Abstract
OBJECTIVE: This study examines the differential predictive value of baseline characteristics of clients being treated for an alcohol problem with respect to the development of an opioid use disorder (OUD) or opioid overdose (OD) within 1 year, between 1 and 4 years, and beyond 4 years after treatment. METHOD: A cohort of 87,172 patients treated for an alcohol use problem within state treatment centers was examined. We extracted the first OUD/OD diagnosis event within 1 year of, between 1 and 4 years of, and more than 4 years after the patient's first admission to the Office of Addiction Services and Supports. We calculated odds ratios for all predictors and control variables with respect to OUD/OD events and compared the predictive values of these variables for the different periods. RESULTS: Both sociodemographic and clinical factors predicted an OUD/OD overall and in most specific follow-up periods. Sociodemographic factors were more strongly associated with OUD/OD during follow-ups beyond 4 years, perhaps because of the increasing availability of opioids over time. Mental health and alcohol use severity factors were more strongly associated with OUD in the 1-year and 1-to 4-year periods, suggesting a rapid progression to OUD/OD. CONCLUSIONS: Both sociodemographic and clinical factors were predictive of a diagnosis of OUD/OD within a brief period; however, they were only predictive of approximately 40% of those who would develop OUD/OD within any specific period. These findings highlight the need for a more formal assessment of opioid use at treatment entry, and for the implementation of harm reduction measures throughout treatment.