Risk Scores Do Not Predict High Mortality After Coronary Artery Bypass Surgery in the Presence of Diastolic Dysfunction

Background Although global postoperative mortality after on-pump coronary artery bypass grafting is approximately 3%, in some groups it can be considerably higher. Many conditions are known to increase mortality and have been included in well-known scoring systems; however, left ventricular diastolic dysfunction has not been sufficiently evaluated to identify its predictive value for mortality after coronary artery bypass grafting, nor is it integrated in currently used risk scores. Methods Left ventricular filling pattern was prospectively evaluated in 191 patients scheduled for on-pump coronary artery bypass grafting. A follow-up of survival and complications was made for 30 days postoperatively. Observed mortality was compared with the mortality predicted by the scores of EuroSCORE and Parsonnet. Results A correlation was found between diastolic function, the presence of comorbidities, and postoperative survival. There was no mortality in the group with normal filling pattern (0 of 33 patients). In the presence of an alteration of relaxation, mortality was 5 of 129 patients (3.8%); in the pseudonormal group it was 2 of 16 patients (12.5%); and in the restrictive group it was 6 of 13 patients (46.1%; p < 0.01). Parsonnet and EuroSCORE predicted a mortality of 1.5% to 1.6%, 1.5% to 2.0%, 1.5% to 2.2%, and 3.9% to 4.1% for each group, respectively. Mortality in the group with E deceleration time of 150 ms or greater was 2.8% and in the group with E deceleration time less than 150 ms was 17.3% (p < 0.01). Postoperative complications were also more frequent in the group with advanced dysfunction. Conclusions Severe diastolic dysfunction is a strong predictor of adverse outcome and mortality after on-pump coronary artery bypass grafting, and this high risk is not adequately predicted by EuroSCORE and Parsonnet score. Measures of diastolic function should be included in routine preoperative risk assessment.