Abstract
BACKGROUND: Anticoagulation after bioprosthetic mitral valve (MV) replacement (BMVR) and repair (MVrep) is controversial. We explore outcomes among BMVR and MVrep patients in The Society of Thoracic Surgeons Adult Cardiac Surgery Database based on discharge anticoagulation status. METHODS: BMVR and MVrep patients aged ≥65 years in The Society of Thoracic Surgeons Adult Cardiac Surgery Database were linked to the Centers for Medicare and Medicaid Services claims database. Long-term mortality, ischemic stroke, bleeding, and a composite of the primary end points were compared as a function of anticoagulation. Hazard ratios (HRs) were calculated using multivariable Cox regression. RESULTS: A total of 26,199 BMVR and MVrep patients were linked to the Centers for Medicare and Medicaid Services database; of these, 44%, 4%, and 52% were discharged on warfarin, non-vitamin K-dependent anticoagulant (NOAC), and no anticoagulation (no-AC; reference), respectively. Warfarin was associated with increased bleeding in the overall study cohort (HR, 1.38; 95% CI 1.26-1.52) and in the BMVR (HR, 1.32; 95% CI, 1.13-1.55) and MVrep subcohorts (HR, 1.42; 95% CI, 1.26-1.60). Warfarin was associated with decreased mortality only among BMVR patients (HR, 0.87; 95% CI, 0.79-0.96). Stroke and the composite outcome did not differ across cohorts with warfarin. NOAC use was associated with increased mortality (HR, 1.33; 95% CI 1.11-1.59), bleeding (HR, 1.37; 95% CI, 1.07-1.74), and the composite outcome (HR, 1.26; 95% CI, 1.08-1.47). CONCLUSIONS: Anticoagulation was used in fewer than half of mitral valve operations. In MVrep patients, warfarin was associated with increased bleeding and was not protective against stroke or mortality. In BMVR patients, warfarin was associated with a modest survival benefit, increased bleeding, and equivalent stroke risk. NOAC was associated with increased adverse outcomes.