Should we believe the benefit of intravenous erythromycin in critically ill adults with gastric feeding intolerance? Reinspecting the pieces of evidence from a series of meta‐analyses

In critically ill adults, unachieved nutrition targets owing to feeding intolerance are associated with poor outcomes. Erythromycin, a macrolide of antibiotics, exerts its gastrointestinal prokinetic as a motilin receptor agonist and subsequently promotes bowel movement. The application of erythromycin has been mentioned in the European Society for Clinical Nutrition and Metabolism (ESPEN) guideline with a grade B recommendation and strong consensus based on significant results from a meta-analysis of only three randomized controlled trials in a Mantel-Haenszel random-effects model.We reinspected the evidence by a random-effects model of meta-analyses through the frequentist approach with the Hartung-Knapp-Sidik-Jonkman (HKSJ) method, trial sequential analysis (TSA), and Bayesian inference.The results analyzed by the HKSJ method showed a risk ratio (RR) of 0.50 (95% confidence interval, 0.19-1.33; I2 = 74.55%). TSA indicated the false-positive results with insufficient cases. Bayesian inference showed a RR of 0.38 (95% credible interval, 0.04-1.62).The effect of intravenous erythromycin in feeding intolerance for critically ill adults remained inconclusive, with very low certainty of evidence. Further large-scale randomized controlled trials with increased rigor are warranted to provide a solid conclusion.