Selection from multiple data sources in a capture-recapture study to estimate the prevalence of opioid use disorder in Massachusetts

Introduction

Current population estimates of opioid use disorder (OUD) can be derived from a capture-recapture study. However, there is little guidance regarding the appropriate number of databases and the dependence structures between them that is needed for accurate estimates using this method. We propose a process to create a parsimonious model from multiple data sources in a capture-recapture study to estimate the population of OUD in Massachusetts (MA).