Methods: This study used self-report data from 119 patients (58% male, 63% African American, mean age = 52.6) receiving MMT at an urban, university-affiliated clinic. A latent class analysis was conducted to identify underlying patient subtypes. The latent class indicators included ecological factors (histories of family member substance misuse and/or living with someone who misused substances), emotional vulnerability (loneliness, depression, stress, childhood trauma, interpersonal violence, community violence, opioid-related coping motives, opioid-related shame), and antisocial-impulsivist traits (trait impulsivity, illegal activities, meaning-in-life impairments). A multinomial logistic regression analysis was performed to examine the associations of opioid-related history and covariates (race, gender, education, age, age of opioid onset) with latent class membership. Analyses were conducted using Mplus v8.2.
Results: The fit indices (AIC, BIC, Adjusted BIC, Entropy, BLRT) suggested a four-class solution as the best fitting model. The “opioid-related only” class (45.6%) was characterized by lower scores on all indicators, with the exception of opioid-related shame, compared with other classes. The “emotionally-vulnerable” class (29.7%) had greater loneliness, depression, and stress than the first and the third classes. The “emotionally-vulnerable/trauma-exposed” class (10.6%) had greater loneliness, depression, and stress than the first class and a greater degree of childhood trauma, interpersonal and community violence than the first and the second classes. The “pervasive-severity” class (14.1%) showed higher scores on nearly all respective indicators. A multinomial logistic regression analysis revealed that race, gender, education, age, prior MMT history, injection opioid use, prescription opioid use precipitating a relapse, and risk for co-occurring addictive disorders (cocaine, alcohol, gambling) were associated with class membership.
Conclusions/Implications: Our findings provide the first subtyping analysis within an MMT patient population, which may prove helpful in improving treatment outcomes for OUD patients. Our study demonstrates preliminary support for the Pathways Model in identifying MMT patient subtypes. Our first and fourth classes overlap with Pathways 1 and 3, while our second and third classes are similar to Pathway 2. These findings are strengthened by the use of a treatment-seeking sample and an examination grounded in a credible framework. Future research should extend our findings with larger samples, followed by investigations of whether MMT patient subtypes differ in treatment response. In conclusion, subtyping has the potential to improve treatment for MMT patients, thereby offering public health value to address the opioid epidemic.