PROJECT SUMMARY/ABSTRACT Opioid use disorder (OUD) is a public health emergency impacting the lives of an estimated 5.6 million individuals in the United States. OUD is a particular concern for vulnerable populations (e.g., pregnant women, criminal- legal system involved populations, etc.), where the barriers to evidence-based care and the economic burden remain significant. Randomized controlled trials (RCTs) are considered the gold standard for identifying causal treatment effects and is the primary scientific tool to inform clinical decisions. RCT research designs are also growing in use for conducting robust health economic studies to ensure that evidence-based treatments can be cost-effective and sustainable—reaching the broadest number of individuals with OUD while efficiently using scarce healthcare resources. As such, health economic RCT research designs have become a critical tool in identifying sustainable and cost-effective solutions to prevent and mitigate the opioid epidemic. The NIH HEAL initiative has championed such research by funding economic evaluations alongside clinical trials in the Healing Communities Study, NIDA's Clinical Trials Network (CTN) and Justice Community Opioid Innovation Network (JCOIN), and The HEAL Prevention Cooperative (HPC), among others. Traditional RCTs have demonstrated the effectiveness of medications for OUD (MOUD) but have limitations in addressing the unique needs of these populations. Real-world data and evidence (RWE) research designs that use observational data from healthcare claims, electronic medical records, and other sources have been used to answer critical questions in medicine that are difficult or impossible to implement in equivalent RCTs. The 21st Century CURES Act reaffirmed the use of RWE and provided greater research flexibility of data sources for the FDA drug approval process. Despite this, significant concerns remain about the reliability of RWE due to its observational nature and potential for confounding bias. To address these concerns, RCTs have been used to inform RWE research designs (e.g., the popular “target RCT” framework) and augment RWE findings. The alternative and unconventional approach of integrating RWE to inform inconclusive RCTs to support robust and causal conclusions that can inform clinical practice has been systematically ignored but represents a potential opportunity to reduce research waste and produce more reliable findings to inform clinical decision-making and improve outcomes for at-risk populations with OUD. This project will develop novel econometric methods and unified framework for integrating RWE in the analysis of inconclusive RCTs by adapting existing econometric and biostatistical techniques into comparative economic and effectiveness assessments of OUD treatments conducted alongside RCTs. The project will evaluate multiple maximum likelihood estimation (MLE) approaches combined with propensity score- based causal inference methodologies to achieve this goal. This novel approach, which is the reverse of traditional practice, offers a unique and highly risky "out-of-the-box" solution to address emerging barriers in generating evidence for health interventions targeting vulnerable populations affected by the opioid epidemic.

PROJECT NARRATIVE The personal, public-health, and economic consequences associated with insufficiently treated opioid use disorder (OUD) are enormous, especially for vulnerable populations who face significant barriers to evidence- based care. While traditional randomized controlled trials (RCTs) have demonstrated the effectiveness of medications for OUD, their limitations in generating evidence that addresses the unique needs of these populations call for the development and application of novel methodologies to improve health economic RCT research designs. The goal of this project is to develop an empirical framework that integrates real-world evidence (RWE) into RCT data analysis using innovative maximum likelihood estimation (MLE) methods to produce more comprehensive, robust, and actionable evidence that can directly inform clinical decision-making and improve patient care for vulnerable individuals affected by the opioid epidemic.