimum score of 10, indicating probable clinically significant anxiety (Lovibond & Lovibond, 1995), were included in the sample. A total of 379 participants (17%) met this criterion. Coefficient alpha was .92 in the current sample. Center for Epidemiologic Studies Depression Scale. The Center for Epidemiologic Studies Depression Scale (CES-D; Radloff, 1977) is a widely used 10-item questionnaire designed to measure the frequency of depressive symptoms in the past week from 0 (none of the time or less than one day) to 3 (5–7 days). Responses to each item were summed to create a total score. On the basis of commonly used criteria (Andresen, Malmgren, Carter, & Patrick, 1994), those who had a minimum score of 10 were included in the sample (731 participants, 33%). Coefficient alpha was .86 in the current sample. CAGE. The CAGE (Ewing, 1984) is a 4-item questionnaire that assesses the presence of probable clinically significant alcohol use. In this study, items were assessed for the period since one’s most recent deployment. A correlation of 0.89 has been shown between CAGE scores and diagnoses when using the dichotomous, two-item cutoff method for scoring (Bradley, Kivlahan, Bush, McDonell, & Fihn, 2001). On the basis of commonly used criteria (Buchsbaum, Buchanan, Centor, Schnoll, & Lawton, 1991), those who had a minimum score of 2 (460 participants, 20%) were

included in the sample. Coefficient alpha was .68 in the current sample.

Data Analysis Prior to examining predictors of use of any mental health care and VA mental health care, we first calculated descriptive statistics for the sample. We then calculated Spearman bivariate correlation analyses separately for women and men to examine associations between potential predictors and use of care (scored dichotomously). Following bivariate analyses, we conducted a total of four hierarchical multivariate logistic regression analyses to identify unique predictors of use of any mental health care and VA care separately for women and men. All variables that were significant at the bivariate level were included as predictors in the regression analyses. Analyses revealed no problems with multicollinearity among variables included in multivariate analyses. Specifically, the SWEEP algorithm in SPSS retained all variables in the model using a fairly conservative epsilon value (.00001). In addition, all tolerances were greater than .10, and all VIFs were below 5.

Results Use of Any Mental Health Care As shown in Table 2, bivariate correlations between use of any mental health care and variables associated with probable need for care (i.e., severity of posttraumatic stress symptomatology, anxiety symptomatology, depression symptomatology) were moderately sized and significant for both women and men, as predicted. Contrary to expectations, alcohol abuse was not significantly associated with use of any mental health care for women or men. In terms of deployment variables, warfare exposure was significantly associated with use of any mental health care for both women and men, whereas sexual harassment was associated with use among women only. Contrary to predictions, income was significantly negatively associated with use of any mental health care for women and men. Minority racial status was significantly associated with use of any mental health care for men but not women. Table 2 Bivariate Correlation Analyses Use of any care Variable PTSD (PCL score) Anxiety (DASS score) Depression (CES-D score) Alcohol abuse (CAGE score) Race Income Warfare exposure Sexual harassment Perceptions of VA care

Women ⴱ

.38 .30ⴱ .37ⴱ ⫺.02 .02 ⫺.17ⴱ .18ⴱ .23ⴱ —a

Use of VA care

Men ⴱ

.49 .37ⴱ .30ⴱ ⫺.08 .13ⴱ ⫺.15ⴱ .30ⴱ .00 —a

Women ⴱ

.45 .32ⴱ .31ⴱ .00 .08 ⫺.20ⴱ .29ⴱ .18ⴱ .22ⴱ

Men .46ⴱ .36ⴱ .27ⴱ ⫺.04 .15ⴱ ⫺.21ⴱ .31ⴱ ⫺.01 .15ⴱ

Note. VA ⫽ Veterans Affairs; PTSD ⫽ posttraumatic stress disorder; PCL ⫽ PTSD Checklist, Military Version; DASS ⫽ Depression, Anxiety, Stress Scale; CES-D ⫽ Center for Epidemiologic Studies Depression Scale. a Variable not included in analyses. ⴱ p ⬍ .05.

This document is copyrighted by the American Psychological Association or one of its allied publishers. This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

PREDICTORS OF MENTAL HEALTH CARE USE

The results of the logistic regression analyses indicate that posttraumatic stress symptomatology (aOR ⫽ 1.03) and depression symptomatology (aOR ⫽ 1.07) predicted women’s use of any mental health care in the expected direction (see Table 3). The only factor that predicted women’s use of any mental health care above and beyond probable need for care was low income (lowest income category compared with the highest aOR ⫽ 1.92); however, this association was in the opposite direction from what was expected based on previous research. For men, contrary to predictions, posttraumatic stress symptomatology (aOR ⫽ 1.06) was the only variable that predicted use of any mental health care in multivariate analyses. Hosmer-Lemeshow statistics for the multivariate models for both women and men suggest that the models fit the data: ␹2 ⫽ 2.55, p ⫽ .96 and ␹2 ⫽ 7.74, p ⫽ .46, respectively.

Use of VA Mental Health Care Bivariate correlations between predictors and use of VA mental health care are shown in Table 2. As predicted, the variables associated with probable need for care (i.e., posttraumatic stress symptomatology, anxiety symptomatology, and depression symptomatology) were each significantly associated with use of VA mental health care for both women and men. Similar to results for use of any care, alcohol abuse was not significantly associated with use of VA mental health care for women or men. In contrast, income, warfare exposure, and perceptions of VA care were significantly associated with use of VA mental health care for both women and men, while sexual harassment was associated with use of VA mental health care among women only. Contrary to expectations, minority racial status was significantly associated with men’s use of VA mental health care but not women’s. Logistic regression analyses (see Table 4) revealed that, as expected, posttraumatic stress symptomatology significantly predicted use of VA mental health care for women (aOR ⫽ 1.05). However, contrary to predictions, other variables associated with probable need for care did not significantly predict women’s use of VA mental health care in multivariate analyses. Low income

149

Table 4 Separate Multivariate Regression Analyses by Gender Predicting Use of Veterans Affairs Mental Health Care, Including All Variables Significant at the Bivariate Level Women Variable PTSD Anxiety Depression Minority race Lowest income categoryb Warfare exposure Sexual harassment Perceptions of VA care ⫺2LL Cox & Snell R2 Nagelkerke R2

aOR ⴱ

1.05 0.99 1.03 —a 2.71ⴱ 1.01 1.07ⴱ 1.17ⴱ

Men

(95% CI)

aOR

(95% CI)

1.03, 1.07 0.95, 1.03 0.99, 1.08 —a 1.25, 5.88 1.00, 1.03 1.01, 1.13 1.10, 1.24 437.10 .29 .40



1.03, 1.08 0.96, 1.05 0.96, 1.06 1.10, 3.56 0.93, 4.55 1.00, 1.02 —a 1.04, 1.16 419.31 .25 .34

1.05 1.01 1.01 1.98ⴱ 2.06 1.01 —a 1.10ⴱ

Note. OR ⫽ odds ratio; CI ⫽ confidence interval; PTSD ⫽ posttraumatic stress disorder. a Variable not included in analyses because relationship was not significant in bivariate analyses. b Comparison group is highest income category. ⴱ p ⬍ .05.

(lowest income category compared with the highest aOR ⫽ 2.71), experiences of sexual harassment during deployment (aOR ⫽ 1.07), and more positive perceptions of VA care (aOR ⫽ 1.17) predicted women’s use of VA mental health care above and beyond probable need for care. For men as well, posttraumatic stress symptomatology (aOR ⫽ 1.05) was the only significant need-related predictor of use of VA mental health care. As expected, minority racial status (aOR ⫽ 1.98) and perceptions of VA care (aOR ⫽ 1.10) predicted men’s use of VA mental health care above and beyond probable need for care. Hosmer-Lemeshow statistics for the multivariate models for both women and men suggest that the models fit the data: ␹2 ⫽ 2.22, p ⫽ .97 and ␹2 ⫽ 1.10, p ⫽ .99, respectively.

Discussion Table 3 Separate Multivariate Regression Analyses by Gender Predicting Use of Any Mental Health Care, Including All Variables Significant at the Bivariate Level Women Variable PTSD Anxiety Depression Minority race Lowest income categoryb Warfare exposure Sexual harassment ⫺2LL Cox & Snell R2 Nagelkerke R2

aOR ⴱ

1.03 1.01 1.07ⴱ —a 1.92ⴱ 1.00 1.05

Men

(95% CI)

aOR

(95% CI)

1.01, 1.04 0.97, 1.04 1.03, 1.12 —a 1.05, 3.53 0.99, 1.02 0.99, 1.10 555.93 .19 .26



1.03, 1.08 0.95, 1.03 0.98, 1.08 0.94, 2.97 0.62, 2.49 1.00, 1.02 —a 487.34 .23 .31

1.06 0.99 1.03 1.67 1.24 1.01 —a

Note. OR ⫽ odds ratio; CI ⫽ confidence interval; PTSD ⫽ posttraumatic stress disorder. Variable not included in analyses because relationship was not significant in bivariate analyses. b Comparison group is highest income category. ⴱ p ⬍ .05. a

This study examined factors associated with overall use of mental health care and VA mental health care in a national sample of male and female OEF/OIF veterans with probable need for mental health care. In terms of use of any mental health care, predictors in multivariate analyses were similar for women and men, with a single exception: Lower income predicted use of mental health care for women but not men. For men, we did not identify any factors beyond probable need for care that predicted use of mental health care generally. Analyses for use of VA mental health care again revealed similar predictors for men and women. Not surprisingly, more PTSD symptomatology and positive perceptions of VA care were associated with use of VA mental health care, regardless of gender. However, low income and sexual harassment were predictive for women only, whereas minority race was predictive for men only. For men, it may be that the demographic and military variables included in the multivariate logistical regression model did not adequately capture reasons behind decisions to use mental health care generally. For example, factors not measured here, such as stigma associated with seeking mental health care and beliefs

This document is copyrighted by the American Psychological Association or one of its allied publishers. This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

150

DI LEONE ET AL.

about the usefulness of mental health care, could be significant predictors of use beyond the variables we examined (Hoge et al., 2004; Vogt et al., 2011). Relevantly, factors associated with masculine self-identity may be associated with less use of care if men (and particularly men affiliated with the military) perceive care seeking as a threat to their masculinity (Addis & Mahalik, 2003). Minority racial status was modestly associated with use of VA mental health care for men, replicating results from prior studies with other cohorts of veterans (e.g., Diala et al., 2000). Unfortunately, more detailed comparisons of multiple ethnic/racial subgroups were not possible in this sample due to small numbers.2 For women, income was moderately negatively associated with use of any mental health care. Interestingly, although income has been shown to be a predictor of use of mental health care in previous research (e.g., Karlin et al., 2008), it is usually found to have a positive relationship with use outside of VA. That women veterans in this sample who had lower annual income were significantly more likely to use mental health care generally suggests that income may be a proxy for stressors that contribute to a need for mental health care. In terms of predictors of women’s use of VA care, more sexual harassment during deployment was modestly associated with greater use, even after controlling for mental health symptom severity. This result is particularly important, because few previous studies have explored the relationship between deployment sexual harassment and use of mental health care. This is consistent with prior research suggesting that women who have experienced sexual assault while deployed (a component of deployment sexual harassment) are more likely to use VA services (Kelly et al., 2008). (It is also important to note that those who have experienced military sexual trauma have access to free VA care for mental health conditions that are determined to be associated with their experience of trauma.) Low income was also moderately associated with use of VA mental health care for women. Given that low income remained a predictor of women’s use of care when controlling for need-related factors such as PTSD and depression symptomatology, it is possible that the stress associated with low income for women is manifest in other mental health-related factors. It is important that VA consider the unique needs of women veterans with lower income, given that they are particularly likely to seek mental health care at VA. There are limitations of the present research. First, there are a number of factors that may be important in mental health care seeking that we were not able to examine in this study (e.g., insurance coverage, stigma). In terms of examining deployment stressors as predictors of health care use, combat exposure and sexual harassment were assessed only for events that happened during OEF/OIF deployments and do not include exposure during prior conflicts or nonwar zone military service. Accordingly, our estimates of the strength of the associations between deployment stressors and health care use are likely slight underestimates. In addition, stigma related to mental health care seeking may have an impact on use of mental health care, but was not included in this research. Despite our inability to examine these potential predictors, it is important to note that model fit statistics suggested that our multivariate models fit the data in all analyses conducted. Last, given that the data are cross-sectional and it was not possible to randomly assign participants to experimental conditions, these data cannot answer causal questions. Future research that explores

similar factors longitudinally or experimentally would be valuable for answering questions of mechanism and causality. Similarly, although some aspects of need for mental health care were taken into account in this research, others were not considered. For example, measures of bipolar disorder and schizophrenia (which were not included in the survey from which data were drawn) might have a different relationship with use of mental health care than depression and anxiety. However, by including PTSD, depression, anxiety, and alcohol abuse symptomatology, we have accounted for the most commonly reported mental health conditions among this cohort (Tanielian & Jaycox, 2008). In addition, our measure of use of mental health care did not allow us to parse out those individuals who received care at VA alone from those who sought care at VA as well as other facilities. That is, we could identify those who received mental health care at non-VA facilities alone and those who received some care at VA facilities, but not those who received VA care alone or both VA care and care elsewhere. Despite these limitations, this research offers a valuable contribution to the literature. We found a number of novel results and disambiguated other results that have been found in previous research. Our simultaneous examination of multiple categories of predictors of mental health care use for OEF/OIF veterans is unique in its analysis of men and women separately. This methodology allowed us to conclude that patterns found in previous research do not necessarily function in the same way for women and men. In addition, given that a sizable proportion of veterans who would benefit from mental health services do not receive treatment, our focus on identifying predictors of service use among this population ensures that we are considering the needs of a particularly important group. Knowledge regarding the factors that are associated with use of mental health care (broadly and at VA) is critical to ensuring that veterans who need mental health care receive it.

2 For Black Hispanic, Asian, Pacific Islander/Native Hawaiian, and American Indian/Alaska Native categories, all Ns were less than 10. We were, however, able to examine differences within the group of participants who identified as White, comparing White Hispanic participants with White non-Hispanic participants. These analyses revealed no statistically significant differences, suggesting that combining these groups did not obscure important findings. Thus, for the purposes of this study we maintain the White/non-White operationalization.

References Addis, M. E., & Mahalik, J. R. (2003). Men, masculinity, and the contexts of help seeking. American Psychologist, 58, 5–14. doi:10.1037/0003066X.58.1.5 American Psychiatric Association. (1994). Diagnostic and statistical manual of mental disorders (4th ed.). Washington, DC: Author. Andresen, E. M., Malmgren, J. A., Carter, W. B., & Patrick, D. L. (1994). Screening for depression in well older adults: Evaluation of a short form of the CES–D (Center for Epidemiologic Studies Depression scale). American Journal of Preventive Medicine, 10, 77– 84. Bradley, K. A., Kivlahan, D. R., Bush, K. R., McDonell, M. B., & Fihn, S. D. (2001). Variations on the CAGE alcohol screening questionnaire: Strengths and limitations in VA general medical patients. Alcoholism: Clinical and Experimental Research, 25, 1472–1478. doi:10.1111/j .1530-0277.2001.tb02149.x

This document is copyrighted by the American Psychological Association or one of its allied publishers. This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

PREDICTORS OF MENTAL HEALTH CARE USE Buchsbaum, D. G., Buchanan, R. G., Centor, R. M., Schnoll, S. H., & Lawton, M. J. (1991). Screening for alcohol abuse using CAGE scores and likelihood ratios. Annals of Internal Medicine, 115, 774 –777. Chatterjee, S., Rath, M. E., Spiro, A., III, Eisen, S., Sloan, K. L., & Rosen, A. K. (2009). Gender differences in veterans Health Administration mental health service use: Effects of age and psychiatric diagnosis. Women’s Health Issues, 19, 176 –184. doi:10.1016/j.whi.2009.03.002 Diala, C., Muntaner, C., Walrath, C., Nickerson, K. J., LaVeist, T. A., & Leaf, P. J. (2000). Racial differences in attitudes toward professional mental health care and in the use of services. American Journal of Orthopsychiatry, 70, 455– 464. doi:10.1037/h0087736 Dillman, D. A., Smyth, J. D., & Christian, L. M. (2009). Internet, mail, and mixed-mode surveys: The tailored design method (3rd ed.). New York, NY: Wiley. Ewing, J. A. (1984). Detecting alcoholism: The CAGE questionnaire. JAMA: Journal of the American Medical Association, 252, 1905–1907. doi:10.1001/jama.1984.03350140051025 Fikretoglu, D., Guay, S., Pedlar, D., & Brunet, A. (2008). Twelve month use of mental health services in a nationally representative, active military sample. Medical Care, 46, 217–223. doi:10.1097/MLR .0b013e31815b979a Hoff, R. A., & Rosenheck, R. A. (1998). Female veterans’ use of Department of veterans Affairs health care services. Medical Care, 36, 1114 – 1119. doi:10.1097/00005650-199807000-00017 Hoge, C. W., Auchterlonie, J. L., & Milliken, C. S. (2006). Mental health problems, use of mental health services, and attrition from military service after returning from deployment to Iraq or Afghanistan. JAMA: Journal of the American Medical Association, 295, 1023–1032. doi: 10.1001/jama.295.9.1023 Hoge, C. W., Castro, C. A., Messer, S. C., McGurk, D., Cotting, D. I., & Koffman, R. L. (2004). Combat duty in Iraq and Afghanistan, mental health problems, and barriers to care. The New England Journal of Medicine, 351, 13–22. doi:10.1056/NEJMoa040603 Karlin, B. E., Duffy, M., & Gleaves, D. H. (2008). Patterns and predictors of mental health service use and mental illness among older and younger adults in the United States. Psychological Services, 5, 275–294. doi: 10.1037/1541-1559.5.3.275 Katz, S. J., Kessler, R. C., Frank, R. G., Leaf, P., Lin, E., & Edlund, M. (1997). The use of outpatient mental health services in the United States and Ontario: The impact of mental morbidity and perceived need for care. American Journal of Public Health, 87, 1136 –1143. doi:10.2105/ AJPH.87.7.1136 Kehle, S. M., Polusny, M. A., Murdoch, M., Erbes, C. R., Arbisi, P. A., Thuras, P., & Meis, L. A. (2010). Early mental health treatment-seeking among U.S. National Guard soldiers deployed to Iraq. Journal of Traumatic Stress, 23, 33– 40. Kelly, M. M., Vogt, D. S., Scheiderer, E. M., Ouimette, P., Daley, J., & Wolfe, J. (2008). Effects of military trauma exposure on women veterans’ use and perceptions of Veterans Health Administration care. Journal of General Internal Medicine, 23, 741–747. doi:10.1007/s11606008-0589-x King, L., King, D., Vogt, D., Knight, J. A., & Samper, R. (2006). Deployment Risk and Resilience Inventory: A collection of measures for studying deployment-related experiences of military personnel and veterans. Military Psychology, 18, 89 –120. doi:10.1207/s15327876mp1802_1 Kressin, N. R., Skinner, K., Sullivan, L., Miller, D. R., Frayne, S., Kazis, L., & Tripp, T. (1999). Patient satisfaction with Department of Veterans Affairs health care: Do women differ from men? Military Medicine, 164, 283–288. Lovibond, S. H., & Lovibond, P. F. (1995). Manual for the Depression Anxiety Stress Scales. Sydney, Australia: The Psychology Foundation of Australia.

151

Ouimette, P., Wolfe, J., Daley, J., & Gima, K. (2003). Use of VA health care services by women veterans: Findings from a national sample. Women & Health, 38, 77–91. doi:10.1300/J013v38n02_06 Radloff, L. S. (1977). The CES-D scale: A self-report depression scale for research in the general population. Applied Psychological Measurement, 1, 385– 401. doi:10.1177/014662167700100306 Romeis, J. C., Gillespie, K. N., & Thorman, K. E. (1988). Female veterans’ use of health care services. Medical Care, 26, 589 –595. doi:10.1097/ 00005650-198806000-00007 Rosenheck, R., & Massari, L. (1993). Wartime military service and utilization of VA health care services. Military Medicine, 158, 223–228. Schell, T. L., & Tanielian, T. (2011). A needs assessment of New York state veterans: Final report to the New York State Health Foundation. RAND Corporation monograph series. Santa Monica, CA: The RAND Corporation. Smith, B., Smith, T. C., Gray, G. C., & Ryan, M. A. (2007). When epidemiology meets the Internet: Web-based surveys in the Millennium Cohort Study. American Journal of Epidemiology, 166, 1345–1354. doi:10.1093/aje/kwm212 Tanielian, T., & Jaycox, L. H. (Eds.). (2008). Invisible wounds of war: Psychological and cognitive injuries, their consequences, and services to assist recovery. RAND Corporation monograph series. Santa Monica, CA: The RAND Corporation. Vogt, D. S., Barry, A. A., & King, L. A. (2008). Toward gender-aware health care: Evaluation of an intervention to enhance care for female patients in the VA setting. Journal of Health Psychology, 13, 624 – 638. Vogt, D., Bergeron, A., Salgado, D., Daley, J., Ouimette, P., & Wolfe, J. (2006). Barriers to Veterans Health Administration care in a nationally representative sample of women veterans. Journal of General Internal Medicine, 21, S19 –S25. doi:10.1111/j.1525-1497.2006.00370.x Vogt, D. S., Proctor, S. P., King, D. W., King, L. A., & Vasterling, J. J. (2008). Validation of scales from the Deployment Risk and Resilience Inventory in a sample of Operation Iraqi Freedom veterans. Assessment, 15, 391– 403. doi:10.1177/1073191108316030 Vogt, D., Smith, B., Elwy, R., Martin, J., Schultz, M., Drainoni, M. L., & Eisen, S. (2011). Predeployment, deployment, and postdeployment risk factors for posttraumatic stress symptomatology in female and male OEF/OIF veterans. Journal of Abnormal Psychology, 120, 819 – 831. doi:10.1037/a0024457 Vogt, D. S., Stone, E. R., Salgado, D. M., King, L. A., King, D. W., & Savarese, V. W. (2001). Gender awareness among Veterans Administration health-care workers: Existing strengths and areas for improvement. Women & Health, 34, 65– 83. doi:10.1300/J013v34n04_05 Vogt, D. S., & Tanner, L. R. (2007). Risk and resilience factors for posttraumatic stress symptomatology in Gulf War I veterans. Journal of Traumatic Stress, 20, 27–38. doi:10.1002/jts.20187 Washington, D. L., Yano, E. M., Simon, B., & Sun, S. (2006). To use or not to use: What influences why women veterans choose VA health care. Journal of General Internal Medicine, 21, S11–S18. doi:10.1111/j.15251497.2006.00369.x Weathers, F. W., Litz, B. T., Herman, D. S., Huska, J. A., & Keane, T. M. (1993). The PTSD checklist (PCL): Reliability, validity, and diagnostic utility. Paper presented at the International Society of Traumatic Stress Studies, San Antonio, Texas. http://www.pdhealth.mil/library/ downloads/PCL_sychometrics.doc

Received March 12, 2012 Revision received October 16, 2012 Accepted November 19, 2012 䡲