Skip to main content
NCBI home page
VA Author Manuscripts logoLink to VA Author Manuscripts
. Author manuscript; available in PMC: 2024 Jan 4.
Published in final edited form as: Psychol Serv. 2022 Jul 4;20(Suppl 2):213–221. doi: 10.1037/ser0000683

The Veterans Health Administration’s Integrated Model of Care Increases Accessibility and Delivery of Mental Health Services

Samuel C Peter 1,*, Tate F Halverson 1,2, Shannon M Blakey 1,2, Mary Jo Pugh 3,4, Jean C Beckham 1,2,5, Patrick S Calhoun 1,2,5,6, Nathan A Kimbrel 1,2,5,6
PMCID: PMC9810760  NIHMSID: NIHMS1804886  PMID: 35787032

Abstract

Depression and posttraumatic stress disorder (PTSD) are two of the most common mental health conditions experienced by veterans. It is unclear what individual and system level factors are associated with receiving mental health treatment for these concerns. Using a national sample of Gulf War Era veterans who endorsed lifetime diagnoses of either depression or PTSD (N = 425), regression analyses were used to predict past-year treatment utilization. Predictor variables were those indicated in the behavioral model of healthcare utilization, including predisposing demographic variables (e.g., age, race), enabling variables (e.g., service connection, enrollment in Veterans Health Administration [VHA])), and need-based variables (e.g., current symptom severity). VHA enrollment was associated with a three- and five-times higher odds of being treated for depression or PTSD, respectively. Income and symptom severity were also positively associated with treatment utilization. Among individuals with diagnoses of depression and/or PTSD, VHA enrollment was the strongest predictor of receiving mental health treatment for these diagnoses, controlling for all other variables in the model including recent contact with the health care system, current symptom severity, and the presence of other enabling resources. Results suggest that the VHA’s integrated model of care increases accessibility and delivery of effective mental health services.

Keywords: PTSD, depression, health care utilization, veterans

Health care utilization represents an adaptive strategy for coping with and/or recovering from various mental health concerns. Across diagnoses, mental health care utilization is associated with symptom reduction and improvements in psychosocial functioning; this applies to mental health care for some of the most common diagnoses among veterans, including depression (Cuijpers et al., 2020), posttraumatic stress disorder (PTSD; Kline et al., 2018), and alcohol use disorder (AUD; Magill et al., 2019; Seal et al., 2009). Understanding which individuals will access mental health care under what circumstances may aid in increasing appropriate delivery of these largely beneficial treatments.

The Veterans Heath Administration (VHA) is the largest integrated health care system in the United States (US) and was established in order to provide medical care, including mental health services, to US military veterans (U. S. Department of Veterans Affairs, 2021). Approximately 60% of veterans are eligible for VHA care, with eligibility being dependent upon length of service, combat experience, injuries accrued or made worse during serve, and income (Farmer et al., 2016). Recent legislation has been passed in response to a growing need for mental health care in the veteran population. For example, the Commander John Scott Hannon Veterans Mental Health Care Improvement Act of 2019 and the Veterans Comprehensive Prevention, Access to Care, and Treatment Act of 2020 aimed to increase access to and quality of the VHA mental health and suicide prevention services (Congressional Research Service [CRS], 2021). Moreover, these legislative acts were specifically designed to address the needs of certain underserved populations, including veterans living in rural areas, women veterans, veterans experiencing homelessness, and veterans at an increased risk for suicide. Additional legislation has also been enacted aimed at increasing availability of community care for veterans who are unable to receive timely or appropriate care through VHA, potentially due to distance from a facility or unavailability of the type of care needed (CRS, 2018). Outpatient VHA services are generally rated as comparable or higher quality than community medical care based on various measures of quality of care, such as those focused on safety and effectiveness of treatments (Farmer et al., 2016). Though most veterans are eligible for VHA care, it is unclear which veterans are most likely to access VHA services. Studies examining health care utilization have been mixed in their findings regarding what factors are predictive of seeking mental health care. One commonly used model for predicting mental health care attendance and/or other forms of help seeking is the behavioral model of health care utilization (BMHU; Andersen, 1995). The BMHU groups explanatory variables into three groups: (1) immutable demographic variables that make someone more or less predisposed to access health care (e.g., age); (2) enabling or hindering variables (e.g., distance from a mental health clinic, having insurance); and (3) need-based variables (e.g., symptom severity).

The BMHU has shown promise in its ability to predict mental health care utilization across multiple settings and populations. Seal and colleagues (2010) applied the BMHU to a sample of United States (US) military veterans who had served during Operation Enduring Freedom and Operation Iraqi Freedom. They found that older age was positively associated with mental health care utilization, whereas living further away from a VHA clinic was negatively associated with utilization. In addition, having PTSD plus other concurrent mental health diagnoses (as opposed to PTSD alone) was also associated with a greater likelihood of health care utilization. These associations are what would be predicted based on the BMHU.

Other studies examining the BMHU have found less intuitive results. For example, Peter and colleagues (2021) applied the BMHU to a sample of female survivors of intimate partner violence who had begun a multi-session mental health evaluation. The objective of this study was to predict which individuals would complete the evaluation versus discontinue prior to receiving a referral to treatment. Unexpectedly, Peter and colleagues (2021) found that depression and PTSD symptom severity were both negatively associated with completing the evaluation. The authors surmised that although the BMHU assumes that greater symptom severity would predict a greater likelihood of seeking treatment, certain mental health symptoms may serve as barriers to continuing care once initially accessed. For example, avoidance of trauma-related memories and stimuli is a core feature of PTSD (Pietrzak et al., 2011), and some people with PTSD may discontinue treatment or treatment-seeking out of a desire to avoid thinking about and discussing their trauma histories. Similarly, behavioral models of depression emphasize how efforts to avoid aversive or minimally rewarding situations and stimuli leads to a pattern of withdrawal and passivity that could, in turn, interfere with health care utilization (Carvalho & Hopko, 2011). PTSD- and depression-related avoidance, paired with typical features such as amotivation/anhedonia, cognitive difficulties, and difficulties with interpersonal functioning, might therefore prevent the people who need care most from seeking or fully engaging with it.

Finally, Fleming and Resick (2017) tested the BMHU among trauma survivors and found mixed associations between symptom severity and mental health care utilization, depending on type of symptom. They found that depressive symptoms were negatively associated with utilization, but that PTSD-related arousal symptoms were positively associated with utilization. Similar to previous studies, they found that age was positively associated with utilization. These studies suggest that the BMHU is a helpful framework for predicting health care utilization, but that the relationship between symptom severity and accessing care is complex in the context of mental health services.

One relevant variable that has not been evaluated in these previous tests of the BMHU is alcohol use. Alcohol use is likely to be important to consider when understanding mental health care utilization for depression or PTSD for multiple reasons. For PTSD, alcohol use is often conceptualized as serving an avoidance function in the form of “self-medication” (Hawn et al., 2020), which may deter treatment seeking to the extent that alcohol use helps people with PTSD to “successfully” cope with their symptoms. Self-medication has also been proposed as an explanation for alcohol use in the context of depression and other mood disorders (Bolton et al., 2009), in that people with AUD might in part drink alcohol to cope with low mood and depressive symptoms (e.g., Daughters et al., 2016). However, at higher levels of alcohol consumption, alcohol use might be associated with greater treatment seeking, which could be due to greater overall mental health symptom severity or greater likelihood of legally mandated substance use treatment. One recent review found that most studies have identified a positive association between substance use and mental health service utilization among trauma survivors, although findings were largely mixed and varied substantially based on population and setting (Hawn et al., 2020).

There are multiple veteran-specific enabling resources that may connect veterans to mental health care that have not been studied in previous tests of the BMHU. For example, VHA enrollment may significantly impact treatment seeking. Individuals appear to be more likely to avail themselves of mental health referrals if they are referred within an integrated health care setting, such as the VHA, as opposed to an outside agency (Bartels et al., 2004). Additionally, the service connection disability system is among one of the most common ways that military veterans report being connected to mental health services (Bovin et al., 2019). “Service connection” refers to a disability program operated by the US Veterans Benefits Administration (U.S. Department of Veterans Affairs, 2019). Disability compensation is a tax-free monetary benefit paid to veterans with disabilities incurred or aggravated during active military service. Veterans that have a service-connected condition may be eligible for not only monetary benefits, but also free or reduced cost health care, access to vocational rehabilitation services, insurance, and travel reimbursement for appointments related to their service-connected conditions, among other potential benefits (U.S. Department of Veterans Affairs, 2019). Evidence regarding the relationship between service connection and treatment utilization is mixed. Sripada and colleagues (2018) examined longitudinal changes in service utilization following a new or increased PTSD-related service connection status and found that the vast majority of veterans neither increased, nor decreased mental health care utilization. Other longitudinal investigations have found that mental health service utilization increases following establishment of PTSD-related service connection, suggesting that the service connection disability program is an effective means of promoting receipt of mental health care for PTSD (Sayer et al., 2004; Spoont et al., 2007). Relatively less research has been published examining what the impact of service connection status is on mental health treatment seeking for depression, although multiple longitudinal studies have found that service connection is a protective factor against dying by suicide (Desai et al., 2005; Zivin et al., 2007). Thus, VHA enrollment and service connection may both be related to treatment seeking among veterans, and consideration of both variables may better elucidate the relationship between these veteran-specific enabling resources and mental health care utilization.

Gulf War Era veterans represent roughly half of all living veterans and are defined as any veterans who served in active duty on or after August 2, 1990 (National Center for Veterans Analysis and Statistics, 2020). Among VHA’s many priorities is connecting veterans to care from which they might benefit (US Department of Veterans Affairs, 2018). The BMHU is a useful framework for not only identifying individual factors associated with greater health care utilization (e.g., symptom severity), but it can also be used to examine whether theoretically enabling resources such as VHA enrollment and service connection are associated with greater rates of treatment utilization. Thus, the present test of the BMHU may aid in identifying underserved populations and if environmental and/or system level factors are associated with treatment utilization.

The Present Study

The present study aimed to test the BMHU among a sample of US Gulf War Era veterans who had a self-reported history of depression or PTSD. Based on previous work in this area (Seal et al., 2010; Peter et al., 2021; Fleming & Resick, 2017), we hypothesized that older age would be associated with a greater odds of receiving mental health care for these diagnoses. We also hypothesized that enabling resources, such as service connection and VHA enrollment, would be associated with a greater likelihood of receiving mental health care for depression and PTSD (Peter et al., 2021; Seal et al., 2010). Given mixed findings in the literature, we considered all other analyses to be exploratory in nature, including those focused on the relationship between health care utilization and mental health symptoms and/or alcohol use.

Method

Participants and Procedures

This sample included Gulf War Era veterans recruited for a larger project, titled “Gulf War Research and Individual Testimony” (Project GRIT; Grant 1I01HX001682), Project GRIT was a national survey of Gulf War Era veterans’ healthcare needs, utilization patterns, and associated costs. Veterans identified for inclusion in Project GRIT were required to have served in active-duty capacity between May 1, 1990, and February 29, 1991. Based on this criteria, 1,098,991 Gulf War Era veterans were identified in a VHA administrative database and stratified random sampling identified a subsample of 6,000 veterans to contact for participation. Sampling was stratified to oversample women veterans for at least 25% representation and to include 750 veterans from each of eight identified United States geographic regions. A modified Dillman approach was utilized to achieve a high response rate which included multiple contact letters (Dillman et al., 2014).

From the initial pool of 6,000, a total of 3,272 surveys were mailed (i.e., veterans who did not initially opt out of participating) and 548 surveys were returned due to incorrect mailing address. Altogether, 1,153 veterans completed and returned the survey and 32 opted out after receiving the survey. The response rate for the study was 42.33% (1,153 completed surveys out of 2,724 received surveys). All participants were considered Gulf War Era veterans, defined as individuals serving in active military, naval, or air service during the Gulf War Era who were discharged or released from service under conditions other than dishonorable. Project GRIT veterans were similar to the national cohort of VHA-utilizing veterans from which they were sampled with women veterans intentionally oversampled (for further detail regarding recruitment methodology, please see Blakey et al., 2021).These participants had served in the U.S. military during the 1991 Gulf War (regardless of whether they had deployed to the Persian Gulf region). While they received VHA care at some point and were all currently eligible for VHA services, they were not necessarily currently utilizing any form of VHA health care at the time of the survey. For example, some veterans, although eligible for VHA services, reported exclusively use of private insurance for healthcare needs. For the present study, only veterans who self-reported that they had ever (i.e., lifetime) been diagnosed with PTSD (n = 281) or depression (n = 362) were included. These two overlapping subsamples comprised a total of 425 unique veterans (Table 1). The sample was geographically representative of the US (11.76% Midwest, 6.82%. Northeast, 53.41% South, 14.35% West, 13.66% did not disclose state of residence).

Table 1.

Sample Characteristics (N = 425)

Variable
Age, M (SD) 57.12 (6.97)
Sex (% Female) 26.01
Race (% Racial Minority) 41.08
Current enrollment in VHA for some or all of health care (% enrolled) 78.34
SC (% connected for any condition) 88.51
Utilize any healthcare (% utilize) 49.41
Household income (%)
 < $15.000 3.53
 $15,000 – $29,999 10.82
 $30,000 – $44,999 14.59
 $45,000 – $59,999 18.82
 $60,000 – $74,999 15.76
 >$75,000 32.94
PC-PTSD-5, M (SD) 2.70 (1.99)
PHQ-2, M (SD) 2.66 (1.93)
AUDIT-C, M (SD) 2.74 (2.82)
Education Level (%)
 Less than High School 0.47
 High School or GED 10.35
 Some College or Trade/Technical/Vocational School 33.18
 Associate’s Degree 17.41
 Bachelor’s Degree 18.12
 Graduate Degree 20.24
Work Status (%)a
 Full-Time 42.59
 Part-Time 7.06
 Unemployed 5.65
 Retired 42.12
 Student 1.88
 Disabled 32.94
Relationship Status (%)
 Married/Committed Relationship 66.82
 Separated/Divorced 24.47
 Widowed 2.12
 Single, Never Married 5.18
Open in a new tab

Note.

a.

Veterans endorsed all that applied,

AUDIT-C = Alcohol Use Disorders Identification Test, 3-item version; PC-PTSD = Primary Care Posttraumatic Stress Disorder Screen for DSM-5; PHQ-2 = Patient Health Questionnaire-2; SC = Service Connection; VHA = Veterans Health Administration

Measures

Mental Health Diagnoses and Treatment Utilization.

Veterans were presented with a list of 28 health diagnoses and asked, “Have you ever been diagnosed with the following?” A follow-up item asked, “If yes, have you been treated for this condition in the last 12 months?” This was used to capture the first two outcome variables of interest, which were whether veterans who reported being diagnosed with PTSD had received treatment for PTSD in the past twelve months, and likewise for depression. In a separate section of the questionnaire packet, an item that was not specific to any diagnosis asked, “How many times have you seen a mental health provider during the past 12 months?” Response options were: (0) none; (1) 1–2 times; (2) 3–4 times; (3) 5–6 times; or (4) more than 6 times.

Predictor Variables.

Variables are grouped below based on their theoretical conceptualization, according to the BMHU, as either being indicators of predisposition/demographics, enabling factors, or need for services. Below, we indicate how each was measured and incorporated into the analyses.

Predisposing Variables.

Age, sex, and race were included as predisposing demographic variables. Age was reported as continuous and sex as dichotomous male (0) or female (1). Given the distribution of the race variable in the sample, which was approximately 59% white, race was entered dichotomously into the present analyses as either white (0) or any other race (1). A fourth variable conceptualized as being an indicator of overall predisposition for health care utilization was measured using an item that asked, “Have you received healthcare services (primary, mental health, specialty care) in the past 12 months?” Participants responded either (0) no; or (1) yes.

Enabling Variables.

VA service connection status, VHA enrollment status, and household income were conceptualized as enabling variables. Service connection was examined dichotomously as either having (1) or not having (0) any service-connected condition as indicated by participant response to a yes/no questions phrased, “Are you service connected?”.1 VHA enrollment status was measured with a single item that asked, “Have you ever been enrolled in VA health care?”. Participants who endorsed using VHA health care for either some or all of their current health care needs were coded as being currently VHA-enrolled (1), while participants who reported never or only previously (but not now) using VHA health care services were coded as not being currently VHA-enrolled (0). Finally, income was measured with a single item that read, “Which of the following categories best describes your annual household income before taxes or any other deductions? Please include income from all sources such as salaries and wages, social security, retirement income, investments, rental incomes and other sources.” Response options were: (0) less than $15,000; (1) $15,00 to $29,999; (2) $30,000 to $44,999; (3) $45,000 to $59,999; (4) $60,000 to $74,999; or (5) $75,000 or more.

Need-Based Variables.

Three continuous variables were included as clinical indicators of need. Current PTSD symptom severity was measured via the five-item Primary Care PTSD Screen for DSM-5 (PC-PTSD-5; Prins et al., 2016). This measure comprises five dichotomous items (yes/no), with an affirmative answer indicating the experience of that symptom within the past month. Respondents report on the experience of nightmares (re-experiencing), cognitive or behavioral avoidance (avoidance), hypervigilance (hyperarousal), numbness or detachment (negative alterations in cognition/mood), and guilt or blame (negative alterations in cognition/mood). Total scores can range from 0–5. The PC-PTSD-5 has been found to correlate strongly with a clinical interview of PTSD among veterans, with a cut-off score of 3 being associated with .95 specificity and .85 sensitivity in detecting a PTSD diagnosis (Prins et al., 2016). Recent additional psychometric evaluations support the diagnostic accuracy of the PC-PTSD-5 and furthermore provide supporting evidence of the measure’s acceptability regarding clarity and comfort completing this measure on their own as opposed to with a provider (Bovin et al., 2021). In the present sample, internal consistency was adequate (α = .79).

Current depression severity was measured via the two-item version of the Patient Health Questionnaire (PHQ-2; Kroenke et al., 2003). This questionnaire begins with a stem question of, “Over the last two weeks, how often have you been bothered by any of the following problems,” followed by two items that ask about the experience of, “feeling down, depressed, or hopeless” (item 1) and, “little interest or pleasure in doing things” (item 2). Response options range from 0 (not at all) to 3 (nearly every day), with total scores ranging from 0–6. A meta-analysis of 21 studies found a pooled sensitivity of .91 and specificity of .70 at a cut-off point of greater than or equal to 2 (Manea et al., 2016). Additional psychometric evaluations have demonstrated that the PHQ-2 is comparable to longer versions (e.g., the PHQ-9) regarding diagnostic accuracy and sensitivity to change during treatment (Staples et al., 2019). Internal consistency in the present sample was excellent (α = .90).

Finally, current alcohol use severity was measured via the three-item version of the Alcohol Use Disorders Identification Test (AUDIT-C; Bush et al., 1998). This measure differs from the full version of the AUDIT in that it only includes the first three items, which ask about the frequency of consuming any amount of alcohol, the number of drinks consumed in a typical drinking day, and the frequency of consuming six or more drinks in one occasion. Total scores range from 0–12, and cut-off score of 4 or more adequately detected current heavy drinking and/or alcohol abuse or dependence (sensitivity of 86%; specificity of 72%). A validation study sampling post-9/11 US veterans across four VA medical centers supported its diagnostic accuracy and found that it was invariant across age and race (Crawford et al., 2013). Internal consistency in the present sample was good (α = .85).

Results

Preliminary Analyses

Analyses were run using R version 4.0.2. Missingness analyses revealed that less than 5% of data were missing and that data were missing at random. Listwise deletion was used for remaining analyses, which is considered appropriate under these conditions per recommendations by Tabachnick and Fidell (2007).

On average, the participants included in the present analyses were 57.12 years of age (SD = 6.97), 26% were female, and 41% identified as a racial minority. Most veterans (68%) reported household incomes totaling over $45,00, with the modal response being greater than $75,000 (33%). Nearly 90% had a service-connected condition, and 78% reported current VHA enrollment. About half (49%) reported that they had utilized any form of health care over the past year. The majority of veterans in our sample reported being diagnosed with both depression and PTSD (n = 218), whereas it was less frequent to be diagnosed with only depression (n = 144) and even less common to be diagnosed with only PTSD (n = 63). The average PHQ-2 score was 2.66 (SD = 1.93), and 74.35% of veterans screened positive for depression based on the recommended cut-off score of greater than or equal to 2. The average PC-PTSD-5 score was 2.70 (SD = 1.99), with 57.04% of veterans screening positive for PTSD based on the recommended cut-off score of greater than or equal to 3. The average AUDIT-C score was 2.74 (SD = 2.82), with 32.78% of veterans screening positive for likely current alcohol abuse or dependence based on the recommended cut-off score of greater than or equal to 4 (Table 1). Regarding treatment utilization, 197 (46%) participants reported that they had been treated for depression in the past year, whereas 153 (36%) reported that they had been treated for PTSD in the past year.

Bivariate correlations of predictor variables are presented in Table 2. Correlations were generally as would be expected and were weak to moderate in magnitude. The AUDIT-C, PHQ-2, and PC-PTSD-5 all correlated positively with one another, with the strongest correlation being between the PHQ-2 and PC-PTSD-5 (r = .48). There was also a notably high correlation between service connection status and whether individuals reported current VHA enrollment (r = .47).

Table 2.

Predictor Variable Bivariate Correlations

1 2 3 4 5 6 7 8 9 10
1. Age -
2. Sex −.01 -
3. Race .08 .08 -
4. VHA enrollment −.09 .04 .13 -
5. SC .12 −.07 .10 .47** -
6. Any healthcare .06 −.07 −.03 −.08 .07 -
7. Household income .08 −.17* −.14* −.24** .16 .12 -
8. PC-PTSD-5 −.22** −.13 .11 .33** .33** −.02 −.10 -
9. PHQ-2 −.26** −.12 .13* .11 .21** .05 −.17** .48** -
10. AUDIT-C −.19** −.25** −.02 −.07 .07 .06 .03 .14** .15** -
Open in a new tab

Note.

*

p<.05,

**

p<.01.

Pearson coefficients are provided for correlations between continuous variables. Spearman coefficients are presented for correlations that included categorical variables. Variables were coded as they were used in the primary analyses (e.g., race was dichotomized as White and non-White). AUDIT-C = Alcohol Use Disorders Identification Test, 3-item version; PC-PTSD = Primary Care Posttraumatic Stress Disorder Screen for DSM-5; PHQ-2 = Patient Health Questionnaire-2; SC = Service Connection; VHA = Veterans Health Administration

Regression Models

Treatment Utilization for Depression (Table 3).

Table 3.

Logistic Regression Results, Depression Model

Predictor Variable Estimate Standard Error Odds Ratio 95% Confidence Interval for Odds Ratio
Age −0.01 0.02 0.99 0.95–1.03
Sex 0.47 0.30 1.60 0.90–2.90
Race 0.05 0.27 1.05 0.62–1.78
VHA Enrollment 1.19 0.37 3.30 1.62–6.99
SC −0.33 0.41 0.72 0.32–1.59
Any Healthcare 0.03 0.26 1.03 0.62–1.71
Household Income 0.22 0.09 1.24 1.04–1.49
PC-PTSD-5 0.06 0.08 1.06 0.91–1.23
PHQ-2 0.11 0.08 1.11 0.95–1.31
AUDIT-C −0.06 0.05 0.94 0.86–1.04
Open in a new tab

Note. AUDIT-C = Alcohol Use Disorders Identification Test, 3-item version; PC-PTSD = Primary Care Posttraumatic Stress Disorder Screen for DSM-5; PHQ-2 = Patient Health Questionnaire-2; SC = Service Connection; VHA = Veterans Health Administration; Confidence intervals that did not include 1.00 were interpreted to be statistically significant.

Of the 425 veterans included in the present study, 362 (85%) reported having ever been diagnosed with depression, and 197 (46%) reported being treated for depression in the past year. Among veterans who reported being diagnosed with depression, VHA enrollees were significantly more likely to report being treated for depression in the past year, OR = 3.30, 95% CI [1.62, 6.99], p = .001. Household income was also positively associated with participants’ likelihood of being treated for depression in the past year, OR = 1.24, 95% CI [1.04, 1.49], p = .017.

Treatment Utilization for PTSD (Table 4).

Table 4.

Logistic Regression Results, PTSD Model

Predictor Variable Estimate Standard Error Odds Ratio 95% Confidence Interval for Odds Ratio
Age −0.02 0.03 0.98 0.93–1.03
Sex 0.30 0.38 1.35 0.65–2.88
Race −0.01 0.30 0.99 0.55–1.81
VHA Enrollment 1.68 0.58 5.36 1.85–18.41
SC 0.73 0.67 2.08 0.59–8.49
Any Healthcare 0.18 0.30 1.20 0.67–2.16
Household Income 0.29 0.11 1.33 1.08–1.65
PC-PTSD-5 0.08 0.09 1.09 0.91–1.30
PHQ-2 0.23 0.09 1.27 1.06–1.52
AUDIT-C −0.09 0.05 0.92 0.82–1.01
Open in a new tab

Note. AUDIT-C = Alcohol Use Disorders Identification Test, 3-item version; PC-PTSD = Primary Care Posttraumatic Stress Disorder Screen for DSM-5; PHQ-2 = Patient Health Questionnaire-2; SC = Service Connection; VHA = Veterans Health Administration; Confidence intervals that did not include 1.00 were interpreted to be statistically significant.

Of the 425 veterans included in the present study, 281 (66%) reported having ever been diagnosed with PTSD, and 153 (36%) reported being treated for PTSD in the past year. Among veterans who reported being diagnosed with PTSD, VHA enrollees were again significantly more likely to report being treated for PTSD in the past year, OR = 5.36, 95% CI [1.85, 18.41], p = .003. Greater depression severity was also associated with increased treatment utilization, OR = 1.27, 95% CI [1.06, 1.52), p = .011. Finally, household income was associated with increased treatment utilization, OR = 1.33, 95% CI [1.08, 1.65], p = .007.

Overall Frequency of Mental Health Treatment Appointments (Table 5).

Table 5.

Full Sample Linear Regression Predicting Number of Mental Health Appointments

Predictor Variable Estimate Standard Error p-value
Age −0.00 0.01 .803
Sex −0.15 0.20 .461
Race 0.10 0.18 .582
VHA Enrollment 0.73 0.26 .005
SC −0.12 0.29 .690
Any Healthcare −0.26 0.17 .126
Household Income 0.08 0.06 .161
PC-PTSD-5 0.05 0.05 .315
PHQ-2 0.26 0.05 <.001
AUDIT-C −0.02 0.03 .537
Open in a new tab

Note. AUDIT-C = Alcohol Use Disorders Identification Test, 3-item version; PC-PTSD = Primary Care Posttraumatic Stress Disorder Screen for DSM-5; PHQ-2 = Patient Health Questionnaire-2; SC = Service Connection; VHA = Veterans Health Administration

A linear regression was computed to evaluate the relationship between predictor variables and frequency of mental health appointments of any kind in the past year (i.e., not diagnosis specific). VHA enrollment (B = .73, SE = .26, p = .005) and higher PHQ-2 scores were both associated with significantly higher mental health appointment frequency (B = .26, SE = .05, p < .0001).

Discussion

The aim of the present study was to explore factors associated with treatment utilization for depression and PTSD—two of the most common diagnoses among US Gulf War Era veterans (Toomey et al., 2007)—using a national sample of US Gulf War Era veterans. Our theoretical framework was based on the Behavioral Model of Healthcare Utilization (BMHU; Andersen, 1995). Our hypotheses were generally supported, in that variables from multiple theoretical domains were significantly associated with treatment utilization, including variables not included in previous tests of the BMHU.

We included veteran-specific variables that we conceptualized as potentially enabling service utilization; namely, whether veterans were VHA-enrolled and whether veterans were designated as having a service-connected health condition. We found that among veterans diagnosed with depression, VHA enrollees had three times greater odds of receiving depression treatment in the past year. Similarly, we found that among veterans diagnosed with PTSD, VHA enrollment was associated with five-times greater odds of receiving PTSD treatment. These findings were significant even when models controlled for enrollment in any health care system, suggesting that this was not likely due to engagement with the health care system in general, but that there may be characteristics specific to VHA that increase the likelihood that veterans will receive mental health care. Some VHA-specific characteristics that may contribute to this main finding include the service connection model, or the way in which mental health care is integrated into primary care settings (Pomerantz & Sayers, 2010). Service connection status was not significantly related to treatment utilization, although this may have been due to a relatively high correlation between VHA enrollment and service connection. In other words, it may have been that because VHA enrollment status was controlled for, service connection was no longer significantly related to treatment utilization. Regarding other VHA-specific characteristics, the VHA has prioritized increasing co-located collaborative care via strategies such as the establishment of primary care-mental health integration programs (PC-MHI; Post et al., 2010). These programs require periodic screens for depression and PTSD in primary care settings, where mental health providers are readily available to provide brief counseling and/or referral for other psychosocial treatments as needed. Although not specifically measured in the present study, PC-MHI initiatives may have contributed to the observed relationship between VHA enrollment and greater mental health treatment utilization.

An additional enabling resource significantly associated with treatment utilization in our sample was household income. Higher reported income was associated with a higher likelihood of receiving treatment for both PTSD and depression. Indicators of socioeconomic status such as income and education have been found to correlate with mental health utilization in other investigations, including those examining treatment for depression (Wittayanukorn et al., 2014) and PTSD (Sripada et al., 2015). This is especially noteworthy in the context of the VHA, in which the higher a veteran’s income is, the lower priority group they may be placed in; this priority group placement is used to determine how soon a veteran may be eligible for health care benefits and the amount of money they may have to pay towards the cost of their care (U.S. Department of Veterans Affairs, 2022). Thus, despite higher income increasing the likelihood of veterans being placed in a lower priority group, they were still more likely to receive mental health care in the present study. Further investigations into what drives this disparity are warranted.

The BMHU also includes demographic variables such as age, race, and sex or gender as variables that are theorized to be associated with a general predisposition for treatment seeking. Age is the most consistently significant variable regarding a relationship with treatment seeking (Peter et al., 2021; Seal et al., 2010; Fleming & Resick, 2017), but age was not significantly associated with treatment receipt in the present investigation. It may be that the present sample, which was comprised of only Gulf War Era veterans, was too narrowly constricted in its age range for analyses to detect true effects. On average, our sample was nearly 60 years old, with a standard deviation of approximately 7 years, and the youngest age represented was 47 years old. It may be that a larger age range would have increased statistical power to detect age-related effects.

Finally, our model included symptom-based variables, with our hypothesis being that the more severe veterans’ mental health symptoms were, the more likely they would be to receive treatment for their diagnoses. This hypothesis was generally supported. Higher scores on the PHQ-2 were associated with a greater likelihood of receiving treatment for PTSD and were associated with a higher overall frequency of mental health visits. Extended versions of this measures, such as the PHQ-9, have been shown to correlate with symptom-based measures across a variety of related disorders, and the present findings support the use of this tool as a potential indicator of transdiagnostic distress in settings where there are high rates of comorbid mental health conditions (Katz et al., 2021).

There are several notable strengths and limitations to the present study that should be considered when evaluating our findings. Strengths include the use of a national sample of Gulf War Era veterans, the use of a theory-driven modeling-based approach to understand treatment utilization, and inclusion of multiple relevant variables that have not been included in previous tests of this model, such as alcohol use, service connection, and VHA enrollment. One limitation to the present investigation was a reliance on self-report as opposed to an objective measurement of diagnoses and health care utilization such as a review of medical records. Indeed, self-reported health care utilization has been shown to moderately diverge from administrative records, especially when reporting on more remote time periods (i.e., outside of the past year) and thus replication of these results using alternative measurement strategies, such as gathering diagnoses and appointment visits via chart review, is warranted (Rhodes et al., 2002). Using record review may also allow exploration of treatment seeking behaviors across further time periods. We were not able to account for more remote treatment seeking behaviors in our models, which may have predictive value in understanding how current symptoms and resources relate to current treatment utilization. We ran multiple analyses examining both any usage (i.e., have you been treated for this diagnosis) and volume of usage (how many appointments attended) as a way to provide multiple measurements of health care utilization, but these nevertheless all relied on self-report data. Additionally, unmeasured variables could have accounted for variance in treatment seeking, such as distance from VHA facilities or cognitive factors such as beliefs about help-seeking (Seal et al., 2010; Elbogen et al., 2013). Future investigations may benefit from inclusion of these types of variables.

Other future directions may explore the relationship between VHA enrollment and mental health treatment utilization using longitudinal designs, which would provide further evidence supporting VHA enrollment as a strategy for increasing appropriate mental health service utilization.. Researchers may also explore how efficient the VHA community care referral process is, and whether being referred for care in the community versus within VHA results in varying likelihoods of patients availing themselves of those referrals. There is some evidence to suggest that veterans have more positive experiences within VHA than through community partners, and this has important implications for how VHA may address veterans’ mental health care needs (Vanneman et al., 2020).

Impact Statement.

This study recruited a national sample of Gulf War Era veterans with diagnoses of depression and/or posttraumatic stress disorder and examined factors associated with mental health treatment utilization. VHA-enrolled veterans were significantly more likely to receive mental health treatment than those who were not. VHA enrollment may reduce barriers to receiving mental health care in the veteran population.

Acknowledgements

This work was supported by grant # I01HX001682 from the Health Services Research & Development Service (HSRD) from the VA Office of Research and Development (ORD), which was awarded to Drs. Kimbrel & Pugh. Drs. Halverson and Blakey were supported by a VA Office of Academic Affiliations Advanced Fellowship in Mental Illness Research and Treatment. Dr. Beckham was funded by a Senior Research Career Scientist award from VA Clinical Sciences Research and Development (IK6BX00377). Dr. Pugh was funded by a Research Career Scientist Award from VA Health Services Research and Development (IK6HX002608). The views expressed in this article are those of the authors and do not necessarily reflect the position or policy of the United States Government, the Department of Veterans Affairs, or Duke University. The authors have no other conflicts of interest to disclose.

Footnotes

Ethics Approval Statement

This research was approved by the Durham Veterans Affairs Health Care System.

1

In a follow-up item, for those who indicated that they were service connected, participants also reported their service connection percentage, ranging from 0–100%. We ran analyses using the continuous 0%–100% SC rating and obtained the identical pattern of findings. For ease of interpretation, we present analyses with service connection coded dichotomously as having (1) or not having (0) a service connected condition.

References

  1. Andersen RM (1995). Revisiting the behavioral model and access to medical care: Does it matter? Journal of Health and Social Behavior, 36, 1–10. doi: 10.2307/2137284 [DOI] [PubMed] [Google Scholar]
  2. Baker TB, Piper ME, McCarthy DE, Majeskie MR, & Fiore MC (2004). Addiction motivation reformulated: An affective processing model of negative reinforcement. Psychological Review, 111(1), 33–51. 10.1037/0033-295X.111.1.33 [DOI] [PubMed] [Google Scholar]
  3. Bartels SJ, Coakley EH, Zubritsky C, Ware JH, Miles KM, Areán PA, Chen H, Oslin DW, Llorente MD, Costantino G, Quijano L, McIntyre JS, Linkins KW, Oxman TE, Maxwell J, Levkoff SE, & PRISM-E Investigators. (2004). Improving access to geriatric mental health services: A randomized trial comparing treatment engagement with integrated versus enhanced referral care for depression, anxiety, and at-risk alcohol use. American Journal of Psychiatry, 161(8), 1455–1462. [DOI] [PubMed] [Google Scholar]
  4. Blakey SM, Halverson TF, Evans MK, Patel TA, Hair LP, Meyer EC, DeBeer BB, Beckham JC, Pugh MJ, Calhoun PS, & Kimbrel NA (2021). Experiential avoidance is associated with mental and medical health diagnoses in a national sample of deployed Gulf War veterans. Journal of Psychiatric Research, 142, 17–24. doi: 10.1016/j.jpsychires.2021.07.033 [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Bolton JM, Robinson J, & Sareen J (2009). Self-medication of mood disorders with alcohol and drugs in the National Epidemiologic Survey on Alcohol and Related Conditions. Journal of Affective Disorders, 115(3), 367–375. [DOI] [PubMed] [Google Scholar]
  6. Bovin MJ, Koenig CJ, Zamora KA, Pyne JM, Miller CJ, Lipschitz JM, Wright PB, & Burgess JF Jr (2019). Veterans’ experiences initiating VA-based mental health care. Psychological Services, 16(4), 612–620. [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Bovin MJ, Kimerling R, Weathers FW, Prins A, Marx BP, Post EP, & Schnurr PP (2021). Diagnostic accuracy and acceptability of the primary care posttraumatic stress disorder screen for the Diagnostic and Statistical Manual of Mental Disorders (Fifth Edition) among US veterans. JAMA Network Open, 4(2), e2036733–e2036733. [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Bush K, Kivlahan DR, McDonell MB, Fihn SD, & Bradley KA (1998). The AUDIT alcohol consumption questions (AUDIT-C): An effective brief screening test for problem drinking. Archives of Internal Medicine, 158(16), 1789–1795. 10.1001/archinte.158.16.1789 [DOI] [PubMed] [Google Scholar]
  9. Carvalho JP, & Hopko DR (2011). Behavioral theory of depression: Reinforcement as a mediating variable between avoidance and depression. Journal of Behavior Therapy and Experimental Psychiatry, 42(2), 154–162. [DOI] [PubMed] [Google Scholar]
  10. Congressional Research Service (2018, November 1). VA Maintaining Internal Systems and Strengthening Integrated Outside Networks Act of 2018 (VA MISSION Act; P.L.115–182). https://crsreports.congress.gov/product/pdf/R/R45390
  11. Congressional Research Service (2021, July 19). Commander John Scott Hannon Veterans Mental Health Care Improvement Act of 2019 (P.L. 116–171) and Veterans COMPACT Act of 2020 (P.L. 116–214). https://crsreports.congress.gov/product/pdf/R/R46848.
  12. Crawford EF, Fulton JJ, Swinkels CM, Beckham JC, Calhoun PS, & VA Mid-Atlantic MIRECC OEF/OIF Registry Workgroup. (2013). Diagnostic efficiency of the AUDIT-C in US veterans with military service since September 11, 2001. Drug and Alcohol Dependence, 132(1–2), 101–106. [DOI] [PubMed] [Google Scholar]
  13. Cuijpers P, Noma H, Karyotaki E, Vinkers CH, Cipriani A, & Furukawa TA (2020). A network meta‐analysis of the effects of psychotherapies, pharmacotherapies and their combination in the treatment of adult depression. World Psychiatry, 19(1), 92–107. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Daughters SB, Magidson JF, Lejuez CW & Chen Y (2016). LETS Act: A behavioral activation treatment for substance use and depression. Advances in Dual Diagnosis, 9(2/3). 10.1108/ADD-02-2016-0006 [DOI] [Google Scholar]
  15. Desai RA, Dausey DJ, & Rosenheck RA (2005). Mental health service delivery and suicide risk: The role of individual patient and facility factors. American Journal of Psychiatry, 162(2), 311–318. [DOI] [PubMed] [Google Scholar]
  16. Dillman DA, Smyth JD, & Christian LM (2014). Internet, phone, mail, and mixed-mode surveys: The tailored design method (4th ed.). John Wiley & Sons. [Google Scholar]
  17. Elbogen EB, Wagner HR, Johnson SC, Kinneer P, Kang H, Vasterling JJ, Timko C, & Beckham JC (2013). Are Iraq and Afghanistan veterans using mental health services? New data from a national random-sample survey. Psychiatric Services, 64(2), 134–141. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Farmer CM, Hosek SD, & Adamson DM (2016). Balancing demand and supply for veterans’ health care: A summary of three RAND assessments conducted under the veterans choice act. RAND Health Quarterly, 6(1):12. [PMC free article] [PubMed] [Google Scholar]
  19. Fleming CE, & Resick PA (2017). Help-seeking behavior in survivors of intimate partner violence: Toward an integrated behavioral model of individual factors. Violence and Victims, 32, 195–209. doi: 10.1891/0886-6708.VV-D-15-00065 [DOI] [PubMed] [Google Scholar]
  20. Hawn SE, Cusack SE, & Amstadter AB (2020). A systematic review of the self‐medication hypothesis in the context of posttraumatic stress disorder and comorbid problematic alcohol use. Journal of Traumatic Stress, 33(5), 699–708. [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Katz IR, Liebmann EP, Resnick SG, & Hoff RA (2021). Performance of the PHQ-9 across conditions and comorbidities: Findings from the Veterans Outcome Assessment survey. Journal of Affective Disorders, 294, 864–867. [DOI] [PubMed] [Google Scholar]
  22. Kline AC, Cooper AA, Rytwinksi NK, & Feeny NC (2018). Long-term efficacy of psychotherapy for posttraumatic stress disorder: A meta-analysis of randomized controlled trials. Clinical psychology review, 59, 30–40. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Kroenke K, Spitzer RL, & Williams JB (2003). The Patient Health Questionnaire-2: Validity of a two-item depression screener. Medical Care, 1284–1292. [DOI] [PubMed] [Google Scholar]
  24. Magill M, Ray L, Kiluk B, Hoadley A, Bernstein M, Tonigan JS, & Carroll K (2019). A meta-analysis of cognitive-behavioral therapy for alcohol or other drug use disorders: Treatment efficacy by contrast condition. Journal of Consulting and Clinical Psychology, 87(12), 1093. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Manea L, Gilbody S, Hewitt C, North A, Plummer F, Richardson R, Thombs BD, Williams B, & McMillan D (2016). Identifying depression with the PHQ-2: A diagnostic meta-analysis. Journal of Affective Disorders, 203, 382–395. [DOI] [PubMed] [Google Scholar]
  26. National Center for Veterans Analysis and Statistics. (2020). Veteran population projection model 2018 (VetPop 2018). Retrieved from https://www.va.gov/vetdata/veteran_population.asp.
  27. Peter SC, Lipinksi AJ, Savage US, Dodson TS, Tran HN, Majeed R, & Beck JG (2021). Can the behavioral model of health care utilization be used to predict completion of a mental health assessment following intimate partner violence? Journal of Interpersonal Violence, 36, 7371–7392. [DOI] [PubMed] [Google Scholar]
  28. Pietrzak RH, Harpaz-Rotem I, & Southwick SM (2011). Cognitive-behavioral coping strategies associated with combat-related PTSD in treatment-seeking OEF–OIF veterans. Psychiatry Research, 189(2), 251–258. [DOI] [PubMed] [Google Scholar]
  29. Pomerantz AS, & Sayers SL (2010). Primary care-mental health integration in healthcare in the Department of Veterans Affairs. Families, Systems, & Health, 28(2), 78–82. doi: 10.1037/a0020341 [DOI] [PubMed] [Google Scholar]
  30. Post EP, Metzger M, Dumas P, & Lehmann L (2010). Integrating mental health into primary care within the Veterans Health Administration. Families, Systems, & Health, 28(2), 83–90. [DOI] [PubMed] [Google Scholar]
  31. Prins A, Bovin MJ, Smolenski DJ, Marx BP, Kimerling R, Jenkins-Guarnieri MA, Kaloupek DG, Schnurr PP, Kaiser AP, Leyva YE, & Tiet QQ (2016). The Primary Care PTSD Screen for DSM-5 (PC-PTSD-5): Development and evaluation within a veteran primary care sample. Journal of General Internal Medicine, 31, 1206–1211. 10.1007/s11606-016-3703-5 [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Rhodes AE, Lin E, & Mustard CA (2002). Self‐reported use of mental health services versus administrative records: Should we care?. International Journal of Methods in Psychiatric Research, 11(3), 125–133. [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Sayer NA, Spoont M, & Nelson DB (2004). Disability compensation for PTSD and use of VA mental health care. Psychiatric Services, 55(5), 589–589. [DOI] [PubMed] [Google Scholar]
  34. Seal KH, Metzler TJ, Gima KS, Bertenthal D, Maguen S, & Marmar CR (2009). Trends and risk factors for mental health diagnoses among Iraq and Afghanistan veterans using Department of Veterans Affairs health care, 2002–2008. American Journal of Public Health, 99(9), 1651–1658. [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Seal KH, Maguen S, Cohen B, Gima KS, Metzler TJ, Ren L, & Marmar CR (2010). VA mental health services utilization in Iraq and Afghanistan veterans in the first year of receiving new mental health diagnoses. Journal of Traumatic Stress, 23, 5–16. doi: 10.1002/jts.20493 [DOI] [PubMed] [Google Scholar]
  36. Spoont MR, Sayer NA, Nelson DB, & Nugent S (2007). Does filing a post-traumatic stress disorder disability claim promote mental health care participation among veterans?. Military Medicine, 172(6), 572–575. [DOI] [PubMed] [Google Scholar]
  37. Sripada RK, Richards SK, Rauch SA, Walters HM, Ganoczy D, Bohnert KM, Gorman LA, Kees M, Blow AJ, & Valenstein M (2015). Socioeconomic status and mental health service use among National Guard soldiers. Psychiatric Services, 66(9), 992–995. [DOI] [PubMed] [Google Scholar]
  38. Sripada RK, Hannemann CM, Schnurr PP, Marx BP, Pollack SJ, & McCarthy JF (2018). Mental health service utilization before and after receipt of a service‐connected disability award for PTSD: Findings from a national sample. Health Services Research, 53(6), 4565–4583. [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Staples LG, Dear BF, Gandy M, Fogliati V, Fogliati R, Karin E, Nielssen O, & Titov N (2019). Psychometric properties and clinical utility of brief measures of depression, anxiety, and general distress: The PHQ-2, GAD-2, and K-6. General Hospital Psychiatry, 56, 13–18. [DOI] [PubMed] [Google Scholar]
  40. Tabachnick BG, Fidell LS, & Ullman JB (2007). Using multivariate statistics (Vol. 5, pp. 481–498). Boston, MA: Pearson. [Google Scholar]
  41. Toomey R, Kang HK, Karlinsky J, Baker DG, Vasterling JJ, Alpern R, Reda DJ, Henderson WG, Murphy FM, & Eisen SA (2007). Mental health of US Gulf War veterans 10 years after the war. The British Journal of Psychiatry, 190(5), 385–393. [DOI] [PubMed] [Google Scholar]
  42. U.S. Department of Veterans Affairs (2018). Department of Veterans Affairs FY2018–2024 Strategic Plan. Washington, DC: U.S. Department of Veterans Affairs. [Google Scholar]
  43. U.S. Department of Veterans Affairs. (2019). Federal Benefits for Veterans, Dependents and Survivors. Washington, DC: U.S. Government Printing Office. [Google Scholar]
  44. U. S. Department of Veterans Affairs (2021, April 23). Veterans Health Administration: About VHA. https://www.va.gov/health/aboutvha.asp
  45. U. S. Department of Veterans Affairs (2022, March 8). VA Priority Groups. https://www.va.gov/health-care/eligibility/priority-groups/
  46. Vanneman ME, Wagner TH, Shwartz M, Meterko M, Francis J, Greenstone CL, & Rosen AK (2020). Veterans’ experiences with outpatient care: Comparing the Veterans Affairs system with community-based care. Health Affairs, 39(8), 1368–1376. [DOI] [PMC free article] [PubMed] [Google Scholar]
  47. Wittayanukorn S, Qian J, & Hansen RA (2014). Prevalence of depressive symptoms and predictors of treatment among US adults from 2005 to 2010. General Hospital Psychiatry, 36(3), 330–336. [DOI] [PubMed] [Google Scholar]
  48. Zivin K, Kim HM, McCarthy JF, Austin KL, Hoggatt KJ, Walters H, & Valenstein M (2007). Suicide mortality among individuals receiving treatment for depression in the Veterans Affairs health system: associations with patient and treatment setting characteristics. American journal of public health, 97(12), 2193–2198. [DOI] [PMC free article] [PubMed] [Google Scholar]

RESOURCES