Intensive care unit admission parameters improve the accuracy of operative mortality predictive models in cardiac surgery.

Abstract

BackgroundOperative mortality risk in cardiac surgery is usually assessed using preoperative risk models. However, intraoperative factors may change the risk profile of the patients, and parameters at the admission in the intensive care unit may be relevant in determining the operative mortality. This study investigates the association between a number of parameters at the admission in the intensive care unit and the operative mortality, and verifies the hypothesis that including these parameters into the preoperative risk models may increase the accuracy of prediction of the operative mortality.Methodology929 adult patients who underwent cardiac surgery were admitted to the study. The preoperative risk profile was assessed using the logistic EuroSCORE and the ACEF score. A number of parameters recorded at the admission in the intensive care unit were explored for univariate and multivariable association with the operative mortality.Principal findingsA heart rate higher than 120 beats per minute and a blood lactate value higher than 4 mmol/L at the admission in the intensive care unit were independent predictors of operative mortality, with odds ratio of 6.7 and 13.4 respectively. Including these parameters into the logistic EuroSCORE and the ACEF score increased their accuracy (area under the curve 0.85 to 0.88 for the logistic EuroSCORE and 0.81 to 0.86 for the ACEF score).ConclusionsA double-stage assessment of operative mortality risk provides a higher accuracy of the prediction. Elevated blood lactates and tachycardia reflect a condition of inadequate cardiac output. Their inclusion in the assessment of the severity of the clinical conditions after cardiac surgery may offer a useful tool to introduce more sophisticated hemodynamic monitoring techniques. Comparison between the predicted operative mortality risk before and after the operation may offer an assessment of the operative performance.