Defining and Assessing Tolerance in Enteral Nutrition![]() ![]()
* Section on Trauma and Surgical Critical Care, Correspondence: Andrew C. Bernard, MD, Section on Trauma and Surgical Critical Care, Department of Surgery, University of Kentucky, C224 Division of General Surgery, 800 Rose Street, Lexington, KY 40536-0298. Electronic mail may be sent to acbern00{at}uky.edu. Nutrition support has become widely recognized as an essential component of optimal care for acutely ill patients. Enteral nutrition is preferred over parenteral routes when possible. However, prescribed enteral nutritional regimens are sometimes met with side effects and even complications. These adverse events have been collectively termed "intolerance," and forms of intolerance occur in a spectrum from bothersome at least to life threatening when most severe. Here we discuss nutritional access and its maintenance, introduce and define intolerance, and then review the current literature with regard to principal forms of enteral nutrition intolerance. Over the last several decades, nutrition has been recognized as an increasingly important factor in the treatment of acutely ill patients. Feeding the gastrointestinal tract had at one time been considered unsafe and impractical in this population. Parenteral nutrition (PN) was widely used and became accepted as standard of care in the 1980s. Advantages included rapid central venous access, safe and sterile formulas, and ease of fluid and electrolyte control. PN use increased despite very little clinical data to support efficacy. More recently, a wealth of clinical literature has established that enteral nutrition (EN) is the preferred means of nutrition support in hospitalized patients. EN maintains gut weight, reduces intestinal atrophy, and reduces hospital-associated infectious complications. The burn and trauma patient populations have demonstrated the most clinically significant benefits of EN, with several studies documenting decreased overall infections rates, ventilator and intensive care unit (ICU) days, costs, and mortality.1–3 These benefits have been attributed to maintenance of the gut immune system in the GI tract, attenuation of the postinjury stress response, increases in mesenteric blood flow, prevention of gastric and enteral mucosal atrophy (translocation of enteric bacteria or endotoxins), and the provision of essential nutrients such as glutamine that are not stable in PN. Despite these well-documented benefits, there are cases where EN is not applicable. Oral intake is avoided in patients with gastrointestinal obstruction. Patients with absorption defects, such as inflammatory bowel disease, may be able to use EN in the form of semielemental tube feeding, but if the disease is severe, even this may not be tolerated. Patients with gastrointestinal fistulas, depending on the location, may benefit from low-volume EN, but if fistula outputs are high, PN may be necessary. As will be discussed in further detail later, hemodynamically unstable patients, particularly those requiring the use of vasopressor agents, are at increased risk of nonocclusive mesenteric ischemia, and EN should be held in these cases until the patient is more stable. In general, however, the presence of a functional gastrointestinal tract in a stable patient indicates that EN should be used rather than PN. Despite this, some clinicians remain reluctant to use EN, particularly in the critically ill. Difficulty accessing the small bowel and maintaining tube patency can deter some clinicians and nursing staff from using EN. Other complications, such as gastric reflux, aspiration, diarrhea, severe abdominal distention, or nonocclusive bowel ischemia have been reported in as many as 62% of patients receiving EN.4 Although complete cessation of EN because of such complications is uncommon, these side effects often significantly affect achieving adequate protein and caloric goals. Delayed gastric emptying, aspiration, and other EN complications are discussed elsewhere in this issue. In this article, we will address the challenges of enteral nutritional access, define tolerance to enteral feeding, and describe the assessment and management of EN intolerance.
EN is clearly superior to PN when a functional gastrointestinal tract is present. Historically, obtaining appropriate enteral access was often the initial obstacle to providing EN to the critically ill patient. Currently, with an assortment of available feeding tubes and numerous successful methods of bedside placement, achieving EN access is now routinely feasible and in many facilities is considered the standard of care in the ICU.5–7 If bedside placement fails to achieve small bowel access, fluoroscopic, endoscopic, and operative interventions can be used. Once the appropriate enteral access device is selected and placed in the proper position, the patency must be maintained. Feeding tube occlusion impedes optimal EN therapy and medication delivery. Administration of crushed medications is the most common culprit for feeding tube occlusion. Several nursing interventions should be established to maintain patency. Tubes should be routinely flushed with water. These flushes can also contribute to the patient's daily free water requirements. Tablets should be avoided for a number of reasons. Some specialty tablet formulations (ie, extended or delayed release, enteric-coated, or osmotic pump formulations) are not intended to be crushed or altered. Altering these medications can result in a toxic or subtherapeutic response or feeding tube occlusion. Instead, medications should be administered as liquid formulations or diluted syrups. Thick or gritty suspensions, such as antacids, can contribute to tube occlusion and should be diluted with water. Administration of fiber supplements such a psyllium or polycarbophil is another common cause of tube blockage. Use of a fiber-containing enteral product provides a better alternative to the administration of a bulk-producing laxative and reduces the risk of tube occlusion. If a feeding tube becomes occluded with medication, the safest and most effective way to restore patency is the administration of tepid water through the tube. Tube feeding clogs may be dissolved with pancrelipase (Pancrease, McNeil Pharmaceutical, Spring House, PA).8 A commercial product (Corpak Clog Zapper Enteral Feeding Tube Declogging System, VIASYS Healthcare Medsystems, Wheeling, IL) is also available and contains a "multi-enzyme cocktail" that is designed to break up formula clogs. As with many predicaments, an ounce of prevention is worth a pound of Pancrease.
Intolerances can encompass a wide range of problems. Physical intolerances such as abdominal distention, vomiting, nausea, bloating, gastrointestinal bleeding, and perforation represent mechanical complications and obviously interfere with achieving EN goals. Metabolic complications, such as fluid accumulation, electrolyte derangements, and hyperglycemia can also limit enteral feeding. Most commonly, the intolerance is associated with absorption or elimination, such as aspiration, delayed gastric emptying, gastric reflux, or diarrhea. In general, tolerance can be described as providing EN nutrition without experiencing one of the above complications or "ill effect."9
Diarrhea is the most frequent form of EN intolerance. Although the term is frequently used, a standard definition of diarrhea has not been well established. In fact, a recent paper reported over 33 different definitions of this term.10 Criteria for diarrhea may be considered as an increased number of daily stools, liquid or loose consistency, color change, duration of >2 to 3 days, and occasionally as stool weight. Because the definition of diarrhea is vague and often subjective, reporting is inconsistent among nursing and medical staff. In our institution, we define diarrhea as 3 or more liquid stools per day or 500 mL daily for 2 consecutive days. Several causes of diarrhea should be considered when treating the patient receiving EN (Fig. 1). Although Clostridium difficile infection should be ruled out if the patient has received a course of antibiotics, pseudomembranous colitis is not the most frequent cause of diarrhea in patients receiving EN.11,12 Hypertonic enteral formulas have also been erroneously implicated as causing diarrhea. In fact, electrolyte supplements, elixirs, and other medications are commonly far more hypertonic than standard tube feeding preparations and are a far more likely cause of diarrhea. Many liquid medications and water-insoluble drugs are mixed or suspended in sorbitol. Although 8 to 10 g/day of sorbitol can result in cramping and abdominal bloating, higher doses (10 to 15 g/day) frequently result in diarrhea. In patients receiving enteral medications, sorbitol content should be minimized when diarrhea is present. Medications commonly associated with diarrhea are shown (Table 1). Some authors have suggested that hypoalbuminemia, common in critically ill patients, can result in significant bowel edema with impaired absorption and subsequent diarrhea.13 High infusion rates, particularly if bowel edema and impaired absorption are present, can exacerbate stool output. Higher-density formulas administered at a lower rate can rectify this problem.
Diarrhea in the patient receiving EN should be evaluated and treated in a stepwise approach before using PN. When use of laxatives, sorbitol-containing medication, stool softeners, or hypertonic or electroyte elixirs has been reduced or discontinued and infectious causes have been eliminated, the enteral formulation can be modified to include fiber or be changed to a semielemental product to facilitate absorption. If these initial therapies fail, the tube-feeding rate can be decreased and antidiarrheal agents can be added. In patients who have received antibiotics, the administration of probiotics may restore some balance to the indigenous gut flora, but data supporting this practice in adults are limited.14
The gut is immunologically and metabolically active, especially in critical illness. Failure or malfunction of the mesenteric circulation, mesenteric lymph, and the mucosal gastrointestinal barrier is considered an important contributor to systemic inflammation and multiple-organ failure.15,16 Shock from any cause, combined with sepsis and vasoactive drugs, can lead to inadequate gut mucosal perfusion, even after resuscitation.17 Maintenance of gut mucosal perfusion and integrity are mainstays of modern critical care, and an accepted strategy for achieving these goals is the early administration of EN.3,18,19 Although EN is well tolerated in most patients, intestinal mucosal ischemia and necrosis has been reported with and attributed to EN.20,21 This ischemic necrosis does not involve occlusion of major vasculature and therefore has been termed nonocclusive bowel necrosis. The proposed mechanisms for this include increased metabolic demand of gut mucosa by intraluminal nutrients in the presence of reduced or threatened mucosal perfusion in the critically ill patient. It has also been hypothesized that bowel distention may contribute to this mucosal ischemia and that luminal stasis might permit accumulation of bacterial toxins.
Significant animal data have established that EN improves gastrointestinal
perfusion, reduces splanchnic ischemia, improves mucosal microcirculation, and
reduces mucosal-to-arterial pCO2 gradient in various models of
sepsis and hemorrhagic
shock.21 The effect
on mucosal blood flow may be nutrient
specific.22,23
Clinical data show that although intestinal oxygen tension routinely decreases
after elective upper gastrointestinal surgery, patients fed EN enriched with
arginine and A precise method of clinical detection of nonocclusive bowel necrosis is unavailable, and monitoring for and detecting mucosal hypoperfusion is difficult. Moore and Weisbrodt17 reported gastric tonometry to be valuable in deciding whether to initiate EN and whether tolerance is adequate to permit advancement toward a nutritional goal. Tonometry uses intramucosal CO2 to measure intramucosal pH. A low intramucosal pH has been shown to predict an adverse outcome. Abdominal distention, reflux, or diarrhea may be early signs that warrant reduction or cessation of enteral feedings. EN should be withheld from patients who are thought to be at risk for this rare but devastating complication. We recommend that when these patients are clinically stable, EN should be reinitiated carefully, with consideration for elemental feeding. A general rule is to initiate enteral feeding only after resuscitation targets (ie, blood pressure, cardiac output, and oxygen delivery) have been met; however, clinically useful precise indicators are presently unavailable.
Feeding intolerance may initially present as abdominal distention. EN intolerance has been shown to be an early sign of sepsis in burn patients. Unfortunately, clinical signs indicating gut function are poor, including the passage of flatus or the presence of bowel sounds. Delayed gastrointestinal transit does occur in some patients, particularly in the critically ill, but mechanisms are not well understood. Animal data indicate that impaired enteric nerve function and smooth muscle of the intestinal wall are involved. Human studies measuring gut motor activity using manometry showed that as many as 50% of patients undergoing a secondary laparotomy have migrating motor complexes consistent with the fasted state and that those patients tolerated early EN.26 However, patients without these normal motor complexes were more likely to have EN intolerance. Unfortunately, routine gut manometry is not clinically feasible, but it may be a valuable tool in subsequent human studies of EN intolerance. In the critically ill patient, however, there is often reluctance on the part of the clinician to initiate EN. Absence of bowel sounds, abdominal distension, and fluid overload contributing to bowel edema are frequently cited concerns. Studies have indicated, however, that tolerance of EN and attainment of nutritional goal can be achieved in as many as 85% of patients using a standardized protocol or regimen.27 Recent laparotomy and even intestinal anastomoses are not themselves contraindications to early EN, particularly at low rates.28 Melbert et al29 reported achieving a regular diet in 3.5 days after colon resection using a critical pathway, and Law et al30 reported early initiation of enteral diet in the elderly to be safe after laparoscopic colectomy. EN has even been successfully used in patients requiring the use of laparostomy (Tsuei BJ, Lexington, Kentucky, unpublished data).
Overfeeding can result in serious metabolic complications in EN. Hyperglycemia, caused by endorgan resistance, is common in most critically ill patients and can be exacerbated if the patient is fed in excess of energy expenditure. Control of glucose levels, often requiring the use of insulin infusions, has been associated with a decrease in mortality in the critically ill.31 Chronic overfeeding, especially of carbohydrates, can result in hypercapnia because of increases in CO2 production. This is rarely clinically significant except when it complicates weaning from mechanical ventilation. Indirect calorimetry can assist in determining appropriate caloric intake and respiratory quotient. In some cases, altering the caloric intake or nutrient distribution may improve ventilator weaning for this select patient population.32
In a landmark study published in 1950, Keys et al33 described individuals who had tolerated 6 months of starvation but experienced cardiovascular dysfunction, sometimes severe, in the refeeding and recovery phase. Electrolyte disturbances, most notably hyperglycemia, hypokalemia, hypomagnesemia, and severe hypophosphatemia, can be fatal.34 Although historically described when victims of war and imprisonment with severely depleted nutrient stores were given food and nutrition, refeeding syndrome is often underappreciated in the industrialized world today. This metabolic phenomenon is common in patients with anorexia nervosa, chronic vomiting, malignancy, and alcoholism. In addition, the hypermetabolic/hypercatabolic patient that has not been fed for at least 48 hours may have similar metabolic derangements. Marik and Bedigian35 prospectively studied 62 ICU patients and reported 34% experienced hypophosphatemia after at least 48 hours of starvation. They reported the only significant predictor of the hypophosphatemia was a serum prealbumin concentration <110 g/L. They also reported longer ventilator and hospital days for the patients that experienced hypophosphatemia. Although hypokalemia can occur as rapidly as hypophosphatemia, potassium levels often are more routinely monitored and replaced. Severe hypokalemia, hypophosphatemia, and hypomagnesaemia can result in potentially lethal neurologic and cardiovascular complications.36 The sodium and fluid retention that commonly occurs with refeeding syndrome can result in significant weight gain, edema, and volume overload. As most manifestations of refeeding syndrome appear within the first week of initiating EN, feeding should be increased gradually over a period of several days to avoid these complications. Glucose and electrolyte monitoring and correction should be aggressively pursued.
EN is the preferred method of providing nutrition support in most hospitalized patients. Bedside methods of achieving EN access have improved and can be quite effective. Although there are some clinical entities for which EN may not be applicable, in general, the presence of a functional gastrointestinal tract in the hemodynamically stable patient should prompt the clinician to consider EN over the parenteral route. Intolerance of EN is commonly manifested as abdominal distension, reflux, or diarrhea and may be a sign of early sepsis or clinical deterioration. Ensuring proper tube position, choosing the appropriate enteral formulation, and avoiding medications that contribute to side effects can increase patient tolerance to enteral feeding. Future clinical investigation must target monitors of gut motility, absorption, and blood flow that are easily applied at the bedside.
Nutrition in Clinical Practice, Vol. 19, No. 5,
481-486 (2004) This article has been cited by other articles:
|
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||


500 mL daily for 2 consecutive days. 
-3 fatty acids recovered intestinal oxygen tension faster
than those fed standard diets or not at
all.
