Elective Aortic Root Replacement in North America: Analysis of STS Adult Cardiac Surgery Database

Unlike coronary artery bypass and aortic and mitral valve procedures, there is no predictive risk model for aortic root replacement procedures. As a first step toward development of a risk model, this study analyzed The Society of Thoracic Surgeons (STS) Adult Cardiac Surgery Database to determine factors predictive of mortality and morbidity in patients undergoing elective aortic root replacement (ARR).

The STS database was queried (from July 2011 to June 2016) for elective ARR with the following exclusion criteria: urgent or salvage cases, endocarditis, redo cardiac surgery, circulatory arrest, and aortic arch surgery. Adjusted multivariate logistic regression models for outcomes of mortality and composite STS morbidity were performed using covariates of the STS aortic valve risk set (expressed as odds ratios [ORs]).

Of 24,244 patients undergoing ARR, 8,807 (77.6% male) met inclusion criteria in 808 centers; 33.7% (n = 2,965) had a bicuspid aortic valve, and 3.7% (n = 327) had Marfan syndrome. The median age was 58.0 years (interquartile range, 49 to 67 years). Median intensive care unit and hospital stays were 46 hours and 6 days, respectively. Significant predictors for mortality included: atrial fibrillation (OR, 2.06), body surface area (OR, 0.14), chronic obstructive pulmonary disease (OR, 1.2), New York Heart Association class IV (OR, 2.53), diabetes (OR, 2.48), coronary artery bypass grafting (OR, 2.77), mitral valve surgery (OR, ≥2.18), and Bentall operation (OR, 2.08). Regression analysis for risk factors for STS morbidity yielded 14 significant factors. A glomerular filtration rate increase of 20 units was predictive of improved mortality (OR, 0.85) and morbidity (OR, 0.91).

Elective ARR is performed with excellent postoperative outcomes. Analysis of the STS database reveals several significant risk factors that are independently associated with increased mortality and morbidity. The investigators anticipate that future studies inclusive of the nonelective ARR cases in the database will facilitate development of a risk model for root replacement procedures.