Social Network Characteristics as Predictors of Opioid Use Disorder Treatment Utilization among People Who Use Drugs with Hepatitis C Infections in Baltimore, MD

BACKGROUND: This study investigated social network characteristics that affect opioid use disorder (OUD) treatment utilization among people who use drugs (PWUDs). METHODS: Between December 2016 and April 2019, participants were recruited in Baltimore, Maryland. Eligibility criteria included being over 18 years old, having a positive antibody test result for HCV, and engaging in risky health behaviors, such as injection drug use in the past 6 months, unprotected sex in the past 3 months, and same-sex behaviors in the past 90 days. The primary outcome was the utilization of different types of OUD treatment (i.e., medication therapies, outpatient treatments, and residential/meeting/self-help). The bivariate analysis was conducted to test the associations between treatment utilization and self-reported individual, behavioral, and network characteristics. The logistic regression model was constructed to investigate network characteristics that predict treatment utilization. RESULTS: Of 334 participants, 84% received some type of OUD treatment. The adjusted logistic regression models suggested that total network size, support network, kin network, and networks currently in drug treatment were significantly positively associated with OUD treatment. Networks tied to healthcare utilization, such as networks helping participants get medical care for HCV, presented positive associations with treatment. Having partners living in the same household showed negative associations with OUD treatment. CONCLUSIONS: Our findings suggest the different composition of support and drug network associated with treatment status. The results indicate that networks may differ by treatment status and can inform network-focused interventions.