OUTCOME PREDICTION FOR INDIVIDUAL INTENSIVE-CARE PATIENTS - USEFUL, MISUSED, OR ABUSED

Probabilities of hospital mortality provide meaningful information in many contexts, such as in discussions of patient prognosis by intensiv e care physicians, in patient stratification for analysis of clinical trial data by researchers, and in hospital reimbursement analysis by i nsurers. Use of probabilities as binary predictors based on a cut poin t can be misleading for making treatment decisions for individual pati ents, however, even when model performance is good overall. Alternativ e models for estimating severity of illness in intensive care unit (IC U) patients, while demonstrating good agreement for describing patient s in the aggregate, are shown to differ considerably for individual pa tients. This suggests that identifying patients unlikely to benefit fr om ICU care by using models must be approached with considerable cauti on.