Robotic-assisted reverse shoulder arthroplasty achieves operative time neutrality after an initial learning period

Robotic assistance has recently been introduced for reverse shoulder arthroplasty (RSA) with the goal of improving the accuracy and consistency of implant positioning, but the additional workflow steps required for its use may prolong operative time. Whether operative time returns to a conventional benchmark after an initial learning period remains uncertain. This study sought to characterize the operative time learning curve for robotic-assisted RSA using the Mako robotic system (Stryker, Kalamazoo, MI).

We conducted a retrospective observational study of 30 consecutive elective robotic-assisted primary RSA cases performed by a single shoulder fellowship-trained surgeon using the Mako robotic system at an academic ambulatory surgery centre between October 2025 and February 2026. The operative time benchmark was defined a priori as all consecutive conventional primary RSA cases performed at the same surgery centre from its opening in July 2025 through October 2025 (n = 16). Operative time was defined as incision start to incision closure. Learning curve behaviour among robotic cases was assessed using linear regression of operative time on sequential robotic case number. Robotic cases were also grouped into three prespecified 10-case blocks (1-10, 11-20, 21-30).

Mean operative time for conventional RSA was 74.9 min (95% CI, 67.3-82.5; range, 55-105). Across robotic-assisted cases, mean operative time was 88.6 min (95% CI, 79.8-97.4; range, 55-170). Operative time decreased with increasing robotic experience (- 1.85 min per case; p = 0.00010; R² = 0.42), and the fitted regression reached the conventional benchmark mean at approximately robotic case 23. Mean operative time declined across prespecified adoption phases from 105.5 min (cases 1-10; 95% CI, 89.3-121.7), to 92.6 min (cases 11-20; 95% CI, 79.4-105.8), and to 67.8 min (cases 21-30; 95% CI, 62.6-73.0; p = 0.0034). The final 10 robotic cases had a shorter mean operative time than the conventional benchmark (67.8 vs 74.9 min) but this difference was not statistically significant (p = 0.37).

In a single-surgeon ambulatory surgery centre series, robotic-assisted RSA showed a clear learning curve and achieved operative time neutrality relative to conventional RSA after approximately two dozen cases. These findings support the feasibility of integrating robotic workflows into RSA without a persistent operative time penalty after early adoption, and provide practical expectations for surgeons and institutions planning implementation.