Mortality risk scoring in emergency general surgery: Are we using the best tool?

Abstract

Background: It is imperative that an accurate assessment of risk of death is undertaken preoperatively on all patients undergoing an emergency laparotomy. Portsmouth-Physiological and Operative Severity Score for the enumeration of Mortality and Morbidity (P-POSSUM) is one of the most widely used scores. National Emergency Laparotomy Audit (NELA) presents a novel, validated score, but no direct comparison with P-POSSUM exists. We aimed to determine which would be the best predictor of mortality.

Methods: We analysed all the entries on the online NELA database over a four-and-a-half-year period. The Hosmer-Lemeshow goodness of fit test was performed to assess model calibration. For the outcome of death and for each scoring system, a non-parametric receiver operator characteristic analysis was done. The sensitivity, specificity, area under receiver operator characteristic curve and their standard errors were calculated.

Results: Data pertaining to 650 patients were included. There were 59 deaths, giving an overall observed mortality rate of 9.1%. Predicted mortality rate for the P-POSSUM score and NELA score were 15.2% and 7.8%, respectively. The discriminative power for mortality was highest for the NELA score (C-index = 0.818, CI: 0.769-0.867, p < 0.001), when compared to P-POSSUM (C-index = 0.769, CI: 0.712-0.827, p < 0.001).

Conclusions: The NELA score showed good discrimination in predicting mortality in the entire cohort. The P-POSSUM over-predicted observed mortality and the NELA score under-predicted observed mortality.

Keywords: Emergency surgery; Mortality; NELA; P-POSSUM; Risk scoring.