Methods: Data were drawn from a large administrative data set provided to our team from outpatient substance use treatment agencies in a large, Midwestern U. S. city. Variables in the de-identified data included information about patients (N=2567; including demographics, length of stay, and treatment outcome) and therapists (N=73; demographics). The low percentage of racial minority patients (26%) and therapists (19%) prompted us to recode race into a dichotomous variable (minority and white). We then used the demographic information to compute matching variables based on age, gender, and race.
Patients were considered to have matched ages with their therapist if both were in the same decennial age (e.g., 31-40 or 41-50). Patients were considered to match gender with their therapist if they were the same. Patients were considered to have matched on race if both patient and therapist endorsed the same dichotomous category (minority and white). We then created a count variable of total matches that a patient had with their therapist.
Results: Patients: there were 162 minority females, 641 white females, 514 minority males, and 1071 white males. Therapists: there were 11 minority females, 48 white females, 3 minority males, and 18 white males. It was significantly more common for white patients to match on at least one characteristic with their therapists than it was for minority patients (96.8% vs. 70.8% [χ2 (1, N=2083)=301.55, p=.000].
The rate of successful discharge for minorities was 9.6% with 65.8% dropping out of treatment after 4 weeks. The rate of successful discharge for whites was 28% with 58.1% dropping out after 4 weeks. The rate of successful discharge for females was 18.9% with 56.9% dropping out after 4 weeks. The rate of successful discharge for males was 24.4% with 62.3% dropping out after 4 weeks.
Matching: Each increase in total number of demographic matches for minority patients was associated with a decreased odds of successful treatment discharge (exp(B)=.49, p=.005). Each increase in the total number of demographic matches for white patients was associated with an increased odds of successful treatment discharge (exp(B)=1.21, p=.04).
Conclusions/Implications: Minorities are not succeeding in SUD treatment compared with white patients. They have lower rates of successful discharge and higher dropout rates. Taken together, these findings suggest that demographics matching could improve retention and completion outcomes for minorities as well as non-minorities. As demographic matching is a new idea, additional research is needed in order to address the inequities within our treatment systems and structures.