Search

Hypertonic NG tube feeding in hospitalized patients, whether on the hospital ward or in the ICU, is considered a common etiology of diarrhea. To evaluate the accuracy of this assumption, five normal volunteers, ten hospitalized postoperative patients with head and neck cancer, and 24 ICU patients were given hypertonic (690 mosm), low residue, lactose-free tube feedings starting at 30 kcal/kg.day. There was no prior history of diarrhea in any of the groups studied. There was a significant difference in albumin levels between the three groups, with an average albumin of 2.8 g/dl in the ICU patient group; different from 4.5 g/dl present in both the normal volunteer and non-ICU hospitalized patient groups (general linear models procedure from SAS, p less than .05) (Duncan test). Diarrhea was not present in the normal volunteers or non-ICU patients during the feedings, but did occur in 3/24 ICU patients. This difference was not significant. The three patients with diarrhea had an average albumin level of 3.0 g/dl, while the other ICU patients had an average albumin of 2.7 g/dl. We conclude that hypertonic NG tube feedings do not cause diarrhea in normal volunteers or postoperative head and neck cancer patients. However, in a small statistically insignificant percent of mechanically ventilated ICU patients, this regimen may cause diarrhea although no risk factors can be identified.

We studied the degree of pulmonary and hemodynamic compromise induced by six saline solutions of varying tonicities (0%, 0.225%, 0.45%, 0.9%, 2%, and 3%) instilled down the trachea. We also attempted to resuscitate the dogs after a simulated wet drowning in each of the solutions, with and without chlorine. A control study with similar degree of anoxia but no fluid in the lungs was also performed. For the pulmonary variables of alveolar-arterial oxygen pressure difference, intrapulmonary shunt fraction, and PaO2/fraction of inspired oxygen, the 0.225% NaCl solution was significantly (p less than .001) less injurious to the lungs, especially compared with sterile water. The 0.225% and 0.45% NaCl solutions were the least disruptive to pulmonary function; the pulmonary injury they produced was similar to injury in the control dogs. Chlorine in the solution at one to two parts per million did not influence the data. The 0.9% (isotonic, normal saline), 2%, and 3% NaCl drowning solutions showed little difference from one another in terms of pulmonary injury. They were significantly (p less than .05) better than sterile water and significantly (p less than .05) worse than the 0.225% and 0.45% NaCl solutions. Sterile water, whether chlorinated or not, was the most injurious to the lungs. The hemodynamic effects of the drowning solutions were explainable solely by the effects of anoxia.