Nearly 25% of Veterans seen in VA and diagnosed with posttraumatic stress disorder (PTSD) have a co-occurring substance use disorder (SUD) diagnosis, and over 25% of Veterans with an SUD have co-occurring PTSD. Patients diagnosed with both disorders often have poorer long-term prognoses for each condition than singly diagnosed patients. The high rates of co-occurring SUD and PTSD led to specific recommendations by the VA's Office of Mental Health Operations (OMHO) and Mental Health Services (MHS) for service provision. Specific requirements include provision of coordinated, and, where possible, concurrent treatment of SUD and other co-occurring conditions, and require that PTSD programs address the needs of Veterans with co-occurring PTSD and SUD. However, to date, well-established treatment approaches or models of care for achieving this goal are lacking.
Objectives were to: (a) examine variability among VA facilities in specialty care visits and dosing of psychotherapy among Veterans with PTSD based on SUD comorbidity status; (b) determine at the patient level whether the presence of co-occurring SUD diagnoses explains variation in specialty care visits and dosing of psychotherapy among Veterans with PTSD; and (c) identify facility-level characteristics associated with relatively high and low SUD-related disparities in PTSD care (i.e., dose of psychotherapy and access to PTSD specialist does or does not differ strongly by SUD status, respectively).
Using a retrospective cohort design, we leveraged national VA administrative data to identify Veterans between 18-89 years of age with two or more primary PTSD encounters (i.e., outpatient visits or inpatient stays with a primary PTSD diagnosis) receiving PTSD care in the VA medical system during fiscal years (FY) 2009-2010. To examine all primary PTSD encounters for one year, we included only individuals who were alive at the end of FY11. We examined individual- and facility-level factors that may influence PTSD treatment among Veterans, focusing on those with and without co-occurring SUD. Factors examined included facility-level characteristics of the local VA health care system (e.g., location, mental health staffing levels), and characteristics of the Veteran (e.g., demographics, co-occurring SUD).
The collective 1-year follow-up period among our sample's 424,211 Veterans yielded roughly 5 million outpatient PTSD visits and 21,000 inpatient PTSD stays. Co-occurring SUD prevalence (n=67,717; 16%) was explained nearly equally by alcohol (n=52,389; 77% of those with co-occurring SUD) and non-alcohol disorders (n=49,985; 74% of those with co-occurring SUD); 18,523 Veterans (27% of those with co-occurring SUD) were diagnosed with both alcohol and non-alcohol disorders during follow-up. Among Veterans in the sample with inpatient PTSD stays during follow-up (n=16,150), co-occurring SUD was only slightly more likely than not (52% vs. 48%). Conversely, SUD comorbidity was the exception (16%) rather than the rule among Veterans with outpatient PTSD visits (n=424,107). Regardless of SUD status, most Veterans' outpatient PTSD care occurred through general mental health specialty clinics rather than PTSD specialty clinics. Individuals with co-occurring SUD were more likely to have utilized a mix of specialty (i.e., PTSD and/or general mental health) and non-specialty care services for their PTSD care compared to those without SUD (42% vs. 23%), with very few receiving their PTSD care exclusively in SUD or PTSD-SUD specialty programs (4% vs. 0.2% among those without SUD diagnoses). A small percentage of Veterans with and without co-occurring SUD did not receive any outpatient PTSD care through specialty mental health or PTSD services (4% vs. 6%, respectively). For inpatient care, two-thirds of Veterans with co-occurring SUD (n=5,330) did not receive any PTSD inpatient specialty care when admission was tied to their PTSD symptoms (vs. 54% among the 7,672 Veterans without SUD who had a PTSD-related inpatient stay during their 1-year follow-up). The remaining Veterans in both SUD groups tended to split their inpatient PTSD care between residential and domiciliary PTSD units.
Veterans with co-occurring SUD (vs. without) were more likely to have received any PTSD specialty care (OR=1.19, 95% CI 1.17-1.22) and any mental health specialty services (incl. PTSD, SUD, or general mental health specialty care services; OR=1.66, 95% CI 1.60-1.73) for their PTSD after controlling for demographic and facility-level factors. We examined dose of psychotherapy using VA's PTSD performance measure for Veterans of Operations Enduring Freedom, Iraqi Freedom, and New Dawn (OEF/OIF/OND) as a guide (i.e., 8 psychotherapy visits within 14 weeks in a 1-year period). Among OEF/OIF/OND participants, 13% met the performance measure during their 1-year follow-up period (11% without SUD vs. 23% with SUD). When we examined only those OEF/OIF/OND Veterans who had not met the performance measure in the previous 5 years (inclusion criterion for the official metric), overall and SUD subgroups followed a similar pattern (i.e., 10% without SUD vs. 21% with SUD). Considering all service periods, 15% met the performance measure without regard to the prior 5 years (13% without SUD vs. 24% with SUD).
In comparing facilities with high vs. low SUD-related disparities in PTSD care (i.e., 15% or greater vs. 5% or lower, respectively), the only factors found to significantly predict group membership were the prevalence of comorbid SUD among each facility's PTSD population (outpatient care) and the percentage of a facility population's PTSD visits occurring at a VAMC (inpatient care).
This study directly examined whether co-occurring substance use disorders impart specific barriers to accessing PTSD specialty treatment. Results suggest Veterans with co-occurring SUD are obtaining their PTSD care in specialty care settings comparable to their non-SUD counterparts, even more so in some cases. We expect these findings will guide efforts aimed at further improving access to adequate PTSD treatment, and underscore documenting and evaluating facility-level variability in performance to identify sites that may have best and problematic practices.
External Links for this Project
Grant Number: I21HX001194-01
None at this time.