Objectives: The aims of the study were to (1) identify personal, social, and clinical history for patients with substance use disorder (SUD) and mental illness, (2) measure agreeance between patient self-report versus facility record history for mental illness, substance abuse, and psychotropic medication, (3) investigate the specific role of medication adherence and barriers to use for psychotropic medications upon SUD relapse, and (4) assess follow-up changes in mental illness severity and medication adherence in dual diagnosis patients enrolled in a substance abuse rehabilitation program.

Methods: The pilot study utilized a mixed methodology. Inclusion criteria included male patients at least 18 years of age who were newly admitted at a 90-day residential rehabilitation program with a self-reported diagnosis of SUD, and either major depressive disorder (MDD), bipolar disorder, generalized anxiety disorder (GAD) or schizophrenia. Patients were evaluated within their first week of treatment and follow-up interviews were conducted at 1 and 2 months. Facility records were accessed to cross-reference patient reported data, using Cohen's kappa coefficient to determine agreement. Patient demographic characteristics, substance abuse characteristics, health-related characteristics, and attitude towards medications stratified by adherence rates and relapse rates utilizing ANOVA and t-tests. Pearson's correlation coefficient was utilized to analyze the relationship between medication adherence and SUD relapse. A multivariable logistic regression model was created to assess the impact of adherence on relapse frequency. Patient and clinical characteristics were stratified according to follow-up interviews completed utilizing ANOVA and t-tests. Lastly, changes in patients' self reported adherence from interview to interview were analyzed using mean difference. SPSS Statistics (IBM Corp; Armonk, NY) was utilized for all analyses, with a two-tailed level of significance at 0.05.

Results: The final sample consisted of 38 patients. The majority of patients were white (n=27, 71.1%), unemployed (n=32, 84.2%), and homeless (n=30, 78.9%). Heroin was the most common primary drug of use (n=19, 50%), followed by alcohol (n=12, 31.6%), and crack cocaine (n=4, 10.5%). The average length of substance use was 20.3 years. Half of the patients (n=19, 50%) had two or more mental illness diagnoses and the most common was the combination of MDD and GAD (n=9, 23.7%), followed by MDD alone (n=7, 18.4%), and bipolar disorder (n=6, 15.8%). Significant agreeance was found between patient self-reported data to facility records for primary substance of use (kappa=0.753, p<.001), mental illness diagnosis (kappa=0.434, p<.001), number of mental illness comorbidities (kappa=0.257, p=0.008), and number of psychotropic medications prescribed (kappa=0.094, p<.001). Patients self-reported less comorbid diagnoses and more previous treatment stays compared to facility records. Patients receiving income prior to admission had higher relapse rates (16.9 vs 8.1, p=0.02). Self-reported relapse rate was negatively correlated with the Modified Morisky Adherence Scale (MMAS-8) intentional score (r= .360, p=.026), MMAS-8 total score was positively correlated with self-reported adherence rates (r=.618, p<.001), the MMAS-8 intentional score (r=.869, p<.001), and the MMAS-8 unintentional score (r=.863, p<.001). MMAS-8 intentional score was positively correlated with MMAS-8 unintentional score (r=.552, p<.001) and self reported adherence rate (r=.613, p<.001). Lastly, the MMAS-8 score was positively correlated with self-reported adherence rate (r=.481, p<.001). For the regression model, MMAS-8 total score was a significant predictor of relapse rate (stand. beta = -.443, CI= 6.37-0.23, p=.048) but the linear combination of the measures included was not significantly related to self-reported relapse rate (F=2.25, adjusted R2 =.145, p=.073). A total of 12 patients (31.6%) fully completed the study, 15 patients (39.5%) only participated in the first follow-up, and 11 patients (28.9%) only participated in the primary interview. DAI-10 total scores were lower in patients who only completed the primary interview vs. patients who completed the entire study (4.0 vs 7.0, p=.044). There was a significant increase in adherence at the first (mean difference=5.7, p<.001) and second (mean difference=6.5, p<.001) follow-ups compared to the primary interview.

Conclusions: The study provided valuable insight into the relationship between psychotropic medication adherence and SUD relapse in patients with dual diagnosis which can be used by healthcare professionals and drug abuse rehabilitation programs.