To evaluate the independent effect of feeding on ICU and hospital mortality, we constructed 3 different multivariate logistic models (see “Methods” section). This analysis revealed that regardless of the severity-of-illness score used, early enteral nutrition was consistently associated with a lower risk of hospital mortality . Age, severity of illness, the ICU’s standardized mortality ratio, and source of admission were all significant factors in these analyses. Kaplan-Meier analysis of survival indicated a significant improvement in survival for the patients fed early. The absolute difference in survival between the 2 groups was evident within the first week of mechanical ventilation and remained constant throughout the first 28 days of follow-up after intubation. Cox proportional hazard analyses indicated that after corrections for confounders, early feeding was associated with a 30% to 35% decreased risk of death.