From a theoritical point of vue, experience is an important factor in obtaining a satisfactory result in arthroplasty surgery. We wished to determine whether standard posterior stabilized total knee replacement (PS TKA) performed by young surgeons (Group A) increased rates of mortality and complications compared with PS TKA performed by senior surgeons (Group B) using the same model of arthroplasty. Between 1990 and 1995, 195 patients underwent 250 total knee arthroplasties in the same orthopaedic university department of the hospital by two senior surgeons (100 TKA; Group B) and 12 younger surgeons without senior assistance (150 TKA; Group A). The procedures were undertaken consecutively and the implant was always the same.

There were no significant pre-operative differences between the groups in terms of age, gender, height, weight, body mass index, diagnosis, comorbidity and duration of follow-up, which was more than 15 years in both groups. Pre-and postoperative assessments were made according to the system of the Knee Society. The preoperative and postoperative deformities were measured on weight-bearing radiographs of the whole limb (hip-knee-ankle angle).

The mean Knee Society knee and functional scores were not significantly different (p = 0.125) pre-operatively: 37.5 points (16 to 53) and 15 points (0 to 20) respectively in the Group B, and 36.0 points (10 to 58) and 17 points (0 to 30) respectively in the Group A; and at final follow-up (p = 0.145): 91 points (42 to 100) and 82 points (25 to 100) respectively in the Group B, and 89 points (58 to 100) and 84 points (35 to 100) respectively in the Group A.

The rate of survival at ten years, with revision as the endpoint for failure, was 96% (95% CI, 93 to 100) in both groups; at fifteen years 91% (95% CI, 85 to 97) in group B, and 92% (95% CI, 90 to 94) in group A. There were no significant differences in revision rates in Group B or Group A (p = 0.735). In the Group B group, 1 knee (0.6%) revised for osteolysis, 1 (0.6%) deep infection, 1 (0.6%) aseptic loosening, and 1 (0.6%) for fracture. In the Group A, 1 knee (1%) revised for deep infection, none for osteolysis and 2 (2%) for aseptic loosening, and one for dislocation.

In this series surgeons in the early stages of their careers achieved the same results as seniors; so this study is very reassuring for patients undergoing surgery in a university hospital. Our study has however limitations. All the patients received cemented TKA in a large-volume centre specialising in joint reconstruction. It is not a randomised-controlled trial; but it would be impossible to perform one. Would really patients sign up to a study where they would be randomised between an inexperienced and experienced surgeon? What is difficult to assess is how the ‘learning curve’ of one implant affects the ‘learning curve’ for a different implant.