The impact of substance use disorders on treatment engagement among justice-involved veterans with posttraumatic stress disorder

Abstract

Veterans involved with the criminal justice system represent a particularly vulnerable population who experience high rates of both posttraumatic stress disorder (PTSD) and substance use disorders (SUD). This study sought to investigate whether having co-occurring SUD is a barrier to PTSD treatment. This is a retrospective observational study of a national sample of justice-involved veterans served by the Veterans Health Administration (VA) Veterans Justice Outreach (VJO) program who had a diagnosis of PTSD (N = 27,857). Mixed effects logistic regression models with a random effect for facility (N = 141 medical centers) were utilized to estimate the odds of receiving each type of PTSD treatment as a function of having a SUD diagnosis. Results indicate that a majority of veterans with PTSD served by VJO have a SUD diagnosis (73%), and having a co-occurring SUD was associated with higher odds of receiving PTSD treatment, after adjusting for demographic differences. Although not without limitations, these results suggest that among justice-involved veterans enrolled in VHA with PTSD, having a SUD comorbidity is not a barrier to PTSD treatment and may in fact facilitate access to PTSD treatment.

Posttraumatic stress disorder (PTSD) frequently co-occurs with substance use disorders (SUDs). Research suggests that the two conditions are functionally related with substances often used to help ameliorate symptoms of PTSD (Carter, Capone, & Short, 2011; Jacobsen, Southwick, & Kosten, 2001; Leeies, Pagura, Sareen, & Bolton, 2010; Simpson, Stappenbeck, Luterek, Lehavot, & Kaysen, 2014). Veterans with co-occurring PTSD and SUD have been found to have higher rates of psychotherapy utilization (Harpaz-Rotem & Rosenheck, 2011; Hundt et al., 2014). However, there is a negative association between SUD and mental health treatment utilization among individuals with trauma histories, which has largely been found among younger veterans with more recent trauma exposure (van den Berk-Clark & Wolf, 2015). Although limited in number, some studies investigating treatment utilization among veterans with PTSD found that while the presence of a co-occurring SUD was associated with more visits or initiation of PTSD treatment services, it was also associated with drop out (DeViva, 2014; Harpaz-Rotem & Rosenheck, 2011).

Although recent Veterans Health Administration (VA) guidelines on treating PTSD/SUD have sought to shift the culture to treating both disorders concurrently (Bernardy, Hamblen, Friedman, & Kivlahan, 2011), SUD may still serve as a barrier to access to treatment for PTSD. In VA, the standard of care historically was to treat SUDs first and then address PTSD treatment needs (Ruzek, 2002). Research has shown that many providers believe that SUDs must be treated prior to initiation of PTSD treatment for a variety of reasons, including beliefs that patients with co-occurring PTSD and SUDs are harder to treat than those with only PTSD, (Becker, Zayfert, & Anderson, 2004; Najavits, Norman, Kivlahan, & Kosten, 2010). In fact, research investigating the efficacy of PTSD or SUD treatment often excludes individuals with more complex clinical presentations, despite other research indicating that treating both simultaneously can be effective (Roberts, Roberts, Jones, & Bisson, 2015). Given the functional relationship between substance use and PTSD symptoms themselves, it may also be that providers refer veterans to SUD treatment because it is the presenting problem at the time of assessment.

Veterans in the criminal justice system represent a specific and vulnerable subset of both Veterans and criminal justice populations nationally. Adults in the criminal justice system experience significantly greater exposure to traumatic events over their lives compared to the general population (Bloom, Owen, & Covington, 2003; Weeks & Widom, 1999). The estimated prevalence of lifetime physical or sexual abuse experienced by men in the criminal justice system ranges from 25% to 68% (James & Glaze, 2006; Weeks & Widom, 1999) and the estimated prevalence of posttraumatic stress disorder (PTSD) diagnoses among incarcerated men ranges from 20% to 55% (Proctor & Hoffmann, 2012; Trestman, Ford, Zhang, & Wiesbrock, 2007). As many as 55% of incarcerated women have experienced physical or sexual abuse in their lifetime (Osher & Steadman, 2007), with PTSD prevalence rates ranging from 17% to 48% (Brinded, Simpson, Laidlaw, Fairley, & Malcolm, 2001; Green, Miranda, Daroowalla, & Siddique, 2005; Harner, Budescu, Gillihan, Riley, & Foa, 2015; Osher & Steadman, 2007; Teplin, Abram, & McClelland, 1996). The prevalence rates vary across studies depending on a variety of factors including specific nature of incarceration (e.g., jail versus prison), location (U.S. versus international sample), design (e.g. convenience sampling), type of trauma exposure, and timing of symptom endorsement (i.e., lifetime prevalence, 6-month prevalence, or 1-month prevalence rates).

Military veterans in the justice system may have even higher rates of trauma exposure. One study estimated that among veterans incarcerated in jails, 87% have been exposed to traumatic events, including events in their homes and communities as well as during military service, and 31% have been diagnosed with PTSD (Saxon et al., 2001). This is consistent with the recent Bureau of Justice Statistics report that found 31% of male veterans in jails and 23% of male veterans in prison had been previously diagnosed with PTSD – twice the prevalence of civilian inmates in either sample (Bronson, Carson, Noonan, & Berzofsky, 2015). Another study found that 93% of veterans being diverted from jail experienced some trauma exposure in their lifetime, with 55% endorsing symptoms consistent with a PTSD diagnosis (Hartwell et al., 2014). Although there is a dearth of research on women veterans in the criminal justice system (including being excluded from the BJS report due to small sample size), they may represent a particularly vulnerable population: 83–99% of women veterans report at least one lifetime traumatic event (Zinzow, Grubaugh, Monnier, Suffoletta-Maierle, & Frueh, 2007), and the estimated prevalence of PTSD ranges from 25% to 62% among women veterans depending on type and number of traumatic events endorsed (Yaeger, Himmelfarb, Cammack, & Mintz, 2006).

Veterans Justice Outreach Program

The Veterans Justice Outreach program (VJO) was established in 2009 to help address the extensive needs of veterans involved in the criminal justice system. (Blue-Howells, Clark, van den Berk-Clark, & McGuire, 2013). This includes outreach to and assessment of veterans who are incarcerated in county jails, staffing of veterans treatment courts, training of law enforcement and other criminal justice partners on military culture, PTSD and traumatic brain injury (TBI), and VA resources. The overarching goal of the VJO program is to prevent homelessness and criminal justice recidivism, and to facilitate prosocial engagement in the community life. The primary function of VJO then is to provide linkage to needed services, the specifics of which are determined based on assessments with individual veterans. In sum, over 77,136 veterans were seen by a VJO specialist between 2010–2014 (Finlay, Rosenthal, et al., 2016).

Although VJO specialists interact with a diverse veteran population, given the high rates of trauma exposure and PTSD, as well as PTSD and SUD comorbidity among justice-involved veterans, it is important to determine whether these veterans are receiving appropriate treatment for these conditions. Recent research suggests justice-involved veterans have high rates of accessing mental health treatment generally, including pharmacotherapy (Finlay, Smelson, et al., 2016); however, it is unclear whether justice-involved veterans with PTSD are actually accessing PTSD-specific treatment. Although there are no prior studies of PTSD treatment access among justice-involved veterans, a study of veterans newly diagnosed with PTSD found that less than 10% engaged in at least 9 sessions of PTSD treatment (Seal et al., 2010). Similarly, in a study investigating initiation and completion of two evidence-based treatments recommended by the VA for PTSD (i.e., prolonged exposure and cognitive processing therapy) in a VA PTSD/anxiety clinic, only 11% of veterans with a clinic visit began treatment and 8% completed treatment (Mott et al., 2014). Although SUDs were not associated with initiating or completing PTSD treatment, given the high prevalence of co-occurring SUDs that were present in both the non-initiating and non-completing samples, further investigation is needed to understand whether having co-occurring SUD may be a barrier to PTSD-specific treatment utilization and/or completion.

Given the high comorbidity between PTSD and SUD among veterans, the overall lack of research investigating whether having a SUD is a barrier to receiving PTSD treatment including pharmacotherapy, and the relative lack of information on justice-involved veterans with PTSD and co-occurring SUDs, the present study was designed to investigate the prevalence of PTSD and co-occurring SUD diagnoses among justice-involved veterans with PTSD, and whether or not having a SUD diagnosis was related to PTSD treatment use among justice involved veterans after controlling for demographic variables. We hypothesized that justice-involved veterans with PTSD and no co-occurring SUD would be more likely to enter PTSD treatment and would use more PTSD treatment services than justice-involved veterans with PTSD and a co-occurring SUD.

Method

Participants

Our sample was limited to the 27,653 veterans who received an outreach visit from a VJO Specialist in fiscal years 2010–2014 and had a diagnosis of PTSD documented in their clinical record in the following year. Of these veterans, 7,495 (27%) had a diagnosis of PTSD and no co-occurring SUD, and 20,158 (73%) had a diagnosis of PTSD and comorbid SUD in the same year. Table 1 presents the patient characteristics of the sample, stratified by SUD. A higher proportion of women compared to men were diagnosed with PTSD and not SUD (10% vs. 7%; p < .001). Compared to those with PTSD only, veterans with co-occurring SUD were more likely to be single (40% vs. 31%) or divorced/separated (33% vs. 29%) and more likely to have a mental health disorder other than PTSD (87% vs. 76%). Almost twice as many veterans with a co-occurring SUD were homeless compared to those with PTSD only (55% vs. 28%).

Table 1.

Demographic Characteristics of Veterans Health Administration Patients With Post-Traumatic Stress Disorder Seen by Veteran Justice Outreach Specialists in Fiscal Years 2010–2014

	PTSD		PTSD+SUD
	n	(%)	n	(%)	Χ2	p	φ
Gender					114.80	< .001	−.06
Women	781	10	1,325	7
Men	6,712	90	18,822	93
Age
< 25	644	9	1,850	9	1.57	.21	.01
25–34	2,374	32	6,432	32	.88	.35	.01
35–44	1,305	17	3,255	16	8.48	< .01	−.02
45–54	1,206	16	4,142	21	68.04	< .001	.05
55+	1,966	26	4,478	22	47.23	< .001	−.04
Ethnicity/Race
Hispanic	790	11	1,602	8	51.66	< .001	−.04
Non-Hispanic
American Indian/Alaskan Native	148	2	447	2	1.18	.28	.01
Asian	131	2	285	2	4.69	< .05	−.01
Black/African American	1,520	22	4,651	25	19.85	< .001	.03
White	4,352	63	11,950	63	.37	.54	.00
Marital status					485.71	< .001	.13
Single	2,316	31	7,976	40
Married	2,867	38	5,039	25
Divorced/Separated	2,137	29	6,711	33
Widowed	114	2	320	2
Residence					21.82	< .001	−.03
Rural	2,257	30	5,504	27
Urban	5,225	70	14,640	73
Homeless					1595.68	< .001	.24
No	5,424	72	9,149	46
Yes	2,071	28	11,009	55
Iraq or Afghanistan veterans					27.26	< .001	−.03
No	3,879	52	11,142	55
Yes	3,616	48	9,016	45
Service-connected disability rating					225.67	< .001	.09
No	1,247	17	4,440	22
< 50%	2,605	35	7,865	39
≥ 50%	3,643	49	7,853	39
Co-occurring mental health disorders					455.53	< .001	.13
No	1,791	24	2,675	13
Yes	5,704	76	17,483	87

Note. PTSD group: n = 7,495; PTSD+SUD group: n = 20,158. Service-connected = a disability caused by medical or psychiatric illness or injury occurring during or aggravated by military service. PTSD = post traumatic stress disorder. SUD = substance use disorder.

^*p < .05.

^**p < .01.

^***p < .001.

Procedure

Using national VA electronic health records, we conducted a retrospective observational study of veterans with justice involvement who were diagnosed with PTSD. Health care utilization data were drawn from the Corporate Data Warehouse, a repository of all electronic health records in the VA. Of the ~6 million veterans served each fiscal year at VA facilities, we identified all veterans who had contact with the Veterans Justice Outreach (VJO) program anytime from fiscal year 2010 through fiscal year 2014 (October 1, 2009-September 30, 2014) using clinic stop code 592. In addition, veterans who had a VJO outreach record indicating contact with the program but did not have a 592 clinic stop code were also included in the sample. We excluded veterans who had an outreach visit in prison (provided by VA’s Health Care for Reentry Veterans program; clinic stop code 591) because veterans who are prison-involved tend to differ in their sociodemographic characteristics and treatment use from veterans who are jail/court-involved (Finlay & Rosenthal, 2015), including having lower prevalence of mental health and substance use disorders. This study is part of a larger project on justice-involved veterans which was approved by the Stanford University Institutional Review Board, which serves the VA Palo Alto Health Care System.

Measures

PTSD and SUDs

Diagnoses were drawn from outpatient encounters or residential bed sections from patients’ electronic health records. A veteran was considered to have PTSD if they had at least one instance of the diagnosis documented in their outpatient health record using International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM) code 309.81 in the one-year period after their initial VJO encounter. Diagnosis of a SUD in the same period was based on at least one ICD-9-CM diagnosis of an alcohol use disorder, opioid use disorder, cocaine use disorder, amphetamine use disorder, cannabis use disorder, sedative use disorder, or other drug use disorder.

Patient characteristics

Demographic variables included gender (male, female), age (<25, 25–34, 35–44., 45–54, 55+), ethnicity/race (Hispanic, non-Hispanic: American Indian/Alaskan Native, Asian, Black/African American, White), and marital status (single, married, divorced/separated, widowed). Other patient characteristics included residence (living in an urban or rural area), homeless status (no, yes), which indicates use of homeless services or being homeless or at-risk for homelessness (based on outpatient clinic codes and residential treatment codes for homeless services, and ICD-9-CM codes related to housing or homelessness), Iraq or Afghanistan veteran status (no, yes) and overall service-connected disability rating (none, <50%, 50–100%), indicating a disability caused by medical or psychiatric illness or injury occurring during or aggravated by military service. Patient records were also coded for co-occurring psychiatric disorders (no, yes), including depressive disorders, anxiety disorders, bipolar disorder, schizophrenia, other psychosis, or personality disorders.

Treatment utilization

PTSD outpatient visits counted the number of visits to VA PTSD treatment clinics for each veteran in the one year period after their VJO encounter. Outpatient visits were coded using clinic stop codes and included Post Traumatic Stress Disorder Group (516), Substance Use Disorder / Post Traumatic Stress Disorder Teams (519), Active Duty Sex Trauma (524), Post Traumatic Stress Disorder Clinical Team - Individual (540), Post Traumatic Stress Disorder Clinical Team - Group (561), Post Traumatic Stress Disorder - Individual (562), and Post Traumatic Stress Disorder Day Hospitalization (580). PTSD residential days counted for each patient the number of days in a VA residential treatment program in the year after their VJO encounter. Residential days were coded using bed sections and included Domiciliary PTSD (code 88) and PTSD Residential Rehabilitation Treatment Program (110). Pharmacotherapy for PTSD symptoms measured whether or not a patient filled a prescription for paroxetine, sertraline fluoxetine, fluvoxamine, citalopram, escitalopram, venlafaxine, imipramine, amitripryline, desipramine, nortriptyline, protriptyline, clomipramine, phenelzine, tranylcypromine, mirtazapine, nefazodone, prazosin, or duloxetine at a VA pharmacy in the year after their VJO visit. While these medications are not “PTSD medications” they are often provided to treat symptoms of PTSD (including negative mood, anxiety, and nightmares) and co-occurring mental health disorders, and two (sertraline and paroxetine) are approved for PTSD by the Food and Drug Administration. Selection of medications was based on the 2010 VA Clinical Guidelines and included medications that were recommended as first-line or second-line (Department of Veterans Affairs and Department of Defense, 2010). Medications with insufficient evidence to recommend for use or against use, those unlikely to be beneficial, and those thought to be harmful were not included.

Results

Data Analysis

We examined descriptive statistics of patient characteristics and tested differences between patients with and without a SUD diagnosis using Χ² tests. We also examined utilization of PTSD treatment stratified by SUD. For outpatient visits, the sample was limited to veterans who had one or more outpatient visits. For residential days, the sample was limited to veterans who had one or more residential days. Means and standard deviations were reported and differences were tested using t-tests. For pharmacotherapy, all veterans in the sample were included and number (percentage) who received treatment was reported with a Χ² test used to test differences between the groups. Finally, we used all veterans in the sample and conducted three mixed effects logistic regression models to examine the odds of receiving each type of PTSD treatment (yes or no received outpatient [model 1], residential [model 2] or pharmacotherapy [model 3] treatment) as a function of having a SUD diagnosis. The mixed effects logistic regression models (Gelman & Hill, 2007) included a random effect for facility (N = 141 medical centers) to account for patients nested within VA facilities and were adjusted for all other patient characteristics including gender, age, race/ethnicity, marital status, residence, homeless status, service in Iraq or Afghanistan, service-connected disability rating, and co-occurring psychiatric disorder. A Bonferroni correction was used to adjust for multiple regression models. All data analyses were conducting using SAS 9.2 statistical software (SAS Institute, Cary NC).

Treatment Use

Among veterans with PTSD but no documented SUD, 35% had one or more PTSD outpatient visits, 2% had one or more days in residential treatment, and 66% received a prescription for one or more medications to treat PTSD symptoms. Among veterans with PTSD and co-occurring SUD, 41% had one or more PTSD outpatient visits, 7% had one or more days in residential treatment, and 76% received a prescription for one or more medications to treat PTSD symptoms (Table 2). Additionally, veterans with PTSD+SUD had one additional PTSD outpatient visit on average compared to those with PTSD without SUD (10.98 vs. 9.67, t = −4.49, p<0.001 (Table 2). Mean number residential days was similar regardless of SUD status.

Table 2.

Use of Posttraumatic Stress Disorder Treatment by VHA Veteran Patients Seen by Veterans Justice Outreach Specialists in Fiscal Years 2010–2014

	PTSD only			PTSD/SUD				Cohen’s
PTSD Treatment	n	M	SD	n	M	SD	t	p	d
Outpatient visits	2,590	9.67	11.99	8,328	10.98	13.35	−4.49	< .001	.10
Residential days	123	46.37	29.10	1,334	46.90	28.45	−.20	.84	.02
	n	%		n			Χ2	p	φ
Pharmacotherapy	4,961	66		15,277	76		256.34	< .001	.10

PTSD = post traumatic stress disorder. VHA = Veterans Health Administration.

Table 3 displays results of the logistic regression models Veterans with PTSD+SUD had higher odds of receiving PTSD outpatient, residential, or pharmacotherapy compared to veterans with PTSD only. Women had higher odds of receiving outpatient and residential treatment and pharmacotherapy compared to men. Non-Hispanic Black/African American veterans had higher odds of receiving outpatient treatment, but lower odds of receiving pharmacotherapy compared to non-Hispanic White veterans. Veterans living in rural areas had higher odds of receiving residential treatment compared to veterans in urban areas. Veterans who were receiving homeless services or at risk for homelessness had higher odds of receiving outpatient or residential PTSD treatment compared to other veterans. Having a service-connected disability rating and having a co-occurring mental health disorder was associated with higher odds of receiving all types of PTSD treatment.

Table 3.

Characteristics Associated with PTSD Treatment Use Among Veterans with an Veterans Justice Outreach Visit in Fiscal Years 2010–2014

	Outpatient		Pharmacotherapy		Residential
Characteristics	OR	95% CI	OR	95% CI	OR	95% CI
Substance use disorder	1.28a	[1.21, 1.37]	3.83a	[3.14, 4.66]	1.43a	[1.34, 1.52]
Gender (ref: male)	1.20a	[1.09, 1.32]	0.83	[0.64, 1.06]	1.14	[1.02, 1.27]
Age, years (ref: <25)
25–34	1.06	[0.94, 1.19]	0.96	[0.76, 1.22]	0.99	[0.87, 1.13]
35–44	1.24a	[1.09, 1.42]	1.27	[0.97, 1.66]	1.13	[0.98, 1.30]
45–54	1.29a	[1.12, 1.48]	0.96	[0.71, 1.30]	1.24a	[1.06, 1.45]
55+	1.11	[0.96, 1.29]	0.74	[0.53, 1.03]	1.02	[0.87, 1.19]
Race/ethnicity (ref: non-Hispanic White)
American Indian/Alaskan Native	1.12	[0.93, 1.34]	0.90	[0.59, 1.36]	0.82	[0.67, 0.98]
Asian	0.89	[0.72, 1.11]	1.55	[1.01, 2.36]	0.88	[0.70, 1.10]
Black/African American	1.16a	[1.08, 1.24]	0.94	[0.80, 1.11]	0.81a	[0.75, 0.87]
Hispanic	1.13	[1.02, 1.25]	0.91	[0.71, 1.17]	1.02	[0.92, 1.13]
Unknown	0.95	[0.84, 1.06]	0.68	[0.46, 0.96]	0.77a	[0.68, 0.86]
Marital status (ref: married)
Single	0.87a	[0.82, 0.93]	0.86	[0.75, 1.00]	0.74a	[0.69, 0.79]
Divorced/separated	0.90a	[0.84, 0.96]	0.93	[0.80, 1.08]	0.82a	[0.76, 0.88]
Widowed	0.83	[0.66, 1.03]	1.19	[0.69, 2.05]	0.75	[0.60, 0.94]
Rural (ref: urban)	1.04	[0.98, 1.12]	1.37a	[1.20, 1.56]	1.09	[1.02, 1.17]
Homeless (ref: no)	1.18a	[1.12, 1.25]	1.56a	[1.38, 1.77]	1.34	[1.26, 1.43]
Service in Iraq/Afghanistan (ref: no)	1.74a	[1.60, 1.88]	2.35a	[1.96, 2.82]	1.09	[1.00, 1.20]
Service-connected disability rating (ref: no)
< 50%	1.65a	[1.54, 1.78]	2.45a	[2.00, 3.01]	1.39a	[1.29, 1.50]
≥ 50%	1.58a	[1.46, 1.70]	2.83a	[2.31, 3.48]	1.44a	[1.33, 1.55]
Other mental health disorder (ref: no)	1.31a	[1.22, 1.40]	2.10a	[1.71, 2.57]	2.9a	[2.75, 3.16]

Note. N =27,612. Cases with missing data (n = 41; <1%) were excluded from the logistic regression models. AOR = adjusted odds ratio. CI = confidence interval.

^aSignificant with Bonferroni correction (p < .05 threshold lowered to p < .016).

Discussion

The current study indicated that overall there was high rate of treatment access for justice-involved Veterans who had PTSD diagnoses as well as co-occurring PTSD and SUD diagnoses. More specifically, justice-involved veterans diagnosed with co-occurring PTSD and SUDs have higher odds of receiving outpatient, residential and pharmacotherapy PTSD treatment than veterans diagnosed with PTSD but no co-occurring SUD. The actual difference in number of visits received, particularly residential treatment days, is negligible, which limits clinical inferences that can be made based on the results. However, these findings are clinically significant in that that they demonstrate broad utilization of PTSD services among Veterans working with the VJO program. Additionally, although the research design of the current study is not amenable to investigating clinical outcomes of specific PTSD interventions (e.g., symptom inventory scores), it is noteworthy that among VA patients with PTSD served by the VJO program, having a co-occurring SUD disorder may not be a barrier to PTSD treatment, and may marginally increase the likelihood of accessing PTSD treatment. Recent research investigating mental health and SUD treatment engagement among justice-involved veterans finds high rates of linkage to treatment in this population (Finlay, Smelson, et al., 2016), which is likely facilitated by emphasis the VJO program places on connecting incarcerated veterans to mental health and substance use treatment within VA health care settings.

Furthermore, justice-involvement may itself also account for higher rates of treatment engagement as veterans may be mandated to SUD treatment, thereby opening the door for referrals to additional specialized PTSD treatment. Research suggests that mandated substance use and mental health treatment can be effective in treating these conditions (Kelly, Finney, & Moos, 2005; Link, Castille, & Stuber, 2008).

In addition, the high rates of comorbidity between PTSD and SUDs may further explain these results (Ralevski, Olivera-Figueroa, & Petrakis, 2014). In the current study 72% of veterans with VJO contact and a subsequent PTSD diagnosis had a co-occurring SUD. While the reasons for arrest are beyond the scope of this paper, some veterans with co-occurring SUDs were in jail or involved with Veterans Treatment Courts as a result of alcohol- or drug-related charges. As such, specific treatment planning with a VJO specialist (and associated court-based mandates with those in Veterans Treatment Courts) would be likely to include referral for SUD specific treatment. This treatment connection may have helped veterans connect to the VA, entered them into a healthcare system that de-stigmatizes trauma and mental health treatment, increased the opportunity for additional PTSD assessment, introduced them to other veterans with similar issues, and provided them with tools to discuss difficult problems (including acknowledging and discussing trauma exposures both in combat and at home). While these specific mechanisms of engagement require further research, evidence does support higher utilization of SUD care among justice-involved veterans compared to other veterans who use VA services (Finlay, Smelson, et al., 2016). Furthermore, the growing number of Veteran Treatment Courts (VTCs) nationally provide legal mandates to enter treatment and the high rate of PTSD treatment among veterans with VJO contact in this study is consistent with studies of VTCs demonstrating engagement in PTSD treatment and improvement in PTSD symptoms (Knudsen & Wingenfeld, 2016).

In addition, there have been increased efforts by the VA to address co-occurring PTSD and SUD concurrently in response to evidence supporting the efficacy of treating these conditions together (Seal et al., 2011). For example, while there is limited research on the efficacy of integrated substance use/PTSD programs, there is research indicating that treating PTSD in the context of an alcohol use disorder actually improves alcohol related symptoms prior to improvements in PTSD symptoms (Back, Brady, Sonne, & Verduin, 2006), and a recent study found that having a current or past alcohol use disorder did not reduce benefits of engaging in cognitive processing therapy (Kaysen et al., 2014). Although the effect sizes in these studies are generally small (Roberts et al., 2015), these findings support the notion that treating PTSD and SUDs concurrently is both appropriate and effective, which is consistent with recommendations made by the VA/DoD Clinical Practice Guidelines for PTSD (Department of Veterans Affairs and Department of Defense, 2010).

There are several limitations of the current study. First, the data were collected from encounter level information, meaning little is known about the specifics of the veteran’s PTSD (i.e., combat vs. military sexual trauma), the severity of their PTSD symptoms, the nature of their treatment (e.g., trauma informed therapies vs. trauma focused therapies), whether they received PTSD treatment through a different clinic stop code (e.g., general mental health clinic) or through referral to Vet Centers (affiliated but independently documented treatment centers for combat Veterans), and the overall extent of their sustained engagement in treatment once diagnosed and referred to PTSD treatment. We also do not have information on whether they received an evidence-based treatment (e.g., CPT or PE) during the outpatient visits or residential day, limiting the ability to utilize the clinic codes as a proxy for PTSD-specific therapies. Given that many VA outpatient and residential programs address both SUD and PTSD treatment, and that integrated treatment programs were included in the data analysis, it is possible that increased utilization within this population in the current sample is a result of this integrated treatment approach, and further analyses should be conducted to better understand whether the results were shaped by where the veterans received treatment. Second, the study focused specifically on those treatment encounters at an outpatient clinic or inpatient program specializing in PTSD treatment. However, given the heterogeneous symptoms that comprise a PTSD diagnosis, it may be that justice-involved veterans were receiving other forms of mental health treatment that were actually targeting PTSD-related symptoms such as depression or anxiety. This may be especially problematic with using psychopharmalogical interventions as a proxy for PTSD treatment given the overlap of for PTSD and other mental health disorders and the administration of certain medications to assist with chronic pain and other medical conditions. Third, the sample is restricted to veterans who were eligible for and used VA health care and so conclusions may not be generalizable to other justice-involved veterans with PTSD. Similarly, even for those VA patients with documentation of PTSD, they may be receiving community treatment for PTSD, which is not captured in the current data.

Conclusion

The current study provides important information about the impact of having co-occurring substance use disorders on accessing PTSD treatment among justice involved veterans. Results demonstrate that having a co-occurring SUD does not serve as a barrier to receiving PTSD treatment, and may in fact increase access to PTSD treatment. This is particularly important given the special treatment needs of justice-involved veterans (as demonstrated by the high rates of co-occurring PTSD and SUD), and suggests that despite mixed results on the impact of SUD on PTSD treatment access, having a connection to an additional provider through the VJO program may help this vulnerable population receive much needed services.