NCJ Number
253687
Journal
Drug and Alcohol Dependence Volume: 201 Dated: 2019 Pages: 127-133
Date Published
August 2019
Length
7 pages
Annotation
This projects objective was to develop a predictive risk model for nonfatal opioid overdose in a statewide population of buprenorphine patients.
Abstract
Predicting which individuals who are prescribed buprenorphine for opioid use disorder are most likely to experience an overdose can help target interventions to prevent relapse and subsequent consequences. The current project used Maryland prescription drug monitoring data from 2015 to identify risk factors for nonfatal opioid overdoses identified in hospital discharge records in 2016. A predictive risk model was developed for prospective nonfatal opioid overdoses among 25,487 buprenorphine patients. A series of models was estimated that included demographics plus opioid, buprenorphine, and benzodiazepine prescription variables. Logistic regression was applied to generate performance measures. About 3.24 percent of the study cohort had 1 plus nonfatal opioid overdoses. In the model with all predictors, odds of nonfatal overdoses among buprenorphine patients were higher among males (OR = 1.39, 95 percent CI:1.21 to 1.62) and those with more buprenorphine pharmacies (OR equals 1.19, 95 percent CI:1.11 to 1.28), 1+ buprenorphine prescription paid by Medicaid (OR equals 1.21, 95 percent equals CI:1.02 to 1.48), Medicare (OR equals 1.93, 95 percent CI:1.63 to 2.43), or a commercial plan (OR equals 1.98, 95 percent CI:1.30 to 2.89), 1+ opioid prescription paid by Medicare (OR equals 1.30, 95% CI:1.03 to 1.68), and more benzodiazepine prescriptions (OR equals 1.04, 95% CI:1.02 to 1.05). The odds were lower among those with longer days of buprenorphine (OR equals 0.64, 95 percent CI:0.60-0.69) or opioid (OR equals 0.79, 95 percent CI:0.65-0.95) supply. The model had moderate predictive ability (c-statistic 0.69). Several modifiable risk factors, such as length of buprenorphine treatment, may be targets for interventions to improve clinical care and reduce harms. This model could be implemented with common prescription-related information and enable payers and clinical systems to better target overdose risk-reduction interventions, such as naloxone distribution. (publisher abstract modified)