Accuracy of Robotic Surgical Assistant (ROSA®) Surgical Plan in Total Knee Arthroplasty

Robotic assistance in total knee arthroplasty (TKA) promises better outcomes through data-supported, personalized alignment. The primary goal of this study was to assess the accuracy of surgical plans delivered by the Robotic Surgical Assistant (ROSA®) Knee System. The secondary goal was to evaluate the variability in knee balancing and implant alignment and to assess the effect of ROSA® on surgery duration and complication rates as compared to the conventional TKA cohort.  Materials: Inclusion criteria were primary TKA indicated by osteoarthritis (OA), age of 40-90 years, body mass index (BMI) of 18.5-50 kg/m², and American Society of Anesthesiologists (ASA) I-III. Surgery was performed by surgeons with at least 10 ROSA® surgeries during the study duration. The primary study parameters were accuracy in femoral and tibial resection levels, assessed by Bland-Altman analysis, implant sizes, and final knee alignment.

The accuracy of medial and lateral distal femoral resection levels was high (mean differences between the plan and validation, -0.50 ± 0.91 mm and -0.59 ± 0.96 mm, respectively). Accuracy of the medial and lateral tibia was less with greater variability (-0.44 ± 2.28 mm and 2.50 ± 3.13 mm, respectively). The femoral component size was accurately predicted in 75% of cases, with overestimation occurring in 20% of cases. Surgical steps were repeated in 59% of femoral cuts and 61% of tibial cuts. Surgical duration was significantly different between groups. ROSA TKA surgery was 18 minutes longer and was prolonged by five minutes per repeated surgical step. The hip-knee-ankle (HKA) axis was 1.6°±1.9° varus post-operatively. Complication rates were not significantly different from ROSA®-assisted TKA as compared to conventional TKA.  Discussion: ROSA®-assisted TKA planning demonstrated a high degree of accuracy for the femur, but a lesser degree for the tibia. Predicting implant sizes was highly accurate. With ROSA®-assistance, alignment differed from mechanical alignment. Future studies should provide a higher level of evidence and investigate the relation between anatomical outcomes and clinical outcomes.