Association Between Opioid Prescribing Patterns and Abuse in Ophthalmology.

IMPORTANCE

Drug overdoses have become the number 1 cause of mortality in American adults 50 years and younger. Prescription opioid abuse is a growing concern that has garnered widespread attention among policymakers and the general public.

OBJECTIVE

To determine the opioid prescribing patterns among ophthalmologists and elucidate their role in the prescription opioid abuse epidemic.

DESIGN, SETTING, AND PARTICIPANTS: In this observational cohort study, beneficiaries and their physicians were analyzed using 2013 to 2015 Medicare Part D Prescriber Data. The Centers for Medicare and Medicaid Services Medicare Part D Prescriber Public Use Files for 2013, 2014, and 2015 were accessed. Analysis began in June 2017. Data were collected and analyzed regarding the prescribing patterns for opioid drugs (eg, number of prescriptions written including refills, number of days' supply, and prescriber rates) for all participating ophthalmologists.

MAIN OUTCOMES AND MEASURES

The mean number of opioid prescriptions written annually by ophthalmologists; prescriber rates compared with all prescriptions written; and geographic distribution of opioid prescriptions written per ophthalmologist.

RESULTS

In 2013, 4167 of 19 615 ophthalmologists were women (21.2%). Consistently, most ophthalmologists (88%-89%) wrote 10 opioid prescriptions or fewer annually. Approximately 1% (0.94%-1.03%) of ophthalmologists wrote more than 100 prescriptions per year. On average, ophthalmologists wrote 7 opioid prescriptions per year (134 290 written annually by 19 638 physicians, on average) with a mean supply of 5 days. The 6 states with the highest volume of opioid prescriptions written annually per ophthalmologist were located in the southern United States.

CONCLUSIONS AND RELEVANCE

In general, ophthalmologists show discretion in their opioid prescribing patterns. The present opioid abuse epidemic should prompt physicians to consider revisiting their prescribing protocols given the high risk for dependency.