Impact of Case Numbers on the 5-Year Survival Rate of Unicondylar Knee Replacements in Germany

Abstract

Reported survival rates of unicondylar knee arthroplasty (UKA) vary considerably. The influences of patient characteristics and the type of implant have already been examined. This analysis investigated the influence of hospital volume on 5-year-survival rate, using administrative claims data of Germany's largest health insurance provider. We analysed administrative claims data for 20,946 UKAs covered by the German local healthcare funds (Allgemeine Ortskrankenkasse, AOK) between 2006 and 2012. Survival rates were estimated using Kaplan-Meier analysis. The influence of hospital case numbers on 5-year survival was analysed by means of multivariable Cox regression adjusted for patient characteristics. We estimated hazard ratios (HR) with 95% confidence intervals for five hospital volume categories: < 12 cases, 13 - 24 cases, 25 - 52 cases, 53 - 104 cases, > 104 cases (per hospital and year). The overall 5-year Kaplan-Meier survival rate was 87.8% (95%-CI: 87.3 - 88.3%). This increased with hospital volume (< 12 cases: 84.1% vs. > 104 cases: 93.2%). The analysis identified low hospital volume as an independent risk factor for surgical revision (< 12 cases: HR = 2.13 [95%-CI 1.83 - 2.48]; 13 - 24 cases: HR = 1.94 [95%-CI: 1.67 - 2.25]; 25 - 52 cases: HR = 1.66 [95%-CI: 1.41 - 1.96]; 53 - 104 cases: HR = 1.51 [95%-CI: 1.28 - 1.77]; > 104 cases: reference category). Our analysis revealed a significant relationship between hospital case numbers and 5-year survival rate, which increases with hospital volume. The risk of surgical revision within 5 years in hospitals with fewer than 25 UKAs per year is approximately twice as high as in hospitals with more than 104 cases.