Leveraging Prescription Drug Monitoring Programs to Evaluate Opioid Policy Interventions

Abstract

Research ObjectiveA variety of state and federal policies have been enacted to curtail prescription opioid overdoses, including implementation of statewide opioid prescribing guidelines and rescheduling of hydrocodone from DEA Schedule III to Schedule II. We aimed to use data from the Ohio Prescription Drug Monitoring Program (PDMP) to evaluate the effects of these policies.Study DesignRetrospective study of all opioid prescriptions dispensed in the state of Ohio from 2010 to 2017. We first determined the number and characteristics of opioid prescriptions stratified by specialty. We then used an interrupted time series analysis to determine the effect on opioid prescribing after (a) the release of statewide opioid prescribing guidelines in 2012 and (b) the federal rescheduling of hydrocodone in 2014.Population StudiedThe state of Ohio. Prescriptions for the nine most common pill form opioid pain relievers dispensed in Ohio were included. The PDMP includes all prescriptions dispensed to individuals in the state, regardless of payer.Principal FindingsThe number and characteristics of opioid prescriptions and trends over time varied greatly between different prescriber specialties. The mean number of pills/prescription and morphine content of the prescriptions were highest for physical medicine/rehabilitation (91.2 pills, 1532 mg, n = 1 680 579), anesthesiology/pain (89.3 pills, 1484 mg, n = 3 261 449), hematology/oncology (88.2 pills, 1534 mg, n = 516,596), and neurology (84.4 pills, 1230 mg, n = 573 389). Overall, family medicine (21.8%) and internal medicine (17.6%) prescribed the most opioids. The implementation of emergency department opioid prescribing guidelines was associated with a 12% decrease (95% CI −17.7% to −6.3%) in the level of statewide total prescriptions/month and an additional decrease of 0.9% (95% CI −1.1% to −0.7%) relative to the trend before the guidelines. Rescheduling of hydrocodone was associated with a level change of −26 358 (95% CI −36 700 to −16 016) prescriptions (−5.8%) and an additional decrease of −1568 (95% CI −2296 to −839) prescriptions/month. Rescheduling was associated with a temporary substitution effect (increase) of codeine and with a decrease in the prescribing of tramadol.ConclusionsThe characteristics of opioid prescriptions vary widely by specialty and suggest the need for specialty‐specific guidance. Opioid guidelines targeting the emergency department were associated with a decrease in prescribing beyond the secular trend. The rescheduling of hydrocodone was associated with a decrease in hydrocodone prescriptions beyond the secular trend, but also appeared to affect the prescribing of other opioids as well.Implications for Policy or PracticePDMPs are state‐based prescribing databases that can be used to influence prescriber decision‐making at the point of care. However, their powerful data can also be used to describe prescribing behaviors overall and evaluate the effectiveness of implemented opioid‐related policies.