The impact of high versus standard enteral protein provision on functional recovery following intensive care admission: A pre-planned Bayesian analysis of the PRECISe trial.

Summary

In a pre-specified Bayesian re-analysis of the PRECISe RCT (n=935), targeting 2.0 g/kg/day protein showed a 0% posterior probability of improving EQ-5D-5L and only an 8% chance of any mortality benefit, with a 47% posterior probability of clinically important harm (>5% absolute risk increase) for 60-day mortality. Findings were robust across multiple prior assumptions.

Key Findings

• Posterior probability of benefit on EQ-5D-5L with high protein was 0%.
• For 60-day mortality, high protein had only an 8% chance of any benefit and a 47% probability of clinically important harm (>5% absolute risk increase).
• Results were robust across sensitivity analyses using different priors.

Clinical Implications

Avoid routine high-protein targets (2.0 g/kg/day) in the early ICU course; consider standard protein targets (~1.3 g/kg/day) while individualizing based on catabolic state and tolerance. Discuss potential harm with multidisciplinary nutrition teams.

Limitations

• Secondary analysis; original trial not powered for all functional outcomes under a frequentist framework
• Potential sensitivity to prior assumptions despite comprehensive sensitivity analyses

Future Directions

Prospective, adaptive trials testing personalized protein targets by phenotype (e.g., sarcopenia, AKI) and timing, with Bayesian decision rules and patient-centered outcomes.