Abstract
BACKGROUND: Robotic-assisted total knee arthroplasty (rTKA) has garnered significant interest for its potential to enhance surgical precision and accuracy. However, the adoption of such systems poses concerns, including longer operative times and learning curves, potentially reducing efficiency. This study aimed to evaluate the learning curve associated with the Robotic Surgical Assistant (ROSA) system for rTKA. METHODS: This retrospective review analyzed the first 75 ROSA rTKA procedures performed by each of 2 fellowship-trained arthroplasty surgeons (150 total procedures) at a high-volume institution. Time stamps within the robotic software were recorded for each case, along with tourniquet time. Statistical analyses included descriptive statistics, t-tests, and multilevel regression. RESULTS: Comparison of each surgeon's first 20 and last 20 cases revealed significant decreases in tourniquet time (61.4-56.7 minutes; P = .0417) and planning time (13.49-6.68 minutes; P = .0078). Landmark femur and tibia times remained stable (P = .6542 and P = .9440). Knee state evaluation time showed a trend of reduction from 9.22 to 7.33 minutes (P = .1335), and resection time from 13.66 to 12.92 minutes (P = .4372). Regression analysis indicated significant reductions in tourniquet time (β = -0.11; P = .0089) and planning time (β = -0.08; P = .0064). CONCLUSIONS: This study demonstrates that execution of ROSA rTKA becomes more efficient over the first 75 cases. The greatest improvement with experience is the time spent on the planning panel, the cognitive portion of the procedure. These data provide surgeons with the confidence that the technical portions of the case are quick to learn and guide industry to focus on teaching effective adjustments on the planning panel.