AIMS: To develop a predictive clinical risk score of post-operative morbidity after coronary artery bypass grafting. METHODS AND RESULTS: Data were collected retrospectively from 679 patients undergoing emergency or planned bypass surgery between 1 January and 31 December 1996. The incidence of morbidity was 23%. Multivariate stepwise logistic regression analysis on two-thirds of the patients identified eight independent risk factors for severe morbidity. Six of these were pre-operative: symptomatic right heart failure, previous ventricular arrhythmias, previous coronary bypass surgery, chronic pulmonary disease, ST changes on pre-operative electrocardiogram, body mass index <24 kg. m-2, and two were intra-operative factors: the surgeon who operated, and the cardiopulmonary bypass time. A predictive clinical risk score was developed with the six pre-operative risk factors. The negative predictive value of the model is 87% and the area under the receiver operating characteristic curve is 0.77. When tested on the remaining patients not used for developing the model, the area under the curve is 0.65. CONCLUSION: This pre-operative risk score provides a simple method of risk stratification for patients undergoing coronary artery surgery. However, as for all predictive models, the performance of the score decreases when applied to a population other than that used to develop it. Copyright 1999 The European Society of Cardiology.
AIMS: To develop a predictive clinical risk score of post-operative morbidity after coronary artery bypass grafting. METHODS AND RESULTS: Data were collected retrospectively from 679 patients undergoing emergency or planned bypass surgery between 1 January and 31 December 1996. The incidence of morbidity was 23%. Multivariate stepwise logistic regression analysis on two-thirds of the patients identified eight independent risk factors for severe morbidity. Six of these were pre-operative: symptomatic right heart failure, previous ventricular arrhythmias, previous coronary bypass surgery, chronic pulmonary disease, ST changes on pre-operative electrocardiogram, body mass index <24 kg. m-2, and two were intra-operative factors: the surgeon who operated, and the cardiopulmonary bypass time. A predictive clinical risk score was developed with the six pre-operative risk factors. The negative predictive value of the model is 87% and the area under the receiver operating characteristic curve is 0.77. When tested on the remaining patients not used for developing the model, the area under the curve is 0.65. CONCLUSION: This pre-operative risk score provides a simple method of risk stratification for patients undergoing coronary artery surgery. However, as for all predictive models, the performance of the score decreases when applied to a population other than that used to develop it. Copyright 1999 The European Society of Cardiology.
Authors: David R Lawrence; Rajael Somaskanthan; Matthew J Barnard; Miles Curtis; Bruce E Keogh Journal: Ann R Coll Surg Engl Date: 2009-04-02 Impact factor: 1.891