Robotic-Arm-Assisted Versus Manual Total Knee Arthroplasty: A Comparative Cohort Study of Gait and Postural Outcomes.

Background/objectives: Evidence on gait and postural recovery after robotically assisted total knee arthroplasty (raTKA), particularly with the ROSA system, remains limited. This study compared early gait, postural, functional, and patient-reported outcomes (PROMs) between ROSA raTKA and manual TKA (mTKA), with PROMs also assessed at final follow-up. Methods: This comparative cohort study included primary TKA patients treated by a single senior surgeon using the same implant and alignment strategy. Patients underwent either ROSA raTKA or mTKA. At three months, a senior physiotherapist assessed quadricep and tibialis anterior maximum voluntary isometric strength (MVIS), centre-of-mass (CoM) kinematics, lower-limb weight distribution, timed-up-and-go (TUG), range of motion (ROM), KOOS Pain, activities of daily living (ADL), and quality of life (QoL). The same KOOS domains were compared at final follow-up. Results: Seventy primary TKAs were included: 46 raTKAs and 24 mTKAs. No intraoperative complications occurred. Groups were comparable for age, BMI, sex, grip strength, and preoperative KOOS. At three months, no significant differences were found in quadricep MVIS (p = 0.257), tibialis anterior MVIS (p = 0.327), CoM kinematics (p = 0.066), weight distribution (p = 0.189), TUG (p = 0.599), ROM (p = 0.165), or KOOS domains. At final follow-up, KOOS-ADL (p = 0.041) and QoL (p = 0.032) were better in the raTKA group, but after Holm-Bonferroni correction, they were no longer significant (QoL, p = 0.384; ADL, p = 0.451). Conclusions: ROSA raTKA showed comparable early gait and postural recovery to mTKA. The marginal differences in KOOS domains are exploratory, as they were no longer significant after multiple-comparisons correction.