Comparison of APACHE II and III scoring systems for mortality prediction in critical surgical illness.

Overview

OBJECTIVE: To determine whether the Acute Physiology and Chronic Health Evaluation III (APACHE III), an updated version of APACHE II that contains a larger number of postoperative patients in the normative database, offers better prediction in critical surgical illness. DESIGN: Prospective cohort study. SETTING: Surgical intensive care unit of an urban, tertiary-care university hospital. PARTICIPANTS: Eight hundred forty-four consecutive patients in the surgical intensive care unit. Overall scores were determined, as well as scores for survivor, nonsurvivor, trauma, nontrauma, postoperative, and nonoperative patient subgroups. MAIN OUTCOME MEASURES: Survival to hospital discharge, and survival compared with published normative APACHE II and III databases. RESULTS: Mean age was 65.1 +/- 0.5 years. Overall mortality was 7.0% in the surgical intensive care unit and 9.1% in the hospital. The relationship between APACHE II and APACHE III scores for individual patients was linear and correlated significantly (P < .0001) (range of correlation coefficients, .72 to .86) overall and in all subgroups. Both scoring systems overestimated our mortality, but estimations made by APACHE III were significantly (P < .01) higher overall and in all subgroups. CONCLUSIONS: In institutions or groups of patients where APACHE II underestimates mortality, APACHE III may be corrective. However, the differences are subtle and may be difficult to detect in smaller studies.