Abstract

Our study in critical and semi-critical patients compares ß-hydroxy-ß-methylbutyrate (HMB) diet with other available enteral nutrition in terms of improvement of nutritional parameters,hospital stay and mortality.We analyzed 2339 patients admitted between November 2018 and December 2021 in our intensive and intermediate care units.Patients were analyzed in two groups, one which received HMB rich enteral formula (n:292) and the other group other different formula (n:2046).Means of the two groups were compared using t-Student test and Fishers Exact test.Binary logistic regression for outcomes set at mortality or diet selection was run with all the variables.Since the dataset is imbalanced, we performed principal component analysis (PCA) and a neural network (NN) was performed with subsampling to compensate sizes in the classes.Selection of variables was performed using the Boruta method.We found a significant reduction in mortality in the HMB group (18.5%) compared to other formulas (27.9%) (Fisher exact test p<0,0003),Binary logistic regression shows similar results, area under the curve (AUC) 0,72.Patients in the HBM group developed less pressure ulcers (28% vs 40%;p<0,001) and had a reduction in the ratio of C-protein-to-prealbumin (p<0,001).The NN had an AUC of 0,71 in terms of retinol binding protein,lymphocyte count,albumin and prealbumin,which was also confirmed by binary logistic regression analysis.Hospital length of stay was larger in the HMB group related to survival.Our data suggests that HBM-based diet reduce mortality and pressure ulcers prevalence,improving nutritional parameters in critical and semi-critical patients.