Medication Adherence and Substance Abuse Predict 18-Month Recidivismamong Mental Health Jail Diversion Program ClientsElizabeth N. Burris1, Evan M. Lowder1, Candalyn B. Rade1, Sarah L. Desmarais1, & Richard A. Van Dorn21North Carolina State University, 2RTI International
Introduction
The Present Study
Methods
Discussion
Results
Acknowledgments
This project was funded by the Bristol Myers Squibb Foundation. We thank the 11thJudicial District of Florida Criminal Mental Health Project for its cooperation and support.
Prior studies have established an association between medicationadherence and recidivism among adults with mental illnesses. Forexample, Young and colleagues (1999) found patients withschizophrenia who failed to adhere to medication were at increasedrisk for recidivism. In another study of adults with serious mentalillnesses, monthly medication possession after hospitalization wasassociated with decreased rates of arrest (Van Dorn, Desmarais,Petrila, Hayes & Singh, 2013). Similarly, multiple studies haveestablished an association between co-occurring substance use andrecidivism among justice-involved adults with mental illnesses,including prison inmates (Baillargeon et al., 2010), jail inmates (Wilsonet al., 2011), and probationers (Castillo & Alarid, 2011).
However, there is relatively limited research on the relationshipbetween substance use and medication adherence with recidivismamong justice-involved adults with mental illness. Farabee and Shen(2004) found substance use was associated with increased recidivismrates in a population of parolees with serious mental illnesses whoused cocaine and were prescribed antipsychotic medication.Additionally, there was a significant interaction between substance useand medication adherence on recidivism; however, medicationadherence alone was not significantly associated with decreasedrecidivism (Farabee and Shen, 2004).
To our knowledge, no study to date has examined the associationsof substance use and medication adherence among mental healthdiversion clients with recidivism as the primary outcome.
The present study explored associations between medicationadherence, substance use, and recidivism at6-,12-, and 18-monthfollow-up in a sample of mental health jail diversion clients. Specifically,we examined associations between: 1) medication adherence andrecidivism; 2) substance use and recidivism; 3) medication adherenceand substance use; 4) medication adherence and recidivism aftercontrolling for substance use, and 5) substance use and recidivismafter controlling for medication adherence.
Study Context
Data were collected as part of a larger study evaluating behavioralhealth services delivered to participants in mental health jail diversionprograms in a metropolitan center of a southeastern state.
Sample
Participants were 95 mental health jail diversion clients (81.1%male). Half were Hispanic/Latino (54.7%) and 44.2% identified asAfrican American. The sample included felony (47.4%) andmisdemeanor (52.6%) diversion clients. Participants had an average of6.69 (SD = 5.42) jail bookings in the past three years and 16.07 (SD =14.99) lifetime bookings. Primary diagnoses included schizophrenia(35.8%), psychosis NOS (25.3%), and bipolar disorder (20.0%).
Measures
Medication Adherence. Medication adherence was measured bythe self-reported number of times in the three months prior to baselineparticipants had taken prescribed medication for behavioral healthproblems. Participants who never missed taking their medication ormissed only one or two times were coded as adherent. All otherresponses were coded as non-adherent.
Substance Use. Substance use was defined as whetherparticipants had a current substance use diagnosis at baseline or werecurrently using at baseline as indicated by an Addiction Severity Index(ASI; McLellan et al., 1992) alcohol composite score of ≥0.17 or drugcomposite score of ≥0.16 (Rikoon et al. 2006).
Recidivism. Self-reported and official recidivism records were usedto measure any arrest, number of arrests,any jail, and daysincarcerated at each follow-up period.
Procedures
Semi-structured interviews were conducted by trained researchassistants at baseline 6-,12-, and 18-month follow-up. Participantdemographics were collected at baseline; recidivism was collected atall follow-up periods.
Findings revealed medication adherence and substance usepredicted arrest and incarceration over the 18-month follow-upperiod in a sample of mental health jail diversion program clients.Consistent with research in other populations (Farabee & Shen,2004; Wilson et al., 2011), findings suggest that substance use isassociated with recidivism in a mental health jail diversion sample.Additionally, and inconsistent with prior research in other populations(Young et al.,1999; Van Dorn et al. 2013), medication adherencewas not associated with recidivism in bivariate analyses; however,when controlling for substance use, baseline medication adherencepredicted recidivism at follow-up.
Conclusions are limited by the self-report nature of the data, lowbase rates, and sample homogeneity. Implications include the needfor improved access to services to promote medication adherenceand substance use treatment which may decrease substance useand increase medication adherence rates. Further research isneeded to clarify the role of substance use and medicationadherence in predicting recidivism among mental health jaildiversion samples over longer follow-up periods and with additionalcovariates, including psychiatric symptom severity.
Descriptive Statistics
Overall, few participants wereincarcerated or arrested (see Table1).The majority of participantsreported being medicationadherent, and approximately halfhad co-occurring substance use.
Table 1. Descriptive Statistics
Substance Use
Logistic regression analyses
revealed substance use was
significantly associated with any jail (OR = 1.77; p = .058) and daysincarcerated (OR = 0.44; p = .032) at 18-month follow-up. Specifically,participants who were using substances at baseline were likely tospend more time in jail at 18-month follow-up, compared to participantswho were not using substances at baseline. However, substance useat baseline was not significantly associated with recidivism acrossother follow-up periods.
Medication Adherence
Logistic regression analyses
revealed medication adherence
alone was not significantlyassociated with any arrests
(ps ≥.119), number of arrests
(ps ≥ .179), any jail (ps ≥ .111), ordays incarcerated (ps ≥ .217)across all follow-up periods.
Substance Use, Medication Adherence, and RecidivismA series of logistic regression analyses were conducted withsubstance use and medication adherence as predictors of recidivism(see Table 2). Results showed that participants who were medicationadherent at baseline, compared to those who were not, were morelikely to be arrested or incarcerated during the 18-month follow-upperiod, while controlling for substance use. Additionally, whencontrolling for medication adherence, participants who were usingsubstances at baseline, compared to those who were not, were morelikely to be arrested or incarcerated during the 18-month follow-upperiod. Logistic regression models testing associations of medicationadherence and substance use at baseline with recidivism were notsignificant for 6-month and 12-month follow-up.
Table 2. Summary of Logistic Regression Results
Notes. *p < .05 **p < .01 ***p < .001. OR = Odds Ratio; CI = confidence interval.
Medication Adherence and Substance Use
Chi-squared analyses revealed a marginally significant difference inthe likelihood of substance use as a function of medication adherence(OR = 0.69; p = .060). Participants who were not medication adherentat baseline were more likely to report substance use at baseline,compared to participants who were medication adherent.