# Real-World Outcomes of Robotic Total Knee Arthroplasty: Five Years’ Experience in a Non-Academic Center

## TL;DR

This study shows that robotic-assisted knee surgery works well in non-academic hospitals, with accurate planning and good patient outcomes over five years.

## Contribution

Demonstrates successful long-term implementation of robotic TKA in a non-academic hospital with high planning accuracy and functional outcomes.

## Key findings

- Over 95% of cases had ≤3.0° deviation between planned and executed alignment parameters.
- Gap balancing improved significantly with reduced variability in flexion and extension gaps.
- Functional outcomes in extension, flexion, and deformity correction were significantly improved (p < 0.001).

## Abstract

Background: Robotic-assisted systems have transformed total knee arthroplasty (TKA), promising improved accuracy and intraoperative consistency, yet real-world data from non-academic centers remain limited. Objective: This study evaluates five-year clinical integration of a semi-autonomous, CT-based, robotic-arm-assisted TKA at a tertiary non-teaching hospital in Germany, focusing on planning accuracy, gap balancing, and intraoperative outcomes. Methods: We retrospectively analyzed all patients (n = 457) who underwent MAKO-assisted TKA from 2020 to 2025, performed by three orthopedic surgeons using a standardized subvastus approach. We assessed preoperative deformities, intraoperative alignment, implant sizing, and gap balancing. Surgical plans were adapted intraoperatively when indicated. Pre- vs. post-implantation values were compared using slopes to evaluate execution consistency. Results: Median patient age was 67.0 years (IQR: 60.0–75.0), with varus in 84.1% (7.0°, IQR: 4.0°–10.0°), valgus in 13.2% (3.0°, IQR: 1.5°–5.8°), and neutral alignment in 2.7%. Flexion contracture occurred in 80.4% (6.0°, IQR: 3.0–10.0%), hyperextension in 12.7% (2.0°, IQR: 1.5°–5.0°). Planning-to-execution consistency was high, even with plan adaptations. Slope values for alignment parameters were: tibial rotation in degrees (slope value: 1.0), femoral sagittal angle in degrees (0.8), tibial sagittal angle in degrees (0.9), coronal posterior condylar angle in degrees (0.9), femoral component size (1.0), tibial component size (1.0). Over 95% of cases showed ≤3.0° deviation between planned and final values. Bone resection concordance showed moderate agreement, with slopes from 0.8 (posterior medial femoral cut in mm) to 0.5 (lateral tibial cut in mm). Gap balancing improved at all stages, with reduced variability in medial/lateral extension and flexion gaps (all p < 0.05). Functional reconstruction showed significant improvements in extension, flexion, and deformities (all p < 0.001). Conclusions: Semi-autonomous, CT-based, robotic-arm-assisted TKA was successfully implemented in this non-academic setting, demonstrating acceptable intraoperative and functional reconstruction outcomes, supporting the feasibility of robotic-assisted surgery outside academic centers.