Geographic Variation of High-Risk Opioid Use and Risk of Overdose Among Disabled Medicare Beneficiaries in the US from 2011 to 2015

Abstract

had > 2 opioid prescriptions, we identified beneficiaries with high-dose use (> 120 daily morphine milligram equivalents for ≥ 90 consecutive days) and multiple providers (≥ 4 prescribers and ≥ 4 pharmacies) each year; and concurrent benzodiazepine use (≥ 30 cumulative days) from 2013-2015 when Part D began coverage for benzodiazepines. We obtained adjusted annual rates of high-risk measures across 306 hospital referral regions (HRRs) using multivariable logistic regression and examined the association between these measures and overdose risk(claimbased) in the subsequent year using Cox proportional regression, adjusting for sociodemographic, health status, and access-to-care factors. ReSultS: Adjusted annual rates of high-dose use (~9%), having multiple providers(~5%), and concurrent benzodiazepine use (~34%) remained stable over five years. In 2015, the ratio of 75th-to-25th percentile rates of high-risk measures across HRRs were 1.80 for high-dose use, 1.87 for having multiple providers, and 1.33 for concurrent benzodiazepine use. The top 3 HRRs with the highest rate of: high-dose use were Sarasota, FL(17.2%), Sun City, AZ(17.2%) and Clearwater, FL(16.9%); multiple providers were Slidell, LA(14.0%), Muskegon, MI(12.5%), and Bryan, TX(12.0%); and concurrent benzodiazepine use were Dearborn, MI(58.0%), Miami, FL(55.4%), and Spartanburg, SC(55.1%). These measures were associated with subsequent overdose risk for high-dose (hazard ratio [HR]= 2.19, 95%CI= 1.86-2.57); multiple providers (HR= 1.58,; and concurrent benzodiazepine use (HR= 1.82, 95%CI= 1.58-2.10). ConCluSionS: High-risk opioid use measures were associated with overdose risk among disabled Medicare beneficiaries. Areas and individuals with prevalent high-risk opioid use may benefit from targeted interventions (e.g.,lock-in programs) to prevent overdose. MH4