BACKGROUND: The burden of osteoarthritis (OA) in multiple joints is high. For patients with bilateral knee OA there is no clear recommendation when to time the second surgery. The purpose of this study was therefore to compare revision and mortality rate in bilateral unicondylar and bicondylar knee arthroplasties after different strategies of surgical timing in bilateral knee OA from the German Arthroplasty Registry (EPRD). METHODS: Data from the German Arthroplasty Registry (EPRD) was used. Since 2012 a total of 15,154 patients had bilateral knee arthroplasty within one year. Patellofemoral arthroplasties and constraint total knee arthroplasties (TKA) were excluded. 1,144 TKA and 682 unicondylar arthroplasties (UKA) were simultaneously performed, 772 TKA and 292 UKA between 1 and 90 days (short interval) and 24,496 TKA and 2,922 UKA between 91 and 365 days (intermediate interval). Revision and mortality rates were analyzed up to 7 years after surgery. Cox regression was performed to evaluate the influence of different patient characteristics on these outcomes. RESULTS: The highest cumulative revision rate for any of the bilateral TKA was found for simultaneous surgery with 3.4% (95% CI 2.1-5.5). Lower risk for revision was seen in two-staged surgery in short interval (HR 0.42
95% CI 0.20-0.90) and intermediate interval (HR 0.58
95% CI 0.39-0.85). The cumulative one year mortality rate for TKA was comparable in all three groups with 0.8% for simultaneous TKA, 1.3% for short interval two-staged and 0.7% for intermediate interval. In UKA there were no differences between the groups regarding cumulative revision rate and mortality rate. CONCLUSION: TKA should be performed simultaneously in selected patients only, the two-staged procedure demonstrated lower revision risks. For UKA we found no differences in timing, simultaneous surgery seems to be a safe option. TRIAL REGISTRATION: Clinical trial number not applicable. LEVEL OF EVIDENCE: III.