Abstract
BACKGROUND: Despite extensive use of enteral (EN) and parenteral nutrition (PN) in intensive care unit (ICU) populations for 4 decades, evidence to support their efficacy is extremely limited. METHODS: A prospective randomized trial was conducted evaluate the impact on outcomes of intensive medical nutrition therapy (IMNT; provision of >75% of estimated energy and protein needs per day via EN and adequate oral diet) from diagnosis of acute lung injury (ALI) to hospital discharge compared with standard nutrition support care (SNSC; standard EN and ad lib feeding). The primary outcome was infections; secondary outcomes included number of days on mechanical ventilation, in the ICU, and in the hospital and mortality. RESULTS: Overall, 78 patients (40 IMNT and 38 SNSC) were recruited. No significant differences between groups for age, body mass index, disease severity, white blood cell count, glucose, C-reactive protein, energy or protein needs occurred. The IMNT group received significantly higher percentage of estimated energy (84.7% vs 55.4%, P < .0001) and protein needs (76.1 vs 54.4%, P < .0001) per day compared with SNSC. No differences occurred in length of mechanical ventilation, hospital or ICU stay, or infections. The trial was stopped early because of significantly greater hospital mortality in IMNT vs SNSC (40% vs 16%, P = .02). Cox proportional hazards models indicated the hazard of death in the IMNT group was 5.67 times higher (P = .001) than in the SNSC group. CONCLUSIONS: Provision of IMNT from ALI diagnosis to hospital discharge increases mortality.