Is comorbid status the best predictor of one-year mortality in patients with severe sepsis and sepsis with shock?

Abstract

Understanding longer term outcomes in critically ill patients will assist treatment decisions, allocation of scarce resources and clinical research in that population. The aim of this study was to compare a wellvalidated means of determining comorbidity, the Charlson Comorbidity Score, to other verified risk stratification models in predicting one-year mortality and other outcomes in emergency department patients with severe sepsis and sepsis with shock. We conducted a planned subgroup analysis of a prospective observational study, the Critical Illness and Shock Study, in adult patients with sepsis meeting study criteria for critical illness. From emergency department arrival, patients were prospectively enrolled with data collected for a minimum of one year post-enrolment. Scoring systems were derived from this data and compared using receiver-operating characteristic curves. One hundred and four patients were enrolled. The 28-day mortality was 18% and one-year mortality 40%. For predicting one-year mortality, the area under the receiver-operating characteristic curve for age-weighted Charlson Comorbidity Score (0.71, 95% confidence interval 0.61 to 0.81) was at least as good or superior to other scoring systems analysed. The intensive care unit admission rate was 45% and the median hospital length-of-stay was eight days. We conclude that in patients who present to the emergency department with severe sepsis or sepsis with shock, age-weighted Charlson Comorbidity Score is a predictor of one-year mortality that is simple to calculate and at least as accurate as other validated scoring systems.