Preoperative early warning scores can predict in-hospital mortality and critical care admission following emergency surgery.

ABSTRACT

BACKGROUND: EWS is frequently used to monitor acute admissions requiring emergency surgery. This study examined preoperative early warning scoring (EWS) and its ability to predict mortality and critical care admission. Postoperative EWS was also evaluated as a predictor of mortality. METHODS: Preoperative EWS, age, physiologic and operative severity (POSSUM) scores, ASA grade, and serology were compared in 280 patients undergoing emergency surgery. RESULTS: Two hundred eighty patients were identified with a mortality of 15%. Among the physiological scoring systems, ASA grade and POSSUM scores were the best predictors of mortality (AUC values of 0.81). EWS, APACHE II, and age were the next best predictors (AUC values of 0.70). Postoperative APACHE II and EWS both predicted mortality. EWS on day 2 postoperatively was the best overall predictor of mortality of all the variables studied (AUC value of 0.83). Survival between patients with "improving or stable" EWS and those with "deteriorating or failing to improve" EWS was also found to be significantly different (P < 0.001). In addition, both EWS on admission and EWS 1 h preoperatively were found to predict critical care requirement postoperatively (AUC value of 0.78). CONCLUSIONS: EWS can predict the need for critical care admission and mortality following emergency surgery. In particular, the progression of EWS preoperatively, that is, whether scores improve or deteriorate, is a highly significant factor in predicting survival following emergency surgery. These findings support the use of EWS in monitoring the acute surgical patient.