Medication treatment for opioid use disorder (MOUD) is the most effective, evidence-based treatment for OUD. However, people receiving MOUD have unfulfilled treatment information needs and knowledge gaps regarding their treatment. They frequently engage on social media with peers instead of turning to their healthcare providers to meet these needs for various reasons, including stigma, lack of access, trust, or resources. Analyzing such large-scale social media data can yield insights about treatment information needs regarding MOUD and improve access to, effectiveness, and adherence to OUD treatment.
This pilot project aims to determine the feasibility of using innovative mixed-method natural language processing (NLP) methodologies on large-scale Reddit data to characterize the self-reported treatment information needs regarding buprenorphine. In this pilot, we will collect and annotate a large, novel Reddit dataset from self-identified Reddit users who are receiving OUD treatment and have been prescribed buprenorphine or a buprenorphine combination. Buprenorphine offers the advantage of broader accessibility since it can be prescribed by a range of healthcare providers who have a waiver to prescribe this medication. We will characterize Buprenorphine-related treatment information needs using a mixed-method study. We will also develop and evaluate novel NLP classifiers and inference models to automatically characterize Buprenorphine-related treatment information needs from large-scale Reddit data. The pilot will result in new evidence, and actionable insights about OUD treatment information needs.