Evaluation of the start of surgical closure as a milestone for forecasting the time remaining to exit the operating room: A retrospective, observational cohort study

Abstract

IntroductionForecasting when patients will exit the operating room (OR) is important for managing add-on cases, coordinating breaks and relief, anticipating postanesthesia care unit admissions, and estimating to-follow surgeons’ start times. If coefficients of variation (CVs) for the interval from closing to OR exit (OR exit interval) were small, a reasonable predictive approach would be to initialize a timer at the start of closing using the mean OR exit interval, then decrement the time remaining each successive minute. However, if the CV were ≥0.32 (typical for many surgical procedure times), this approach would be invalid because actual average times remaining would significantly exceed the predictions. Futility would be expected because the corresponding approach is inaccurate for case duration prediction. MethodsWe retrospectively studied all OR anesthetics (n = 44,254) at an adult academic hospital from October 29, 2017, through January 11, 2022. The primary surgeon, scheduled primary Current Procedural Terminology code, OR entry to closing, and OR exit intervals were retrieved electronically. The CVs of the OR exit intervals were calculated, categorized by: primary CPT and surgeon; primary CPT; American Society of Anesthesiologists CPT; and Clinical Classification Software category. We compared, pairwise, the standard deviations (SD) of the OR entry to closing interval to the OR exit interval, determined if most (>50%) of the CVs for the OR exit interval were ≥0.32, and compared CVs for the OR exit interval among categories. We analyzed categories that were performed ≥30 times. ResultsFor all categories, the standard deviations (SD) of the OR exit intervals were considerably smaller than the paired SDs of OR entry to closing (P < 0.0001). The smallest CVs were from combining primary CPT and surgeon (median 0.4, IQR 0.34 to 0.47, P < 0.0001 vs. other categories). However, using the 244 analyzed combinations of primary CPT and surgeon, the best classification category, 81.3% of the CV (99% CI 74.1% to 87.2%, P < 0.0001) were ≥0.32. In other words, at least 74% of combinations would have greater proportional variability than observed commonly for surgical times. ConclusionsA simple timer started when surgical closure begins, decrementing each subsequent minute, is an invalid approach to estimating the time remaining in cases once closure begins. Rather, previously published methods for predicting the time remaining in ongoing OR cases are appropriate, with time-zero set when closing begins. OR status boards would be updated continually with the revised exit times, calculated as the current time plus the time remaining.