NIH Reporter #5R01EB035028-04

Personalized Risk Prediction for Prevention and Early Detection of Postoperative Failure to Rescue

This award funds the development and validation of machine learning-based clinical decision support tools to predict and prevent postoperative complications using multi-modal data sources in acute care settings.

Description

The goal of this proposal is to apply machine learning approaches to multiple data sources including electronic health record data, high-fidelity physiological waveform data, and genomic data that have never been used together in the acute care setting to predict postoperative major complications and decrease failure-to-rescue after surgery. These inputs will be used in simulated real-time bedside management to iteratively design a prototype clinical decision support tool and evaluate its use in simulation upon use of postoperative continuous remote monitoring and early warning systems, accuracy and time to correct diagnosis, accuracy of intervention, and time-to-intervention. The feasibility and acceptability of this clinical decision support tool will then be assessed in a small-scale prospective, longitudinal pilot evaluation in sequential 10-week, 13-week, 10-week phases to help design a future, large-scale clinical trial.

View original record