Center and Surgeon Volume Influence the Revision Rate Following Unicondylar Knee Replacement

Background: Revision rates following unicondylar-knee replacement vary among reporting institutions. Revision rates from institutions involved in the design of these implants and independent single-center series are comparable with those following total knee replacement, suggesting that higher operative volumes and surgical enthusiasm improve revision outcomes. Methods: This registry-based cohort study involved the analysis of 23,400 medial cemented Oxford unicondylar knee replacements for the treatment of osteoarthritis. Total center and surgeon operative volumes were calculated over an eight-year time span since the inception of the registry (April 2003 to December 2010). The revision rate was calculated according to center volume and surgeon volume, each of which was grouped into five categories. The groups were compared with use of life tables, Kaplan-Meier plots, and Cox regression models that adjusted for variations in age, sex, and American Society of Anesthesiologists (ASA) grade among the groups. Results: A total of 919 surgeons and a total of 366 centers performed at least one replacement, with the majority performing a small number of procedures. The revision rate for the centers with the lowest volume (fifty or fewer procedures over the eight-year study period) was 1.62 (95% confidence interval [CI], 1.42 to 1.82) revisions per 100 component years; this was significantly higher than the rate for the centers with the highest volume (more than 400 procedures), which was 1.16 (95% CI, 0.97 to 1.36) revisions per 100 component years. The five-year implant survival rate of 92.3% (95% CI, 91.2% to 93.3%) for the lowest-volume centers was significantly lower than the rate of 94.1% (95% CI, 93.0% to 95.2%) for the highest-volume centers. Similarly, the revision rate for the surgeons with the lowest volume (twenty-five or fewer procedures), 2.16 (95% CI, 1.91 to 2.41) revisions per 100 component years, was significantly higher than that for the surgeons with the highest volume (more than 200 procedures), 0.80 (95% CI, 0.62 to 0.98) revisions per 100 component years. The five-year survival rate of 90.1% (95% CI, 88.8% to 91.3%) for the lowest-volume surgeons was also significantly lower than the rate of 96.0% (95% CI, 95.0% to 97.0%) for the highest-volume surgeons. When center and surgeon volume were considered simultaneously, the hazard of revision was greater for lower-volume surgeons at lower-volume centers compared with higher-volume surgeons at higher-volume centers (hazard ratio = 1.87 [95% CI, 1.58 to 2.22], p < 0.001). Conclusions: High-volume centers and surgeons specializing in such procedures had superior results following unicondylar knee replacement compared with their low-volume counterparts. These results suggest that centers and surgeons should undertake a minimum of thirteen such procedures per year to achieve results comparable with the high-volume operators.