Developing and Evaluating a Machine-Learning Opioid Prediction & Risk-Stratification E-Platform (DEMONSTRATE)

The proposed study aims to harness advanced natural language processing and longitudinal neural network approaches to build on our previously developed machine-learning prediction algorithms to identify patients at risk for opioid overdose or opioid use disorder. We developed a prediction tool using all-payer electronic health records (EHR), Medicaid claims, and Medicaid claims linked with EHR data from the One Florida Clinical Research Consortium and will translate the risk prediction algorithms into a clinical decision support platform integrated into the EHR system to identify patients at high risk of overdose and opioid use disorder. This innovative and integrated platform will better guide clinical providers and health care systems for improving safety of opioid prescribing in clinical practice, and prevent opioid-associated adverse outcomes.