Use of Causal Framework to Evaluate Effect of Abuse Deterrent Properties of Extended-Release Oxycodone on Tampering in a Real-World Settings

Purpose: To assess whether exposure to an extended-release (ER) oxycodone with abuse deterrent properties (ADF) reduced tampering of oxycodone in a real-world, postmarket setting to address the thinking behind Category 4 labeling by the FDA. Methods: Data from an observational cross-sectional study of the general adult population (2022) was used under a causal framework to estimate the confounding-adjusted odds of tampering oxycodone after exposure to two types of ADF ER oxycodone. The tampering behaviors of those who used only single entity immediate-release (SE-IR) oxycodone was used as a comparison. The tampering outcome was defined as use by snorting, smoking, or injecting any oxycodone (ER or SE-IR). A directed acyclic graph was used to identify covariates. Average treatment effect among the treated was estimated using inverse propensity score weighting combined with survey weights in a regression. Results: In 2022, 0.14% and 3.0% among the general population reported using the two ER oxycodone groups, while 2.4% used SE-IR oxycodone. Propensity score analyses with both comparators (common support > 98%) balanced demographic, health, and drug use covariates. After adjustment for selection and confounding bias, among those who used ER oxycodone group 1 the odds ratio of tampering with any form of oxycodone was elevated but not statistically significant (2.25; 95% CI: 0.94, 5.39). The odds ratio of tampering by users of ER oxycodone group 2 was significantly elevated (1.90; 95% CI: 1.08, 3.19). Conclusions: Tampering of ER oxycodone products by individuals was rare. We found evidence suggestive of elevated odds of tampering behaviors with use of an ADF ER oxycodone.