[Prospects for prognosis in cardiac surgery: experience from using discriminant linear analysis].

The authors evaluate the possibility of using the discriminant linear analysis in predicting the unfavourable outcomes at aortocoronary bypass. There were investigated 98 patients (the mean age 64.4 +/- 3.0 years) who underwent the surgical revascularization of myocardium in 1992-1997. 56 signs characterizing the demographic, clinical-and-instrumental and operational indices were analyzed. The obtained discriminant equation of operation outcome prediction included 15 signs. The operation for unstable angina, the cholesterol level and the patient's age were the most informative parameters. The total per cent of correct classification was 87.8%. According to the authors' experience the discriminant linear analysis is the informative method of multi-dimensional statistical analysis permitting to reveal the most significant signs that influence on the prognosis of surgical intervention.