Prediction of Operative Mortality for Patients Undergoing Cardiac Surgical Procedures Without Established Risk Scores

The authors of this study used four supervised machine learning models to analyze a data set from the Society of Thoracic Surgeons (STS) of Massachusetts General Hospital to predict and classify operative mortality of procedures which did not already receive a risk score from the STS. The researchers found that the machine learning models used on procedures which did not receive a risk score could accurately predict their operative mortality rate. This suggests that these models can be used on a variety of cardiac procedures which do not receive a risk score, as they make up 23% of cardiac procedures nationwide, and may be used on local institution level data to account for institution-specific practices.

Citation for the Full-text Article

Ong CS, Reinertsen E, Sun H, Moonsamy P, Mohan N, Funamoto M, Kaneko T, Shekar PS, Schena S, Lawton JS, D’Alessandro DA, Westover MB, Aguirre AD, Sundt TM. Prediction of operative mortality for patients undergoing cardiac surgical procedures without established risk scores. The Journal of Thoracic and Cardiovascular Surgery. September 14, 2021. Article in Press. doi. 10.1016/j.jtcvs.2021.09.010.