Two stage analysis of learning curves on laparoscopic study of surgeons

Abstract

The performance of many repeated tasks change with experience, with improvement being most rapid at first and then tails off over time until a steady state is reached. The term ‘learning curve’ is often used as shorthand to describe this phenomenon. The objectives of this study were to evaluate individual learning curves for surgeons performing laparoscopic activities for two different tasks important for a surgeon (the nuts and the ropes) that are timed and also to evaluate the possibility of using the information from psychological test and gender in predicting surgical performance. Moreover, it was of interest to know if the measures of performance from the learning curve, psychological test and gender can be used for predicting the results during the 'real life test’. Two-stage analysis was implemented to achieve the stated objectives. In the first stage two approaches were used: the simple method which summarises individual measures by initial measurement, the difference between the first and last measurement or just the final measurement, and the complex analysis. For the complex method, the exponential curve was used to derive the measures of performance of a surgeon from which proxies for learning such as initial level, length of learning, final skill level, rate of learning and time taken to reach a plateau were derived. In the second stage, the proxies for learning were regressed against several covariates. When we look at the training ropes, females performed better than males according to most occasions of measures of performance. Surgeons with higher special cognitive ability were found to have lower starting level. It is also noted that surgeons with larger motivation perform better as compared to those with lower motivation.