Editor's Note
In this study, researchers developed service- and surgeon-specific statistical models using linear regression and machine learning to predict case-time duration at a large academic medical center.
Results showed:
- The machine-learning algorithm had the highest predictive capability.
- The surgeon-specific was superior to the service-specific model, with higher accuracies and percentage of cases in the 10% threshold as well as lower percentages of overage and underage.
- The ability to predict cases within 10% improved from 32% using the institutional standard to 39% using the machine-learning, surgeon-specific model.
Machine-learning approaches may improve case duration estimations, which will improve OR scheduling, efficiency, and reduced costs, the researchers say.
