Predictive risk factors for prolonged stay in intensive care unit in patients undergoing coronary artery bypass grafting surgery / 中华胸心血管外科杂志

ABSTRACT

Objective To describe the preoperative factors of prolonged intensive care unit length of stay after coronary artery bypass grafting. Methods From 1997 to 2009, 1318 patients underwent isolated CABG in our hospital. Retrospective analysis was performed on these cases. Univariate and multivariate analyses for preoperative risk factors were performed. Prolonged length of ICU stay was defined as initial admission to ICU exceeding 72 h. Results The mean age of patients ( 322women and 996 men) was (67.4±9.4) years. Of 1318 patients undergoing isolated CABG from 1997 to 2009, 205 experienced prolonged length of ICU stay. The length of ICU stay was (40.1 ± 22.5 ) hours and ( 122.6 ± 48.7 ) hours separately.Overall in-hospital mortality was higher among these 205 patients ( 13.7% vs. 1.2%, P ＜0.05 ). The overall mortality was 3.1%. In univariate analyses, there were statistically significant differences with respect to the percentage of CPB, total bypass time, cross-clamp time, number of distal anastomoses, use of pressor agent, use of intro-aortic balloon pump,time of ventilation and hospital mortality. The significant risk factors were age, NYHA class Ⅲ/Ⅳ, left ventricular ejection fraction(LVEF) ＜0.40, renal failure, cerebrovascular and/or peripheral vascular disease, chronic obstructive pulmonary disease, recent acute myocardial infarction, prior percutaneous coronary intervention, left main stenosi, three-vessels disease. The variables entered into the multivariate logistic regression were age, NYHA class Ⅲ/Ⅳ, LVEF ＜0.40, renal failure, chronic obstructive pulmonary disease, recent acute myocardial infarction, prior percutaneous coronary intervention, three-vessels disease. According to the outcome of multivariate logistic regression, we can conclude the model of probability forecast and create a new variable named Pre. The area under ROC curve of the new variable Pre was larger than other variables. Conclusion The main risk factors of prolonged ICU length of stay were LVEF ＜ 0.40, recent acute myocardial infarction, renal failure and chronic obstructive pulmorary disease. The AUC of variable Pre was higher than other' s , which indicated that new variable Pre combining each variable was more valuable than single variable with respect to prediction. A predicted probability of 0. 754 was used as cut-off point for the prognostic test.