Total knee arthroplasty (TKA) achieves good overall outcomes, yet 15–30% of patients remain dissatisfied, often attributed in part to variability in implant positioning and soft-tissue balance. Robotic-assisted TKA (RA-TKA) aims to improve intraoperative accuracy, but its real-world association with postoperative healthcare utilization and mid-term mechanical outcomes remains uncertain at scale. This study compared 90-day utilization and complication outcomes and 2-year mechanical outcomes of RA-TKA versus conventional TKA (C-TKA) in a large, national propensity-matched cohort. This retrospective cohort study used the TriNetX US Collaborative Network to evaluate outcomes following primary TKA among adults undergoing TKA between January 2018 and January 2023. Patients receiving RA-TKA were identified and 1:1 propensity-matched to C-TKA patients based on demographics and comorbidities. Primary outcomes were 90-day inpatient hospitalization encounter and 90-day opioid prescription fill. Secondary outcomes included 2-year mechanical outcomes (revision TKA and mechanical loosening diagnoses). Analyses were performed within TriNetX using built-in propensity matching and risk ratio calculations with 95% confidence intervals. After propensity score matching, covariate balance was assessed using standardized mean differences, with values 0.05). At 2 years, RA-TKA demonstrated lower rates of revision (1.3% vs. 2.2%, RR 0.58; p < 0.001) and mechanical loosening diagnoses (0.28% vs. 0.78%, RR 0.35; p < 0.001), with no difference in manipulation under anesthesia (3.26% vs. 3.22%, p = 0.919). In this large propensity-matched national cohort, RA-TKA was associated with lower 90-day inpatient hospitalization encounters and opioid prescription fills, and lower coded revision and loosening outcomes at 2 years, while early complication rates were similar. These findings should be interpreted as associations given the use of administrative coding and unmeasured confounding.
Ghafarian et al. (2026) studied this question.