Diarrhea associated with tube feeding in mechanically ventilated critically ill patients.

This study was undertaken to determine the incidence and duration of diarrhea associated with tube feeding in critically ill adult patients who require mechanical ventilation. Of the 73 subjects studied, 63% had diarrhea associated with tube feeding. This incidence is higher than that reported from other studies of critically ill patients who were not mechanically ventilated. Multiple logistic regression analysis revealed that three variables (higher rates of infusion, greater tube-feeding osmolality, and change of tube-feeding product) were statistically significant predictors of diarrhea incidence. Antibiotic use and serum albumin levels were not predictors. In a stepwise multiple regression analysis, the duration of diarrhea was predicted by frequency of diarrhea, within the first 5 days of tube-feeding onset, tube-feeding product osmolality, and rate of tube-feeding infusion. Serum albumin levels and frequency of diarrhea beyond 6 days of tube-feeding onset did not predict duration of diarrhea. Higher osmolality and infusion rates of tube-feeding products did contribute to prediction of both incidence and duration of diarrhea.