Background: The Society of Thoracic Surgeons Adult Cardiac Surgery Database (STS ACSD) was expanded in 2017 to include more granular detail on thoracic aortic surgeries. We describe the first validated risk model in thoracic aortic surgery from the STS ACSD.
Methods: The study population consisted of patients undergoing nonemergent isolated ascending aortic aneurysm repair by open or clamped distal anastomoses, including those requiring aortic root or valve replacement. Model outcomes included operative mortality, 30-day major morbidity (cardiac reoperation, deep sternal wound infection, stroke, prolonged ventilation, renal failure), and a composite of both. To select the predictors, univariate associations and clinical face validity of models were examined. Models were evaluated by their ability to distinguish between patients with and without specific outcomes (discrimination) and their predictive accuracy (calibration).
Results: Between 2017 and 2021, 24,051 eligible patients underwent ascending aortic aneurysm surgery at 905 hospitals. Procedures included 8913 aortic root replacements, 2135 valve-sparing root replacements, 7545 ascending aortic replacements with aortic valve replacement, and 5458 ascending aortic replacements. Circulatory arrest was performed in 7316 (30.4%) cases. Operative mortality was 1.9%, and 12.2% of patients experienced major morbidity including 2.4% incidence of stroke. The adjusted C statistics for the model were 0.74, 0.67, and 0.67 for mortality, morbidity, and the composite, respectively. Previous stroke and circulatory arrest were associated with new stroke. Genetic aortopathy was associated with less mortality.
Conclusions: A new STS ACSD risk model to predict mortality and morbidity after ascending aneurysm surgery has been developed, and predictors of better and worse outcomes have been identified.
Copyright © 2024. Published by Elsevier Inc.