Predictive pre-resection robotic ligament balancing in total knee arthroplasty is accurate across surgeons and patient factors

Abstract

New robotic total knee arthroplasty (TKA) technology enables force-controlled pre-resection joint gap assessment. Its accuracy across surgeons and patient factors remains unknown. This study evaluated how accurately this technology achieved planned joint gaps and whether accuracy was influenced by surgeon, preoperative deformity, sex, BMI, and age. Six hundred seventy-two consecutive robotic TKAs from five surgeons were retrospectively reviewed. A robotic ligament tensioner applied 70–100 N medially and laterally before bony resections. Mean (ME) and mean absolute (MAE) predicted gap errors (final minus planned) were calculated at 10°, 45°, and 90° flexion. Linear regression evaluated associations between error and preoperative alignment and sex. A sub-analysis included BMI and age. Wilcoxon rank-sum tests compared errors between surgeons and deformity groups. Overall, MAE was 1.1 mm medially and 1.3 mm laterally. The largest ME occurred laterally in flexion in varus knees (-0.7 mm; tighter than plan), and laterally in extension in valgus knees (0.6 mm; looser than plan). Extension errors were similar in flexion-contracture and sagittal-neutral knees (ME within ± 0.1 mm), while hyper-extended knees were 0.5 mm looser than planned. Increasing BMI was associated with lateral flexion gap tightening (-0.59 mm per 10 kg/m 2 increase, p = 0.018). ME was within 1.4 mm across all surgeons and flexion angles. Robotic TKA with force-controlled pre-resection gap assessment accurately achieved targeted joint gaps across a large patient cohort. While surgeon and patient factors influenced joint gap error, these effects were small. Awareness of these variations may help optimize implant planning.