Is It Possible to Predict How Long a Surgery Will Last?

Alex Macario, MD, MBA


July 14, 2010


The 2 most important determinants of surgical duration are the surgeon and the type of surgical procedure. Averaging historical data for predicting case durations does not increase prediction accuracy as much as one would think. The combination of a wide variety of procedures and a large number of surgeons on staff at most hospitals means that, on average, for half of the cases scheduled in an OR tomorrow, fewer than 5 previous cases of the same primary procedure type and by the same surgeon will have been performed during the preceding year.

If case durations for a surgeon who performs a particular procedure vary significantly, then making good predictions is also intrinsically very difficult regardless of how many cases are collected. Yet another barrier to truth in scheduling is that the statistical distributions of case times are not normal or bell-shaped distributions. This skewness complicates the use of average durations of historical cases because unusually long cases (outliers) have a disproportionately large effect. In the future, the uncertainty associated with predicting case duration will be accepted and better managed. Many hospitals now have a monitor in the waiting area that provides data about when the patient has entered the OR, when the surgery is finishing, and when the patient has left the OR. This provides real-time information to patients and OR staff of the progress of a surgery case.