Evaluation of the society of thoracic surgeons online risk calculator for assessment of risk in patients presenting for aortic valve replacement after prior coronary artery bypass graft: An analysis using the STS Adult Cardiac Surgery Database

Background Accurate risk assessment in patients presenting for aortic valve replacement (AVR) after prior coronary artery bypass grafting (CABG) is essential for appropriate selection of surgical versus percutaneous therapy. Methods We included 6,534 patients in The Society for Thoracic Surgeons (STS) Adult Cardiac Surgery Database (October 2009 through December 2013) who underwent elective, isolated reoperative AVR for aortic stenosis after prior CABG. Case-specific PROM was calculated and observed-to-expected ratios were inspected across the spectrum of risk. A cohort-specific recalibration equation was derived using logistic regression: = expit(-0.6453+0.6147∗logit(PROM) -0.0709∗logit(PROM)^ˆ2), where PROM is the predicted risk of mortality. The proportion of patients reclassified as low (PROM < 4%), intermediate (4% to < 8%), high (8% to < 12%), and very high risk (≥12%) was calculated using the recalibration equation. The performance of the cohort-specific recalibration equation was then compared with the generic recalibration for quarterly STS reports. Results The STS online risk calculator overestimates risk for low, intermediate, and high risk categories. Using the recalibrated risk equation, a substantial proportion of patients were reclassified as the following: 25.5% from intermediate to low risk; 39.7% from high to intermediate risk; and 41.5% from very high to high risk. Comparison of the cohort-specific recalibration equation to the generic quarterly STS recalibration demonstrated very similar results. Conclusions In patients presenting for AVR after prior CABG, the STS online risk calculator overestimates risk for all but the highest risk patients. Using a cohort-specific recalibration equation, a substantial proportion of patients would be downgraded to lower risk categories. The cohort-specific recalibration correlates well with the existing generic quarterly STS recalibration. The findings of this study support recommendations for periodic recalibration of the online risk calculator in order to facilitate clinical decision making in real time.