Prediction of Opioid-Related Overdose and Suicide Events Using Administrative Healthcare Data
Ralph Ward, PhD
Seminar date: 10/18/2023
Description: Veterans suffer disproportionate health impacts from the opioid epidemic, including overdose, suicide, and death. Prediction models based on electronic medical record data can be powerful tools for identifying patients at greatest risk of such outcomes. In this study investigators developed an improved prediction model that built on existing work by incorporated advanced statistical methods, additional data sources and new predictor variables in a longitudinal setting. The proposed model achieved an area under the ROC curve (AUC) of 84% and sensitivity of 71%. The model performed particularly well in identifying patients at risk for suicide related events, where 72% of actual events were accurately predicted among patients with the 100,000 highest risk scores. Some predictors in the proposed model have markedly different associations with overdose and suicide risks, which will allow clinicians to better target interventions to the most relevant risks.