Predictors of Drug Court Graduation and Implications: The Importance of Employment, Education, and Medication Assisted Treatment

Background and Purpose:  Drug courts have been in an important part of the criminal justice system since the first program began in 1989 in Dade County (Miami), Florida.  In nearly three decades, the number of drug courts has increased to 3,057 and these programs are now operating in all 50 United States.  This study contributes to the existing literature by evaluating a drug court in Indiana to identify who is most likely to graduate the program.  Understanding the variables that impact graduate rates is especially important because research has consistently demonstrated that participants who graduate drug court are less likely to recidivate than those who are terminated from the program.  This study has one research question: Who is most likely to graduate drug court? 

Methods:  To answer the research question, data were collected on all participants (n = 248) who started the drug court from 2010 to 2015 and either graduated or were terminated from the program.  The primary analysis was then conducted using hierarchical binary logistic regression.  The first model examined how individual characteristics, including gender, race/ethnicity, age, education level, and employment status predicted graduation.  The second model examined how the addition of other clinically relevant variables, including primary drug of addiction (opiate versus non-opiate), mental health (previous mental health diagnosis or none), number of days between arrest and admission to drug court, and criminal history (previous misdemeanor or felony case prior to the current charge or none) increased the ability to predict graduation.  Finally, whether or not there was a drug court violation (dilute drug test, positive drug test indicating new use, missed treatment or drug court appointment, or new arrest) in the first 30 days in the program was added to the third model to determine whether that variable improved the ability to predict graduation. 

Findings:  Three variables emerged as significant predictors of graduation.  First, participants who were employed or were students at the time of admission were nearly 2.5 times more likely to graduate than participants who were not (Exp(B) = 2.47, Wald χ² = 8.88, p ≤ .01).  Second, participants who were using opiates as their primary drug of choice were over 80% less likely to graduate than participants who were using non-opiates as their primary drug of choice (Exp(B) = .17, Wald χ² = 16.09, p ≤.001).  Third, participants who had violations in the first 30 days of the program were nearly 50% less likely to graduate than participants who did not violate in the first 30 days (Exp(B) = .51, Wald χ² = 5.05, p ≤ .05).  The final Nagelkerke R² was .21.

Conclusion and Implications:  Offering medication-assisted treatment (MAT), such as methadone, Suboxone, or Vivitrol, to participants who have an opiate use disorder may improve graduation rates for this population.  Additionally, graduation rates may also improve by offering more resources to assist participants in gaining and maintaining employment or schooling, and this seems to be especially important within the first month of the program.