Drug availability adjustments in population-based studies of prescription opioid abuse

Purpose: Population-based prescription opioid abuse studies in which one drug is compared to another, or drugs are compared across time, often account for the availability of those drugs in the community. The objective of this investigation is to assess consistency in the relative abuse ratios (RARs) across different approaches for adjusting for drug availability. Methods: For the years 2004 through 2010, RARs for each of four prescription opioids (hydrocodone, oxycodone, hydromorphone, and morphine) were calculated using negative binomial regression. Measures of abuse (outcome) were misuse/abuse-related emergency department visits obtained from the Drug Abuse Warning Network. Measures of drug availability (offsets) were drug utilization estimates obtained from IMS Health. Separate regression models were run using each of five measures of drug utilization: unique patients (URDD), prescriptions dispensed (RX), tablets dispensed (TD), kilograms (KGs) sold, and morphine-equivalents (MEs) of kilograms sold. These results were compared for consistency. Results: Aside from oxycodone-combination products, across molecules, RARs adjusted by RXs, TDs, and URDDs were generally similar to each other while RARs adjusted by KGs and MEs were different. For example, compared to hydrocodone, oxycodone had statistically significantly increased RARs of 3.6 (95%CI: 2.0–6.5), 3.5 (95%CI: 1.9–6.4), and 2.7 (95%CI: 1.5–5.0) when adjusted by URDDs, RXs, and TDs, respectively, but not when adjusted by KGs or MEs. Conclusions: Different drug utilization adjustment approaches may yield inconsistent RAR estimates in population-based prescription opioid abuse analyses. Published 2016. This article is a U.S. Government work and is in the public domain in the USA.