I am concerned by the conclusions of the article titled "The Learning Curve for Laparoscopic Colorectal Surgery" by Bennett et al.1 The authors reported 1194 laparoscopic colectomies recorded in a national registry. They concluded that 40 cases are required to achieve good results, and they suggested that credentialing committees may benefit from studies like theirs.
First, the accuracy of the registry used by the authors is questionable. In addition, results for the group of 40 to 49 cases and the number of patients in each group are missing.
Second, despite the article's title mentioning "learning curve" and its conclusion that "40 cases are required to achieve good outcome," the authors failed to show a learning curve. The authors could have analyzed data of those surgeons who had performed laparoscopic colectomies in more than 40 cases and compared each individual surgeon's results for the groups of 1 to 39 cases