Abstract
IMPORTANCE: Increasing access to naloxone (an opioid antagonist that can reverse overdose) could slow the US opioid epidemic. Prior studies suggest cost sharing may be a barrier to dispensing of naloxone prescriptions, but these studies were limited by their cross-sectional designs and use of databases that do not capture prescriptions that are not filled (abandoned). OBJECTIVE: To evaluate the association between cost sharing and naloxone prescription abandonment (nondispensing of naloxone prescriptions). DESIGN, SETTING, AND PARTICIPANTS: This cross-sectional, regression discontinuity analysis exploited the fact that deductibles typically reset at the beginning of the year in commercial and Medicare plans. The included data were derived from the 2020-2021 IQVIA Formulary Impact Analyzer (a pharmacy transactions database that represents 63% of prescriptions at US pharmacies). The analysis included claims for naloxone nasal spray among commercially insured patients and Medicare patients that occurred during the 60 days before January 1, 2021, through 59 days after January 1, 2021. EXPOSURE: Cost sharing, which is defined as the amount patients would have to pay to fill prescriptions. MAIN OUTCOMES AND MEASURES: Local linear regression models were used to assess for abrupt changes in cost sharing and the probability of prescription abandonment on January 1, 2021. To estimate the association between cost sharing and prescription abandonment, a fuzzy regression discontinuity analysis was conducted. RESULTS: These analyses included naloxone claims for 71 306 commercially insured patients and 101 706 Medicare patients (40 019 [56.1%] and 61 410 [60.4%], respectively, were female). The commercially insured patients and Medicare patients accounted for 73 311 and 106 076 naloxone claims, respectively. On January 1, 2021, the mean cost sharing per claim increased by $15.0 (95% CI, $13.8-$16.2) for commercially insured patients and increased by $12.3 (95% CI, $10.9-$13.6) for Medicare patients and the probability of abandonment increased by 4.7 (95% CI, 3.2-6.2) percentage points and 2.8 (95% CI, 1.6-4.1) percentage points, respectively. The results from the fuzzy regression discontinuity analysis suggest a decision by commercial and Medicare plans to increase naloxone cost sharing by $10 would be associated with percentage-point increases of 3.1 (95% CI, 2.2-4.1) and 2.3 (95% CI, 1.4-3.2), respectively, in the probability of abandonment. CONCLUSIONS: The elimination of cost sharing might be associated with increased naloxone dispensing to commercially insured and Medicare patients.