Evidence on the learning curve associated with robotic-arm-assisted total knee arthroplasty (ra-TKA) is scarce and mostly based on operative time. Thus, the objective of this study was to assess a surgeon's learning experience based on accuracy to reach planned limb alignment and its impact on surgical-characteristics, limb-alignment, and perioperative-outcomes. A retrospective chart review was conducted on a consecutive series of 204 primary ra-TKAs (patients), performed by a single surgeon in a single institution (3/7/2018-to-6/18/2019). Cumulative summation control sequential analysis was used for the assessment of the learning curve using accuracy of reaching the planned limb alignment establishing that surgeries had an initial-learning-phase, followed by a second-consolidation-phase. Baseline demographics, operative/tourniquet times, prosthesis type, and limb alignment were compared between these two phases. Length of stay, discharge disposition, complications, reoperation/readmission (90 days), and total morphine equivalents (TMEs) prescribed were compared between phases. Independent sample t-tests, and chi-squared analyses were performed. ra-TKA demonstrated a learning curve of 110 cases for reaching planned limb alignment (p = 0.012). Robotic experience resulted in significantly more proportion of knees in neutral-axis postoperatively (p = 0.035) and significant reduction in TMEs prescribed (p = 0.04). The mean operative and tourniquet time were found to be significantly lower in second-phase versus the first-phase (p for both < 0.0001). ra-TKA has a significant learning curve in clinical practice. A surgeon can reach the planned limb alignment with increased accuracy over time (110-cases). Progressive robotic learning and associated operative time efficiency can lead to significantly lower opioid consumption in patients undergoing TKA.

No abstract