Contemporary outcomes in reoperative mitral valve surgery

OBJECTIVE: Data suggest that redo mitral valve surgery is being performed in increasing numbers, possibly with superior results according to single-centre studies. The purpose of this study is to describe outcomes of redo mitral valve surgery and identify risk-adjusted predictors of poor outcomes. METHODS: All (11 973) open mitral valve cases were evaluated (2002-2016) from a regional Society of Thoracic Surgery (STS) database. Patients were stratified by primary versus redo mitral valve surgery. Mixed effects logistic regression models including hospital as a random effect were used to identify risk factors for patients undergoing redo mitral valve surgery. RESULTS: Of all mitral valve cases, 1096 (9.7%) had a previous mitral operation. Redo patients had higher rates of valve replacement and preoperative comorbidities resulting in more complications, operative mortalities (11.1%vs6.5%, p<0.0001) and higher resource utilisation. Several factors independently increased risk for composite STS major morbidity and 30-day mortality, including cardiogenic shock (OR 10.3, p=0.0001), severe tricuspid insufficiency (OR 2.3, p=0.001), urgent/emergent status (OR 1.8, p=0.001) and concurrent coronary artery bypass grafting (OR 2.4, p=0.002). The volume of redo mitral valve surgery increased 10% per year and the observed-to-expected ratios (O/E) for operative mortality in redo mitral surgery improved from 1.44 early in the study period to 0.72 in the most recent era. CONCLUSIONS: Redo mitral valve surgery accounts for approximately 10% of mitral valve operations and is associated with increased risk and resource utilisation. However, as the volume of redo mitral surgery increases, outcomes have dramatically improved and are now better than predicted.