A quality control method in cardiac surgical outcome: experience in 462 patients.

The aim of our study is to verify the reliability, reproductiveness and simplicity of a method to control cardiac surgical results. We divided 462 adult patients, operated on for acquired heart disease from October 1989 to January 1991, into five classes according to an individual score which was predictive for their operative mortality risk. The score resulted from 15 different risk factors tested with univariate and multivariate analysis against one event: operative death. The total number of deaths was 12: 2, 2, 1, 2, 5 for each class respectively. When comparing the predicted versus our observed mortality, we found no statistically significant difference, using the chi-squared test. The method we used is highly predictive for surgical mortality risk: it makes the results objectively comparable among different institutions; it is useful as a self-controlled quality method for cardiac surgical activity in any single institution.