HIP ARTHROSCOPY: THE LEARNING CURVE

Abstract

Aim: The purpose of this study is to explore the experience of a consultant orthopaedic surgeon, and to quantitatively describe the learning curve for hip arthroscopy.

Introduction: Arthroscopic surgery in orthopaedics is a well established procedure for both diagnostic and therapeutic purposes. Unlike many other joint arthroscopies, hip arthroscopy has been delayed in its development. It was first pioneered by Burman in 1931, who under-took a study on cadavers, stating that ‘it is manifestly impossible to insert a needle between the head of the femur and the acetabulum’. Over several decades, this technique has developed considerably, but still remains a technically demanding and difficult procedure. The learning curve for hip arthroscopy has not previously been objectively quantified.

Method: We prospectively reviewed the first 100 hip arthroscopies performed in the supine position between 1999 and 2004. Surgery was performed by a single experienced hip and knee consultant orthopaedic surgeon (FH). We assessed the operative time (traction time), surgeon comfort, patient satisfaction at 6 months and operative complications. This was analysed for consecutive blocks of 10 cases. Results of the first 10 and the remaining 90 cases, subsequently the first 20 and remaining 80 cases, and finally the first 30 and remaining 70 cases were compared for a difference.

Results: The mean traction time was 55 minutes (range: 36–94 minutes). Mean surgeon comfort was 73% (range: 52–89%). 49% of patients reported an excellent outcome at 6 months follow – up. Only 8% of patients reported an unsatisfactory outcome. The main complications noted were chondral damage (6 cases) and perineal injuries (4 cases). There was a remarkable decrease in complications from the first 30 cases compared to the remaining 70 operations. 5 cases of chondral damage was noted in the first 30 cases, compared to 1 (1.4%) in the remaining 70 cases. The number of perineal injuries was noted to decrease from 3 cases in the first 30 operations to 1 (1.4%) in the subsequent 70 operations.

There is an overall decrease in operative time over the 100 cases, representing a gradual learning process throughout. However, the fall from an average time of 75 minutes for the first 30 cases, to the average operative time of 30 minutes for the remaining 70 cases, is a significant learning process (40% fall in operative time). We thus, believe the learning curve to be 30 operations.

Conclusion: We have demonstrated that there is a considerable fall in operative time when comparing the first 30 cases with the remaining 70 cases. This quantitative decrease is indicative of a rapid learning curve. This is further suggested by the remarkable fall in complications during this learning phase.