The ability to predict outcome in coronary artery bypass graft (CABG) surgery has always been of interest to physicians and patients. Over the past 10 years, attempts have been made to predict operative mortality. However, the mortality rates, which are below 4%, do not seem to correlate with complication rates or length of hospitalization, suggesting that morbidity might be a more valid and predictable end point. The major morbid events (and their approximate frequencies) are reoperation (8%), mechanical ventilation >48 hours (7%), excessive bleeding (6%), low cardiac output (5.5%), pulmonary infection (5%), other infection (4.5%), myocardial infarction (4.5%), and renal failure (2.5%).
A recent univariate analysis identified over 20 risk factors, including eight that were so robust as to resist multivariate analysis: two intraoperative factors (cardiopulmonary bypass time, and operating surgeon), and six preoperative factors (symptomatic right heart failure, ventricular arrhythmia, reoperation, chronic obstructive pulmonary disease, body mass index <24, and electrocardiographic ST-seg-ment changes). Surprisingly, neither diabetes, hyper-lipidemia, hypertension, nor angiographic left main coronary artery stenosis correlated significantly with severe morbidity. Unfortunately, however, not all studies agree. Twelve simple clinical scoring systems have been published to date for predicting morbidity and mortality in patients undergoing revascularization or cardiac surgery in general. They incorporate a total of 40 predictors of complications at the univariate level. No one factor was present in all models, and only six independent factors were present in over five models: gender, age, emergency, reoperation, left ventricular dysfunction, and diabetes. These discrepancies are not readily explicable, but they raise the following points: • The difficulty of predicting hemorrhage and infection; • The failure of a pragmatic approach to take intra-and postoperative factors into account, eg, operative mortality is lower with surgeons and institutions performing more operations. • The effect of chance and biological diversity. • The frequent absence of a relationship between statistical and clinical significance. The frequency of one of the most significant predictors of morbidity right heart failure is <3%: statistically significant predictors are clinically common, while clinically common predictors are not very statistically significant. This is why assessing morbidity and mortality predictors directly in high-risk CABG candidates could be much more useful. The CABG Patch Trial was a multicenter trial of high-risk patients undergoing surgical revascularization. Patients were under 80 years of age, with a left ventricular ejection fraction <36%, an abnormal signal-averaged electrocardiogram, and no prior sustained ventricular arrhythmias. Paradoxically, the results showed an association between an impaired left ventricular ejection fraction and increased longterm survival. However, this relationship was not as strong as that between New York Heart Association (NYHA) functional class and survival. Adjustment for functional class weakened the association between left ventricular ejection fraction and survival. The message of the CABG Patch Trial was that NYHA class is a more potent predictor of CABG Further reading Green J, Wintfeld N. Report cards on cardiac surgeons. Assessing New York State's approach N Engl J Med 1995;332:1229-1232. Hammermeister KE, Burchfiel C, Johnson R, Grover FL Identification of patients at greatest risk for developing major complications at cardiac surgery. Circulation. 1990;82(suppl):IV380-IV389. Olshansky B, Telfer EA, Curtis AB, Bigger JT. Predictive value of preoperative left ventricular ejection fraction and functional class for mortality and morbidity after high-risk coronary artery bypass grafting. Am J Cardiol. 2000;85:1489-1491. mortality in high-risk patients than the preoperative left ventricular ejection fraction, ie, global clinical evaluation is a better discriminator than instrumental parameters. Petros AJ, Marshall JC, van Saene HK. Should morbidity replace mortality as an endpoint for dinical trials in intensive care? Lancet. 1995;345:369-371. Stoat P, Cure Herat M, George M, et al. Severe morbidity after coronary artery surgery: development and validation of a simple predictive clinical score. Eur Heart J. 1999,20:960-966. Keywords management; surgical revascularization; coronary artery bypass grafting; left ventricular dysfunction; mortality; morbidity; complication; risk