Methodological Advances in Evaluating Abuse Deterrent Opioid Analgesics

This is a joint proposal between the University of North Carolina (UNC) and the University of Kentucky (UK) that aims to advance the methodology and epidemiological tools for observational studies and post-marketing surveillance of abuse-deterrent formulations (ADFs) of opioids. The research studies proposed in this project are in line with the Food and Drug Administration’s agenda on gathering data and testing hypotheses on the role of ADF in meaningful reduction in opioid abuse and related outcomes. The UNC will be the main recipient of the award and the UK will receive a sub award. The project will have two co-principal investigators, one from each university. The study will address important challenges in postmarketing surveillance and research on ADFs, including but not limited to sampling and selection bias, exposure and outcome misclassification, selection of appropriate comparators (i.e. control group drugs with therapeutically equivalent substitution of prescribed ADFs), operationalization of numerator data sources, and measures for utilization adjustment. Longitudinal statistical models will be used to evaluate the relationship between opioid utilization and opioid abuse outcomes in different populations and using different data sources while isolating the secular trends related to changes in laws and policies (e.g., prescription drug monitoring programs, opioid prescribing guidelines, insurance coverage, allowed substitution with therapeutically equivalent non-ADFs), availability of illicit drugs (e.g., heroin and illicit fentanyl), etc. The models will test choices of different and multiple comparators informed by the results of surveys among controlled substance prescribers and pharmacists on choices in the ADF’s prescribing and substituting, as well as focus groups and serial cross-section interviews among people who tamper with ADFs. The intent of this project is to produce open source tools that can be deployed by other researchers. Each research aim will generate computer code, practice guides, webinars, conference presentations, and peer-reviewed publications. Articles will be published in open access journals, programming code will be deposited in online repositories (e.g., GitHub), and instructional material will be posted on a public-facing project-specific website in a timely manner, with Creative Commons licensing to allow sharing.