Accuracy of Orthopedic Surgeons Versus Electronic Health Record in Prediction of Operating Room Times

Despite integration of machine learning in electronic health record (EHR) systems, accurate prediction of case time continues to present variable outcomes. This study compared the accuracy of the EHR versus surgeons in predicting operating room (OR) times.

A retrospective chart review examining orthopedic case times was conducted at a level 1 trauma center. OR durations were calculated, and the difference between predicted and actual times were compared between the EHR system and surgeons. A prediction within 30% of the actual case duration was considered correct. T test and analysis of variance (ANOVA) were used to compare prediction accuracy.

The EHR-predicted OR time demonstrated a 77.9% accuracy, whereas surgeon-prediction demonstrated a 48.2% accuracy. EHR-predicted OR time resulted in a sum discrepancy of a 1,007-minute underestimation with an absolute difference of 9,941 minutes; surgeon-predicted OR time resulted in a sum discrepancy of 13,014-minute underestimation with an absolute difference of 15,850. minutes. ANOVA and t tests between surgeon-predicted case time and EHR-predicted time stratified by subspecialty demonstrated significant differences between spine-joint and spine-trauma. T tests comparing differences between single-procedure cases and multi-procedure cases for both EHR- and surgeon-predicted times demonstrated significantly increased discrepancies in multi-procedure cases.

Although performance varied, the EHR appears to more accurately predict operating time compared to surgeons. Notably, surgeons tend to underestimate operating time. These findings support the use of the EHR when scheduling cases to improve efficiency and maximize OR use.