Long-term comparison of a routine laboratory parameter-based severity score with APACHE II and SAPS II

Risk score models predicting mortality have tremendous value, but because of the additional effort involved, their clinical use remains low. The aim of this study is to compare three different scores that each requires different levels of effort during admission and throughout treatment: the Acute Physiology and Chronic Health Evaluation II (APACHE II), the Simplified Acute Physiology Score II (SAPS II), and the Dense Laboratory Whole Blood Applied Risk Estimation (DELAWARE) score. Of the three, only the DELAWARE is based solely on routine laboratory parameters.

Prospective data of the three scores were collected for 268 surgical patients admitted to the intensive care unit over 1 year. The predicted hospital mortality and survival were evaluated for the first 14 days.

With a cutoff value of 0.65, the sensitivity of the DELAWARE was 71.6%, the specificity, 92.5%, and the correct classification rate, 87.3%. The APACHE II and SAPS II showed values of 41.2%/96.8%/86.2% and 62.7%/87.1%/82.5%, respectively. The r2 value was 0.884 for the DELAWARE, 0.876/0.814 for the APACHE II and SAPS II. Hospital mortality rate was overestimated by 20% to 65% in all scores. The discriminatory ability of the APACHE II and SAPS II increased throughout the course of treatment.

The routine laboratory-based DELAWARE provides a reliable, valid risk assessment of the surgical intensive care patient at admission. It also provides additional information without added effort or poor interobserver reliability, which leads to better data comparability. We have to state that until now the data have been collected in a single-center and their general validity is therefore limited. By the end of treatment, the SAPS II and APACHE II had increased discriminatory ability and are therefore useful as process parameters.