[Computer-assisted total-knee arthroplasty. Comparison of two successive systems. Learning curve].

PURPOSE OF THE STUDY

The increasing popularity of total-knee arthroplasty has led to many technical improvements both in the field of prosthesis design and implanted material and instrumentation. The recent advent of computer-assisted techniques is the fruit of a search for more precision for the bone cuts and better ligament balance. The purpose of the present study was to demonstrate how easy it is to use navigation systems by examining the difficulties encountered by one operator with navigation experience when the material was changed.

MATERIAL AND METHODS

The first 30 total-knee arthroplasties implanted with a new navigation system were investigated. Elements specifically related to navigation difficulties were studied. The series was composed of 16 women and 14 men, mean age 65.9 years at the time of surgery (range, 43 to 84). Mean BMI was 30.66 (range, 23.05 to 39.54). All patients were reviewed by the operator using a standard X-ray protocol. Mean follow-up was six months. The 30 arthroplasties were consecutive, with no exclusions excepting revision procedures. Primary or post-traumatic degeneration was the main reason for surgery. This series was compared with two prior series of 30 prostheses each, implanted with a different navigation system. The first 30 and last 30 implantations using the previous navigation system were thus compared in terms of operative time and precision (comparison of postoperative alignment and implant position). The study focused on difficulties encountered when using the new system, on intra- and postoperative complications and on assessment of implant position.

RESULTS

All procedures were totally performed with the navigation system, no interruptions. Operative time was lengthened by an average of 18 min (range, 0 to 45 min). There were no complications specifically related to the navigation system. The position of the implants was assessed in the frontal and sagittal plane on the plain X-rays and with a goniometer. Computed tomography was used to assess femoral component rotation. The overall alignment of the lower limb was within the "ideal" range of +/-3 degrees in 97% (average 0.1 degrees varus). The position of the femoral implant and the tibial plate was correct in the frontal and sagittal planes and no internal rotation of the femoral piece was noted on the 27 ct scan studies (mean 1.9 degrees external rotation). Implant accuracy was equivalent to that observed in the series of the last 30 implants using the prior navigation system. The learning curve was shorter.

DISCUSSION

This small series demonstrated the absence of major problems with the new navigation system. The length of the learning curve was acceptable. This study demonstrated that prior experience with navigation is beneficial because the learning curve with the new system was shorter and the accuracy of implantation was equivalent to that achieved with the prior system. Widespread use of computer-assisted surgery should enable continued improvement in ancillary systems in the upcoming years, particularly concerning rotatory position of the femoral implants, which is still a problem. Cost containment will also be a necessary goal.