Robotic-assisted unicompartmental knee arthroplasty is associated with lower odds of prolonged hospitalization and no higher odds of high-charge admission during the index hospitalization

BACKGROUND: Robotic-assisted unicompartmental knee arthroplasty (RA-UKA) may improve implant positioning, but its impact on index-hospitalization length of stay (LOS), billed charges, and early inpatient outcomes in routine practice remains unclear. METHODS: Using the National Inpatient Sample (2016-2022), we identified primary medial UKA for osteoarthritis. RA-UKA was defined by ICD-10-PCS robotic-assisted lower-extremity procedure codes. Primary outcomes were prolonged hospitalization (LOS ≥ the 75th percentile, ≥ 2 days), high-charge admission (total hospital charges [TOTCHG] ≥ the 75th percentile, ≥ $72,321), and in-hospital mortality. We performed univariable comparisons and multivariable logistic regression adjusting for demographics, payer, admission type, hospital teaching status, comorbidities, and calendar year fixed effects (YEAR 2016-2022) using the unweighted NIS discharge sample. RESULTS: Among 7,154 UKAs, 1,297 (18.1%) were RA-UKA and 5,857 (81.9%) were C-UKA. Median LOS was 1 day (IQR 1-2) in both groups; however, prolonged hospitalization (LOS ≥ 2 days) occurred less frequently in RA-UKA (412/1,297 [30.7%] vs. 2,162/5,857 [37.2%]; P < 0.001). In adjusted analyses including YEAR fixed effects, RA-UKA was associated with lower odds of prolonged hospitalization (aOR 0.750, 95% CI 0.647-0.870; P < 0.001) and was not associated with high-charge admission (TOTCHG ≥ $72,321; aOR 1.007, 95% CI 0.855-1.186; P = 0.993). In-hospital mortality and other inpatient complications were rare. CONCLUSIONS: RA-UKA was associated with lower odds of prolonged hospitalization (LOS ≥ 2 days) and no higher odds of a high-charge admission (billed charges, TOTCHG) during the index hospitalization. Findings apply to the index hospitalization only.