Multidimensional analysis of learning curves in laparoscopic sigmoid resection: eight-year results.

PURPOSE: The number of operations to reach a plateau phase in colorectal laparoscopic surgery is still under debate. There are few publications reporting on multidimensional assessment of the learning curve, including operation time and complication and conversion rates. The purpose of this study was to define a multidimensional learning curve for sigmoid resection performed by two surgeons with experience in laparoscopic surgery. METHODS: Between 1993 and 2001 from a total of 715 laparoscopic colorectal procedures, two surgeons performed self-educated 338 sigmoid resections. Demographic data, indications for surgery, procedure performed, operation time, frequency and kind of complications, conversion rate, and days to discharge were recorded. The moving average method was used to demonstrate the change of the operation time. The cumulative sum technique was used to chart the changes in the conversion and complication rates. RESULTS: Surgeon A performed 199 and Surgeon B 139 sigmoid resections. The operation time decreased from 225 minutes to 169 minutes after approximately 90 operations for Surgeon A and from 270 minutes to 223 minutes after 110 operations for Surgeon B. Based on a decline in intraoperative complications and conversion rate, the steady state was reached after approximately 70 to 80 interventions for both surgeons. CONCLUSIONS: The assessment of a learning curve should not be limited to measurement of a decrease in operation time but should also include the conversion and complication rates. The cumulative sum technique and moving average method as proposed in this study seem appropriate to evaluate the learning curve in this clinical domain. Our findings might be especially useful for those planning training programs in laparoscopic surgery.