BACKGROUND: Findings on the association between early high protein provision and mortality in ICU patients are inconsistent. The relation between early high protein provision and mortality in patients receiving CRRT remains unclear. The aim was to study the association between early high protein provision and hospital and ICU mortality and consistency in subgroups.

METHODS: A retrospective cohort study was conducted in 2618 ICU patients with a feeding tube and mechanically ventilated ≥48 h (2003-2016). The association between early high protein provision (≥1.2 g/kg/day at day 4 vs. <1.2 g/kg/day) and hospital and ICU mortality was assessed for the total group, for patients receiving CRRT, and for non-septic and septic patients, by Cox proportional hazards analysis. Adjustments were made for APACHE II score, energy provision, BMI, and age.

RESULTS: Mean protein provision at day 4 was 0.96 ± 0.48 g/kg/day. A significant association between early high protein provision and lower hospital mortality was found in the total group (HR 0.48, 95% CI 0.39-0.60, p = <0.001), CRRT-receiving patients (HR 0.62, 95% CI 0.39-0.99, p = 0.045) and non-septic patients (HR 0.56, 95% CI 0.44-0.71, p = <0.001). However, no association was found in septic patients (HR 0.71, 95% CI 0.39-1.29, p = 0.264). These associations were very similar for ICU mortality. In a sensitivity analysis for patients receiving a relative energy provision >50%, results remained robust in all groups except for patients receiving CRRT.

CONCLUSIONS: Early high protein provision is associated with lower hospital and ICU mortality in ICU patients, including CRRT-receiving patients. There was no association for septic patients.