Email Updates
A machine learning-based clinical predictive tool to identify patients at high risk of medication errors.
Abdo A, Gallay L, Vallecillo T, et al. A machine learning-based clinical predictive tool to identify patients at high risk of medication errors. Sci Rep. 2024;14(1):32022. doi:10.1038/s41598-024-83631-w.
Pharmacist-led medication reconciliation is resource intensive; targeting patients most at risk for medication errors can make the best use of this limited resource. In this study, a machine learning (ML) tool was developed to identify hospitalized patients at risk of medication errors. As standard practice in the study hospital, a pharmacist selected a small random sample of patients for medication reconciliation using an existing tool, with approximately 20% having at least one medication error. Using the ML tool, 45% of identified patients had at least one medication error, outperforming the existing tool.
