Abstract
BACKGROUND: High-volume centers report acceptable outcomes for robotic mitral repair, yet population-level data are limited. This study compared late mortality and reintervention rates for robotic vs nonrobotic mitral repair in the United States. METHODS: The Centers for Medicare & Medicaid Services database validated against clinical records was used to identify 26,524 patients with isolated first-time nonemergency mitral repairs with or without tricuspid repairs or ablations without other concomitant procedures. Of these patients, 2227 (8.3%) underwent robotic repair and 24,297 (91.7%) underwent nonrobotic repairs. Propensity score matching was performed on 30 baseline characteristics. The primary end point was a composite of death or mitral reintervention, and the secondary end point was all-cause mortality. Both were compared in a Cox proportional hazards model. Falsification end point analysis assessed for potential unmeasured confounders with death as a competing risk. RESULTS: Matching yielded 2226 patient pairs (mean age, 72 years; 44% female; 9% with concomitant tricuspid repair; 7% ablation). Thirty-day mortality did not differ between the groups (1.3% robotic vs 1.3% nonrobotic; P = .90). Robotic repair was associated with lower postoperative atrial fibrillation (19.1% vs 23.2%; P = .001) and a shorter hospital stay (median 5 days [interquartile range, 4-7 days] vs 7 days [interquartile range, 5-9 days]; P < .001). At 5 years, the composite of death or mitral reintervention (17.8% vs 18.6%; hazard ratio, 0.93; 95% CI, 0.79-1.09; P = .37) and all-cause mortality (14.9% vs 15.6%; hazard ratio, 0.93; 95% CI, 0.77-1.11; P = .40) were similar, with falsification testing confirming minimal confounding (P = .20). CONCLUSIONS: Robotic mitral repair in the United States is safe, yielding outcomes comparable to those of nonrobotic repair.