Incidence rates of and risk factors for opioid overdose in new users of prescription opioids among US Medicaid enrollees: A cohort study.

Abstract

To measure incidence rates of and risk factors for opioid overdose among new users of prescription opioids in the Medicaid population. A cohort study using Medicaid claims from four states (1999-2012) among adults continuously enrolled in Medicaid for ≥3 years free of opioid prescriptions and opioid overdose before cohort entry. Exposure and outcome of interest were prescription opioid use and apparent incident opioid overdose identified in inpatient and outpatient claims (sensitivity ≈ 97%; positive predictive value ≈ 87%), respectively. Among new prescription opioid users (1 336 140 persons; 246 466 person-years), the overall opioid overdose incidence rate per 100 000 person-years was 247.1 (95% confidence interval [CI], 227.5-266.7), with 251.0 (CI, 188.6-313.5) in 2002 and 225.5 (CI, 142.0-309.0) in 2012. A lower hazard for opioid overdose was seen for age 65-80 years (adjusted hazard ratio [HR], 0.50; CI, 0.37-0.66) and 80-100 years (0.35; 0.23-0.52) vs 18-35 years; females (0.79; 0.67-0.93) vs males; and other/unknown race/ethnicity (0.71; 0.54-0.93) vs whites. A higher hazard was seen for initial opioid dose in morphine milligram equivalents (MMEs), 50-100 MME/day (1.52; 1.24-1.86) and >100 MME/day (1.98; 1.55-2.53), vs <50 MME/day; prior diagnosis of substance use disorders (2.30; 1.91-2.79) or mental health conditions (1.75; 1.47-2.08); and prior prescriptions for benzodiazepines (1.43; 1.13-1.81). In Medicaid enrollees in four study states during 2002 to 2012, opioid overdose incidence rate per 100 000 person-years among apparent new users of prescription opioids was 247.1, with 251.0 in 2002 and 225.5 in 2012. Younger ages, white race/ethnicity, higher MME opioid daily doses, prior substance use disorders, mental health conditions, and benzodiazepine prescriptions were associated with a higher risk of opioid overdose incidence.