Evaluation of single intensive care unit performance by simplified acute physiology score II system

To evaluate effectiveness and quality of care in a single intensive care unit (ICU) by the Simplified Acute Physiology Score II (SAPS II).

A prospective study included 395 patients from the ICU at Rijeka University Hospital, Croatia. The sum of the SAPS II points was used for calculating predicted mortality for each patient. The observed death rate was compared with predicted mortality calculated by SAPS II system. The ability of the SAPS II prognostic system to predict probability of hospital mortality was assessed with discrimination (receiver operating characteristic [ROC] curve) and calibration (Hosmer-Lemeshow test) measures.

The SAPS II score on the first ICU day was low (median, 20; range, 3-83). SAPS II system showed a good ability to separate the patients predicted to live from those predicted to die, as shown by an area under the ROC curve of 0.827. The calibration curve demonstrated under-prediction of the actual death rate (Hosmer-Lemeshow goodness-of-fit test, C=22.961; df=8; P=0.003). The observed mortality was higher than predicted (observed-to-predicted ratio was 1.6).

SAPS II system is a useful tool for the assessment of ICU performance. This system demonstrated a good ability of discrimination, but an under-prediction of the actual mortality rate in our ICU.