• Elisabeth U Dexter, Franklin Dexter, Danielle Masursky, and Kimberly A Kasprowicz.
• Department of Anesthesia, University of Iowa, Iowa City, IA 52242, USA.
• Anesth. Analg. 2010 Apr 1;110(4):1164-8.

BackgroundSurgeon estimates of case durations are important for operating room (OR) management decision making because many cases are rare combinations of procedures with few or no historical data. Thoracic and spine surgeons updated their scheduled OR times on the day of surgery just before the "time out" in the OR.MethodsAll elective (scheduled) general thoracic (n = 39) and spine surgery (n = 48) cases at 1 hospital were studied over 3-month and 1.5-month periods, respectively.ResultsAmong cases with a change in predicted duration, most changes were made based on updates to the surgical or anesthetic procedures (thoracic 85%, spine 86%). For thoracic surgery, there was overall no significant median reduction in absolute prediction error (median 0 minutes, 95% confidence interval [CI] 0-0 minutes). Among the 37% of cases with changed predicted durations, there was a significant reduction in absolute error (median 38 minutes, 95% CI >7.5 minutes). For spine surgery, there was overall no reduction in the absolute error (median 0 minutes, 95% CI 0-0 minutes). Among the 29% of cases with changed predicted durations, absolute error was no worse, but not significantly better (point estimate of median reduction 34 minutes, 95% CI >0 minutes). Secondary observations made were no effect of updates on bias, frequent rounding of scheduled durations to the nearest half hour, and increased predictive error caused by decisions that reduced expected overutilized OR time.ConclusionsA systematic program of routinely and/or always asking for updated case duration predictions will not substantively improve OR management decision making. However, when a change in surgical approach, surgical procedure, or anesthetic procedure is identified (e.g., at the intraoperative briefing before case start), the updated estimate of case duration should be used, because such updates are not worse and often better than original estimates.

11 keywords...

hide…