RA-TKA with MAKO system follows structured learning curve offering rapid surgeon adaptation: study

Robotic-assisted total knee arthroplasty (RA-TKA) is increasingly used to improve implant positioning, soft-tissue balance, and procedural reproducibility. Yet, little is known about how different components of the operation independently contribute to the overall learning curve.

Ferdinando Granata et al conducted a study to characterize the learning curve of MAKO-assisted TKA by separately evaluating these components and assessing their differential impact on operative workflow.

A retrospective observational study included 92 consecutive patients who underwent image-based RA-TKA (MAKO, Stryker) for primary knee osteoarthritis. All procedures were performed by a single experienced arthroplasty sur¬geon with no prior robotic or computer-assisted surgery background. Cumulative sum (CUSUM) analysis with piecewise linear regression was applied to total surgical time, pin placement time, and composite robotic workflow time to identify inflection points and define learning curve phases.

The key findings of the study were:

• Piecewise regression of the CUSUM plot for total surgical time revealed two breakpoints at cases 11 and 51, defining three phases of the learning curve: (1) initial learning (cases 1–11), (2) competence (cases 12–51), and (3) opti¬mized performance (cases 52–90).

• Mean surgical time was 68.9 ± 20.1 min, stabilizing around 65 min after 50 cases.

• Along with total surgical time, the initial learning phase ended around cases 10–11 for both robotic workflow and pin place¬ment.

• However, subsequent performance patterns differed: pin placement reached optimized performance by case 52 (mean 8.4 ± 4.3 min), whereas robotic workflow time improved more gradually, without clear stabilization until the end of the series (mean 37.8 ± 10.8 min).

Conclusions RA-TKA with the MAKO system follows a structured learning curve with early achievement of proficiency after 11 cases. Total surgical time and pin placement reached optimized performance by mid-series, whereas robotic work¬flow tasks required a longer consolidation period, likely influenced by patient-specific anatomical variability. These findings support RA-TKA as a safe and effective tool, offering rapid surgeon adaptation.

Level of evidence IV

For further details on the article refer to:

Learning curve in image-based robotic assisted total knee arthroplasty: a MAKO-robot experience

Ferdinando Granata et al

European Journal of Orthopaedic Surgery & Traumatology (2026) 36:217

https://doi.org/10.1007/s00590-026-04803-0