Comparative effects of continuous, cyclic, intermittent, and bolus enteral feeding on feeding intolerance and clinical outcomes in critically ill adults: a protocol for a systematic review and network meta-analysis

Achoimre

Background Enteral nutrition is recommended for critically ill adults unable to eat adequately. Continuous feeding predominates in intensive care, but cyclic, intermittent, and bolus regimens may confer advantages. Prior reviews combined intermittent and bolus feeding into a single category despite meaningful clinical differences. The comparative effects of all four modes remain uncertain. Objective To compare the effects of continuous, cyclic, intermittent, and bolus enteral feeding on feeding intolerance and clinical outcomes in critically ill adults using a systematic review and network meta-analysis. Methods We will conduct a systematic review and Bayesian network meta-analysis of randomised, cluster-randomised, cross-over, and quasi-randomised controlled trials in critically ill adults receiving gastric enteral nutrition. The intervention network will comprise continuous, cyclic, intermittent, and bolus feeding, classified from procedural details of the administered regimen rather than the terminology used by trial authors. MEDLINE, Embase, CENTRAL, CINAHL, Web of Science, Scopus, SinoMed, CNKI, Wanfang, VIP, ClinicalTrials.gov, and the WHO International Clinical Trials Registry Platform will be searched from inception without language restrictions. Two reviewers will independently select studies, extract data, and assess risk of bias using Cochrane RoB 2. The primary outcome will be feeding intolerance as defined by each trial, pooled only where definitions are clinically compatible. Key secondary outcomes, which will be presented in summary-of-findings tables, will be diarrhoea, vomiting, aspiration pneumonia, the proportion of prescribed energy delivered, intensive-care-unit mortality, and duration of mechanical ventilation. Additional secondary outcomes will include further gastrointestinal, respiratory, nutritional, glycaemic, clinical-course, and safety outcomes. Feeding intolerance is the sole confirmatory outcome; all secondary outcomes and all subgroup, meta-regression, and sensitivity analyses are exploratory and will be interpreted according to prespecified rules. The false discovery rate will be controlled using the Benjamini–Hochberg procedure within prespecified families of comparisons. When the evidence network is connected and transitivity is plausible, Bayesian random-effects network meta-analyses will estimate odds ratios, mean differences or standardised mean differences, and hazard ratios with 95% credible intervals, using arm-level likelihoods where arm-level data are available. Risk ratios from a binomial-log model will be examined in sensitivity analysis. Otherwise, random-effects pairwise meta-analyses (restricted maximum-likelihood estimation with the Hartung–Knapp–Sidik–Jonkman adjustment) or synthesis without meta-analysis will be undertaken. Transitivity, heterogeneity, global and local inconsistency, convergence, small-study effects, and robustness to alternative node classifications and prior distributions will be evaluated. Certainty of evidence for the primary and key secondary outcomes will be assessed using CINeMA for network estimates and GRADE for pairwise comparisons. Discussion This review will distinguish four clinically meaningful enteral feeding modes and integrate direct and indirect evidence while explicitly addressing uncertain classification and sparse network nodes. The findings may support nursing protocols, nutrition practice, and future critical-care nutrition research.