Purpose: Minimally-invasive total knee arthroplasty has generated tremendous recent interest, but there have been reports of complications resulting from these new techniques. However, most studies have less than one year follow-up and are anecdotal in their results. The purpose of this study was to retrospectively evaluate a series of 1,000 consecutive minimally invasive total knee arthroplasties, to describe the clinical and radiographic complications of this technique, to determine whether complication rates changed with experience, and finally to compare complication rates to a match group operated by the same surgeon.

Method: Clinical and radiographic data were reviewed for the first 1,000 consecutive minimally-invasive knees (820 patients) performed by a single surgeon with a minimum 2 year post-operative period. All patients were operated using the mid-vastus approach, with a mean incision length of 10 cm (range, 8 to 13 cm). Nine patients were lost prior to 2 year follow-up, leaving 990 knees (811 patients) in this report with annual follow-up. After determining overall results, data was stratified into 100-knee groups to determine whether complication rates changed over time. Finally, complication rates were compared to a matched group of 50 knees operated by the same surgeon.

Results: There were a total of 45 clinical complications (4.5%), including 20 (2.0%) manipulations under anesthesia, 12 (1.2%) arthroscopies for painful crepitus, 4 (0.4%) component changes, and 3 (0.3%) spacer changes. There were 3 (0.3%) impending radiographic failures. Stratification of results showed a complication rate of 6% over the first 200 knees, and 1% over the subsequent 800 knees. In comparison, the matched group had a complication rate of 4% over mean follow-up of 3 years (range, 2 to 4).

Conclusion: This large series with minimum two year follow-up describes many of the long-term complications of this technique. Additionally, this large series suggests that high volume knee surgeons may require as many as 200 minimally invasive arthroplasties before achieving optimum proficiency with this technique. Nonetheless, comparison to a matched group suggests that even with this learning curve, complication rates need not increase markedly during this period if surgeons remain attentive to the potential complications associated with minimally invasive techniques.