OBJECTIVES

To evaluate whether the risk of opioid-related overdose increases with the number of days with overlapping supply of opioid and benzodiazepines in a dose-response fashion.

METHODS

Using 2013-2014 Medicare claims data, we identified beneficiaries who filled at least one prescription for an opioid in 2014. We extracted their prescriptions for benzodiazepines in 2014 and categorized them into three groups: those who never filled prescriptions for benzodiazepines (“opioid-only users”, n=173,228), those who filled prescriptions for benzodiazepines but had no overlapping supply (“not concurrent opioid and benzodiazepine users”, n=15,022), and those with overlapping supply of opioids and benzodiazepines (“concurrent opioid and benzodiazepine users”, n=76,708). We divided the last group into 4 subgroups: those with 1-15 days of overlapping supply (n=21,795), 15-62 days (n=16,594), 62-234 days (n=19,213) and ≥235 days (n=18,493). We collected their claims for opioid-related overdose and constructed Cox proportional hazard models to compare the risk of overdose between treatment groups, controlling for demographics, insurance, clinical characteristics, and number of opioid and benzodiazepine providers.

RESULTS :

Concurrent use of opioids and benzodiazepines increased the risk of overdose in a dose-response fashion. Compared to opioid-only users, the hazard ratios (HR) for overdose were 1.55(95%CI, 1.06-2.26) for concurrent opioid and benzodiazepine users with 1-15 days of overlap, 1.71(95%CI, 1.21-2.42) for 16-62 days, 1.81(95%CI, 1.34-2.43) for 63-234 days, and 1.84(95%CI, 1.37-2.47) for ≥235 days. The risk of overdose was not significantly higher for not concurrent opioid and benzodiazepine users than for opioid-only users (HR 1.58; 95% CI, 0.99-2.51). The risk of overdose increased by 13% for each additional opioid or benzodiazepine provider.

CONCLUSIONS

The risk of opioid-related overdose increases with the number of days with overlapping supply of opioids and benzodiazepines in a dose-response fashion. Fragmentation of care, measured as the number of opioid and benzodiazepine providers, was associated with increased risk of overdose.