Robot-assisted total knee arthroplasty: a limited learning curve for ligament balancing

Robotic-assisted total knee arthroplasty (RA-TKA) aims to improve implant positioning and soft tissue balance, but few studies have evaluated the learning curve (LC) for intraoperative ligament balancing.

We retrospectively analyzed 234 posterior-stabilized RA-TKAs performed by three surgeons with varying prior experience using the MAKO system. The primary endpoint was the learning curve duration for mastering ligament balancing, classified as perfectly balanced, stable, or unstable based on medial and lateral joint line openings in extension and flexion. Learning curves for ligament balancing and operative time were assessed using LC-CUSUM analysis.

Overall, 79.8% of TKAs were perfectly balanced, 19.1% stable, and 1.1% unstable. The LC for overall ligament balancing ranged from 5 to 9 cases depending on the surgeon. Operative time normalized after 12 to 22 cases, shorter for surgeons with prior navigation experience. Complications were minimal, with no early infections and few mechanical issues.

Optimal ligament balancing with RA-TKA is achieved rapidly and remains stable, with operative times comparable to conventional techniques after the LC. Prior experience with navigation-assisted surgery shortens the LC. These findings support the reproducibility and safety of robotic ligament balancing, though further studies are needed to confirm long-term clinical benefits.