Variability of the times remaining in surgical cases and the importance of knowing when closure has started

Abstract

BackgroundWhen a patient enters the operating room, the projected time to room exit is the time of room entry plus the estimated case duration. Subsequent estimates are provided based on a probability distribution for time remaining in the case, conditional on the interval from room entry. There have been several recent studies investigating and showing good accuracy of different suitable methods. However, it is unknown whether and why these methods are useful compared with the historical data and original surgeon/ scheduler estimate known before the case started. We hypothesized that knowing when closure has started, patient has been extubated, etc., gives substantive incremental information. MethodsOur retrospective cohort study included all 23,343 surgical cases at a large hospital over 4.2 years wherein each case's surgeon scheduled the procedure ≥29 times. The sample size was n = 244 combinations of scheduled Current Procedural Terminology code and surgeon. ResultsVariability in the logarithmic scale (i.e., proportional variability in the time scale) increased progressively as cases advanced. Specifically, the 90th percentiles of times to operating room exit were ≈40% longer than 50th percentiles (medians) at the start of cases versus 142% longer at the time of the start of closure. These ratios were less at starts of cases than at closures for 99.2% of n = 244 combinations (P <0.0001). The 5th percentiles were ≈32% briefer than 50th percentiles at the start of cases versus 85% briefer at the time of the start of closure. These ratios were less at starts of cases than at closures for 100% of combinations (P <0.0001). ConclusionsThere is considerable information value in knowing when key milestones have occurred near the end of cases that managers can use for quality decision-making (e.g., when closure has started). Organizations should document these milestones in real-time.