Abstract
PURPOSE: To compare 90-day readmissions, complications, and resource use after robotic-assisted versus conventional total knee arthroplasty (TKA) in a contemporary, nationally representative cohort. METHODS: Retrospective cohort study using the Nationwide Readmissions Database (NRD) 2020-2022. Primary TKA identified from the PR1 field (ICD-10-PCS). Exclusions included non-elective admissions, revisions, bilateral procedures, age < 18, oncology/fracture/reoperation, COVID-19, and discharges after September. Readmissions within 90 days were categorized (prosthesis/SSI, mechanical/implant, VTE). 1:1 propensity score matching (nearest neighbor, caliper 0.01, no replacement) included demographics, comorbidities, hospital factors, and year. RESULTS: After matching, 96,982 patients (48,491 per group) were analyzed. Robotic TKA showed lower all-cause 90-day readmission (5.0% vs 6.5%), superior readmission-free survival (log-rank P < 0.001), shorter readmission LOS (4.8 vs 5.6 days), and lower readmission charges ($66,769 vs $75,544), with slightly higher index charges ($78,125 vs $74,090). Risks were lower for VTE/PE, pneumonia, transfusion, postoperative pain, and prosthesis/SSI-related and mechanical readmissions. CONCLUSIONS: In the largest contemporary national analysis, robotic TKA was associated with fewer early complications, lower 90-day readmissions, and reduced readmission resource use compared with conventional TKA.