Enteral Feeding and Bowel Necrosis: An Uncommon but Perplexing Problem

Abstract

This issue of Nutrition in Clinical Practice contains 3 articles focusing on an important problem. I have seen several episodes of small bowel necrosis in patients who seemed indistinguishable from dozens of other patients who tolerated feedings without major problems. This devastating complication of small bowel (and occasionally proximal colonic) necrosis is one that is of concern to all clinicians administering enteral feeding to critically ill or injured patients. Unfortunately (or fortunately), so few cases occur, in such a sporadic manner, that it has been impossible to definitively identify the factor or factors that increase the patient’s risk. Early in my career, I advanced tube feedings aggressively, even in a few hemodynamically unstable patients. I’d become confident after successful use of intragastric feeding in burn patients who required vasopressors for blood pressure and tolerated feedings with low residuals and little gastrointestinal tract intolerance. One advantage of intragastric feedings is that high gastric residuals result if the small intestine doesn’t tolerate feeding. This protective mechanism is bypassed with administration directly into the small bowel. Reflux of tube feedings from the intestine into the stomach is an important sign of gastrointestinal intolerance that warrants discontinuation of tube feedings. However, in some of the reported cases, such warning signs have not been evident. The most common “party line” I’ve heard and I’ve expressed is to avoid direct small bowel feedings in hemodynamically unstable patients, especially those requiring vasopressors. Clearly, as the authors of these articles point out, not all data substantiate a poor physiologic response when the gut is fed in low-flow states. In fact, some studies show that enteral feeding may actually help. The difficulty is interpreting results from the different models of endotoxin, complete mesenteric occlusion, or other artificially induced low-flow states. The fact that some nutrients are deleterious in some circumstances but beneficial in others, or that some pressor agents improve splanchnic blood flow or increase mucosal blood flow, is difficult to apply to an individual patient. One clinical study showed that enteral feeding actually improved flow in select patients with hemodynamic instability. Even more confusing is the recognition that episodes of necrosis can occur after patients have been fed successfully for 5 to 10 days. My few personal episodes occurred within 24 hours of instituting enteral feeding; the patients became tachycardic and hypotensive and spiked a high fever. I know of 1 patient who had been successfully fed (I’m told) for a number of days but suddenly deteriorated. In the operating room, the necrotic small intestine was filled with a pasty residual from tube feedings. The fluid had been absorbed from the feedings and formed a cast of the intestine. The 3 articles presented in this issue review the existing knowledge regarding this uncommon (0.3% to 0.7%) but devastating problem. Unfortunately, no one has the answer, and until we understand this process, I still believe the wisest approach is to delay direct small bowel feedings until there is evidence of adequate splanchnic perfusion and patients are fully resuscitated. Some hemodynamically unstable patients will tolerate feeding, a good example being the burn patients. Unfortunately, it is difficult to predict who will have significant intolerance to small bowel feeding, and by the time it becomes evident, it is too late. Dr. Zaloga’s comment regarding the paucity of this complication in medical patients is very important. It may be that manipulation of the gastrointestinal tract during laparotomy, the transmural placement of the small bowel tube, the presence of intra-abdominal pathology, or some other issues are important etiologic factors. These issues currently are not being tested in animal models. Many randomized prospective studies have demonstrated that there is a significant benefit when nutrients are delivered into the gut of critically ill patients that is not gained when patients are fed IV Correspondence: H4/736 Clinical Science Center, 600 Highland Avenue, Madison, WI 53792.