RATIONING CRITICAL CARE - WHAT HAPPENS TO PATIENTS WHO ARE NOT ADMITTED

To determine the outcome of patients who are denied ICU admission because of limited resources, we conducted a prospective, nonintervention, cohort survey in a 710-adult bed teaching public hospital. Of 127 critically ill adult medical and surgical patients who were considered ICU candidates by an independent ICU consultant, 63 patients were admitted to the ICU and 64 patients were not admitted. Admitted patients were younger and were more likely to be surgical patients. Raw APACHE II scores and predicted probabilities of death were similar in the two groups, as were other potential confounders. For admitted patients the observed mortality was slightly less than predicted (0.13 versus 0.18 - not significant) for non-admitted patients the observed mortality was 0.49 versus 0.17 (P < 0.0001). The different admission status explained almost all the excess mortality in a logistic regression model. We conclude that critical care delivered in a specialized area by a specialized team can prevent excess mortality in the setting described.