Probability of survival as a prognostic and severity of illness score in critically ill surgical patients

Abstract

A new quantitative method for measuring the prognosis and severity of illness in terms of probability of survival was developed from 224 studies in an index population of 220 critically ill surgical patients. Patients were selected preoperatively to eliminate pre-existing cardiac disease, cirrhosis, nutritional debility, shock or sepsis, in order to evaluate the physiologic relationships of surgical trauma to outcome free of confounding associated medical disorders. The empirically derived numeric severity index was calculated from the probability of survival for each of 28 hemodynamic and oxygen transport variables at each time period after surgery. The score correctly indicated patient outcome in 96% of the index population and 94% of an independent, prospective population. The survivors' score consistently predicted survival within 21.6 +/- 4.4 (SEM) h after the end of surgery. The severity score of those who died consistently predicted nonsurvival within 37 +/- 11 (SEM) h after the end of surgery. We conclude that the score provides a useful, objective, physiologic measure of the severity of illness and prognosis.