BACKGROUND:Diarrhea is an important complication in critically ill patients undergoing enteral feeding. The occurrence of diarrhea may lead to systemic and local complications and negatively impacts on nursing workload and patient's wellbeing. An enteral feeding based on blenderized natural food could be beneficial in reducing the risk of diarrhea. No study has compared natural and commercial enteral feedings in critically ill cardiac surgery patients. OBJECTIVE:The aim of this study was to compare the risk of diarrhea occurrence in two cohorts of patients fed a blenderized natural food diet or commercial enteral feeding preparations, respectively. DESIGN:Retrospective cohort study. SETTING:Cardiac-Surgery Intensive Care Unit of a University Hospital. PARTICIPANTS:Two-hundred and fifteen patients admitted to the postoperative cardiac surgery intensive care unit were included, 103 fed blenderized natural enteral feeding and 112 fed commercial formulas. METHODS:Commercial enteral formulas were delivered by continuous pump administration, while natural enteral feeding by bolus 3 times per day. Diarrhea was documented in the presence of three or more evacuations of loose or watery stool (or an amount above 250 ml) per day. The presence of diarrhea was recorded daily from the beginning to the end of the enteral feeding, up to a maximum of 8 days. The unadjusted time to the first event of diarrhea between the two enteral feeding groups was compared. Adjusted comparison was then performed by fitting a multivariable Cox Proportional-Hazards model, adjusted for potential confounders for diarrhea occurrence (i.e. administration of inotropes, vasopressors, prokinetics, antibiotics, oral nutritional supplements, antifungal agents, sedatives, opioids, probiotics, laxatives). RESULTS:In unadjusted survival analysis the probability of diarrhea was significantly lower in the natural enteral feeding group (log rank test: p = 0.023). In the multivariable model patients in natural enteral feeding cohort showed a non-significant trend towards an almost halved risk of experiencing diarrhea (hazard ratio: 0.584; 95% confidence interval: 0.335-1.018; p = 0.058) compared to those fed commercial enteral feeding. CONCLUSIONS:Administration of a blenderized diet based on natural food for enteral feeding can reduce the incidence of diarrhea in cardiac surgery critically ill patients. This strategy may reduce the risk of diarrhea-associated malnutrition and systemic and local complications, also having a positive impact on nursing workload and patient wellbeing.