Comparison of three severity scores for critically ill cancer patients

Objective: To compare three scoring systems, the Acute Physiology and Chronic Health Evaluation (APACHE) II, the Simplified Acute Physiology Score (SAPS) II and a modified Mortality Probability Model II (ICU cancer mortality model, ICMM) for their prognostic value for mortality during hospital stay in a group of cancer patients admitted to a medical ICU. Design: Prospective cohort study. Setting: Medical ICU of a tertiary care hospital. Patients: Two hundred forty-two consecutive cancer patients admitted to the ICU. Measurements and results: Variables included in APACHE II, SAPS II and the ICMM scores as well as demographic data were assessed during the first 24 h of stay in the ICU. Hospital mortality was measured; it was 44%. Calibration for all three scoring systems was acceptable, SAPS 11 yielded a significantly superior discrimination between survivors and non-survivors. The areas under the receiver operating characteristic curves were 0.776 for APACHE II, 0.825 for SAPS II and 0.698 for the ICMM. Conclusion: The SAPS II was superior to APACHE II and ICMM. The newly developed ICMM does not improve mortality prediction in critically ill cancer patients.