Provision of protein and energy in relation to measured requirements in intensive care patients.

ABSTRACT

BACKGROUND & AIMS: Adequacy of nutritional support in intensive care patients is still a matter of investigation. This study aimed to relate mortality to provision, measured requirements and balances for energy and protein in ICU patients. DESIGN: Prospective observational cohort study of 113 ICU patients in a tertiary referral hospital. RESULTS: Death occurred earlier in the tertile of patients with the lowest provision of protein and amino acids. The results were confirmed in Cox regression analyses which showed a significantly decreased hazard ratio of death with increased protein provision, also when adjusted for baseline prognostic variables (APACHE II, SOFA scores and age). Provision of energy, measured resting energy expenditure or energy and nitrogen balance was not related to mortality. The possible cause-effect relationship is discussed after a more detailed analysis of the initial part of the admission. CONCLUSION: In these severely ill ICU patients, a higher provision of protein and amino acids was associated with a lower mortality. This was not the case for provision of energy or measured resting energy expenditure or energy or nitrogen balances. The hypothesis that higher provision of protein improves outcome should be tested in a randomised trial.