Predicting outcome in ICU patients

Abstract

Considerable time and energy has been invested in the conception, modelling and evaluation of sophisticated severity scoring systems for ICU patients. These systems are created to enhance the precise estimation of hospital mortality for large ICU patient populations. Their current low sensitivity precludes their use for predicting out-come for individual ICU patients. However, severity scores can already be valuable for predicting mortality in groups of general ICU patients, and are very useful in the clinical trial setting. Outcome of ICU therapy, however, should incorporate more than mortality. Morbidity, disability and quality of life should also be taken into account; these factors were not taken into consideration in the design of the currently available severity scoring systems. At present, the severity scores have a very limited or no role in clinical decision-making for an individual patient, because they are based on a number of physiological and disease-oriented variables collected during the first 24 h after ICU admission. Future developments and subsequent validation of the dynamic process of clinical, physiological and organ-specific variables could improve the sensitivity and the value of severity scoring. Further collaborative developmental work in this field should be encouraged and supported across Europe and North America.