Outcome prediction in intensive care: Results of a prospective, multicentre, Portuguese study

Objective: To compare the performance of the New Simplified Acute Physiology Score (SAPS II) and Acute Physiology and Chronic Health Evaluation (APACHE) II in an independent database, using formal statistical assessment. Design: Analysis of the database of a multicentre, prospective study. Setting: 19 intensive care units (ICUs) in Portugal. Patients: Data for 1094 patients consecutively admitted to the ICUs were collected over a period of 4 months. Following the original SAPS II and APACHE II criteria, the analysis excluded patients younger than 18 years of age, readmissions, acute myocardial infarction, burns, patients in the post-operative period after coronary artery bypass surgery, and patients with a length of stay in the ICU of less than 24 h. The group analysed comprised 982 patients. Interventions: Collection of the first 24 h admission data necessary for the calculation of SAPS II, APACHE II, Therapeutic Intervention Scoring System (TISS), Simplified TISS, organ system failure and basic demographic statistics. Vital status at discharge from the hospital was registered. Measurements and results: In this cohort, discrimination was better for SAPS II than for APACHE II (SAPS II: area under the receiver operating characteristic curve 0.817, standard error 0.015; APACHE II: 0.787, 0.015;p < 0.001); however, both models presented a poor calibration, with significant differences between observed and predicted mortality (Hosmer-Lemeshow goodness-of-fit tests H and C, p < 0.001). In a stratified analysis, this study was unable to demonstrate any definite pattern of association between the poor performance of the models and specific subgroups of patients except for the most severely ill patients, where both models overestimated mortality. Conclusions: SAPS II performed better than APACHE II in this independent database, but the results do not allow its use, at least without being customised, to analyse quality of care or performance among ICUs in the target population.