Agreement between planned, robot-validated, and postoperative radiographic posterior tibial slope in ROSA-assisted total knee arthroplasty

Posterior tibial slope (PTS) is a key determinant of sagittal knee mechanics, cruciate ligament loading, and flexion-extension behavior after total knee arthroplasty (TKA). Accurate reproduction of the intended PTS is therefore an important objective in contemporary TKA, particularly when patient-specific alignment targets are used. The aim of this study was to evaluate the radiographic agreement among planned, robot-validated, and postoperative PTS values in ROSA-assisted TKA using standardized true lateral radiographs and blinded repeated measurements, and to determine the reliability of postoperative radiographic measurements. This retrospective observational radiological study, reported in accordance with the STrengthening the Reporting of OBservational studies in Epidemiology statement, included knees that underwent robotic-assisted primary TKA using the ROSA system and had complete intraoperative robotic records and adequate postoperative true lateral radiographs. Postoperative true lateral radiographs were obtained at 6-week follow-up. Planned PTS values were obtained from the robotic planning interface, and final robot-validated PTS values were recorded intraoperatively after tibial preparation. Postoperative PTS was measured on standardized true lateral radiographs by 2 independent blinded orthopedic surgeons, each performing repeated measurements at a 2-week interval. Agreement was assessed using paired comparisons, mean absolute differences, outlier analysis, and Bland-Altman evaluation. A total of 73 knees from 64 patients were initially reviewed, and 72 knees with complete paired measurements were included in the final analysis. There was no statistically significant difference between planned and robot-validated PTS values. The mean absolute difference between planned and robot-validated PTS was 0.57°, and 80.6% of knees were within ±1° of the planned target, while all knees were within ±2° and ±3°. Similarly, no statistically significant difference was observed between robot-validated and postoperative mean radiographic PTS values. Radiographic measurement reliability was excellent, with ICC(3,1) = 0.983 (95% confidence interval = 0.976-0.989) and ICC(3,k) = 0.996 (95% confidence interval = 0.994-0.997). ROSA-assisted TKA demonstrated close radiographic agreement among planned, robot-validated, and postoperative PTS values, with low absolute deviations and excellent measurement reliability. Because this study lacked a comparator group and did not assess clinical outcomes, these findings should be interpreted as evidence of radiographic reproducibility rather than evidence of superiority over conventional instrumentation or proof of clinical benefit.