Reporting of nutritional screening, status and intake in trials of nutritional and physical rehabilitation following critical illness: a systematic review.

Background: Surviving critical illness leads to prolonged physical and functional recovery with both nutritional and physical rehabilitation interventions for prevention and treatment being investigated. Nutritional status and adequacy may influence outcome, but no consensus on which nutritional-related variables should be measured and reported in clinical trials exists.

Objective: To undertake a systematic review investigating the reporting of nutritional screening, nutritional status and nutritional intake/delivery in randomised controlled trials (RCTs) evaluating nutritional and/or physical rehabilitation on physical and functional recovery during and following critical illness.

Methods: Five electronic databases (MEDLINE, Web of Science, EMBASE, CINAHL, Cochrane) were searched (last update 9 August 2023). Search terms included both free text and standardised indexed terms. Studies included were RCTs assessing nutritional and/or physical interventions either during or following intensive care unit (ICU) admission in adults (≥18 years old) with critical illness, and who required invasive mechanical ventilation for any duration during ICU admission. Study quality was assessed using the Cochrane Collaboration Risk of Bias tool for RCTs and descriptive data synthesis was performed and presented as counts (%). Meta-analysis was not performed as this systematic review investigated measurement and reporting, and not effectiveness endpoints.

Results: 123 RCTs (30 nutritional, 87 physical function and 6 combined) were included. At least one nutritional variable was measured and/or reported in 99 (80%) of the studies including BMI (n=69), body weight (n=57), nutritional status (n=11), nutritional risk (n=10), energy delivery (n=41), protein delivery (n=35), handgrip strength (n=40) and other nutritional-related muscle variables (n=41). Only three studies were considered to have a low risk of bias in all categories.

Conclusion: Few RCTs of physical rehabilitation measure and report nutritional or related variables. Future studies should measure and report specific nutritional factors that could impact physical and functional recovery to support interpretation where studies do not show benefit. A protocol was pre-registered on PROSPERO (CRD42022315122).