Nutritional support

Feeding tubes

Nasogastric or nasoenteral tubes are placed into the stomach or bowel through the nose. To prevent sinusitis or nasal decubital ulceration, orogastric and oroenteral tubes are also used. A tube placed directly through the skin into the stomach or bowel is called a gastrostomy or jejunostomy .

Nasogastric tubes may be used depending on composition: polyvinyl chloride (PVC) tubes up to 10 days, polyurethane (PUR) up to 6–8 weeks, and silicone tubes for 6 weeks to 3 months.

Blenderized tube feeds

Low-cost blenderized tube feeds (“home brew”) are used in some parts of the world. Macronutrient content is usually highly variable, often conflicting with daily recommendations. Moreover, there is a high contamination risk, physical and chemical instability, and high osmolarity and viscosity, potentially enhancing intolerance.

Commercially available tube feeds

Tube feeds are available as canned powder to dissolve, liquid-containing glass bottles, or self-collapsible packages. Closed feeding systems involve the use of sterile feeding containers that are spiked with feeding sets. Closed feeding systems use connectors that prevent a connection to intravenous lines and reduce the contamination risk, and the hanging time can be increased up to 24 hours. Higher contract price and increased waste of closed systems lead to higher daily costs. However, after adjusting for nursing time and feed contamination costs, the total costs are lower.

Compositions of tube feeding

Nutrient concentrations of formulas vary from 1.0 to 2.0 kcal/mL. Calorically dense formulas are used in patients requiring fluid restriction. Polymeric high-protein formulas have higher protein to nonprotein energy ratios and aid in achieving protein requirements in obese patients while preventing energy overfeeding. There is no evidence that hydrolyzed protein (peptide-based) formulas are superior with respect to tolerance, absorption, or outcome.

Fibers

There are feeds with and without fibers. Sources of fiber in EN include soluble and insoluble fibers. Soluble fibers (e.g., pectin and guar) are fermented by colonic bacteria and enhance colonic sodium and water absorption to treat EN-associated diarrhea. Insoluble fibers (e.g., soy polysaccharide) increase fecal weight and peristalsis and decrease the fecal transit time. Frequently, mixtures of fibers are used.

Disease-specific feeding formulas

Renal formulas have lower protein concentrations and lower potassium and phosphate levels. Hepatic formulas have increased amounts of branched-chain amino acids (e.g., valine, leucine, and isoleucine) and reduced amounts of aromatic amino acids (e.g., phenylalanine, tyrosine, and tryptophan). Diabetic formulas have lower carbohydrate, higher fat content, and variable types of carbohydrates (e.g., oligosaccharides, fructose, cornstarch, and fiber). In pulmonary formulas, some carbohydrate calories are substituted with fat calories to limit carbon dioxide (CO ₂ ) production and improve ventilation. Acute respiratory distress syndrome (ARDS) formulas combine antioxidants with borage and fish oils to supplement gamma-linoleic acid (GLA), and eicosapentaenoic acid (EPA).

In patients with chyle leakage, low-fat or medium-chain triglycerides (MCTs) or enriched feeds are recommended. MCTs do not require lymphatic transport.

Overall, for all disease-specific feeding formulas, strong evidence to improve outcome is lacking.

Allergies

The majority of adverse reactions to foods are nonimmunologic in origin, with lactose intolerance being the most common. Documented food allergies in adults are likely lower than 8%–10%, and most enteral feeds are gluten-free. For specific allergies, soy-free, casein-free, whey-free, and egg-free feeds are available.

Contraindications to enteral nutrition

In general, EN is safe when contraindications are carefully reviewed ( Table 41.2 ). They are related to obstruction, bowel perforation, and ischemia. Many critically ill patients are at risk of splanchnic hypoperfusion caused by circulatory redistribution. It is essential to provide adequate fluid therapy before commencing EN. There is no definition of hemodynamic stability associated with safe enteral feeding. Retrospective data show that early EN in patients on vasopressor treatment is safe and improves outcome. However, aggressive feeding up to target in the phase of shock resuscitation induces bowel ischemia. Patients should have stable blood pressures without the need to increase vasopressors, with acceptable central venous oxygen saturation (ScVO ₂ ) and/or plasma lactate levels or other indicators of adequate blood flow. For most patients this will be within 12–24 hours after admission, still meeting the window of early EN (24–28 hours).

Timing of initiation

By definition, early EN is initiated within the first 24–48 hours after hospital admission. Admission is considered to be the starting moment for ICU patients. Observational studies show that early EN is superior to late initiation (>48 hours) and is therefore recommended in the guidelines. The practical inability to initiate early EN may reflect the severity of illness.

Trophic or permissive underfeeding versus full enteral nutrition

The optimal dose of nutritional support is heavily debated. Permissive underfeeding, trophic, trickle, and hypocaloric feeding are frequently used and confusing.

Permissive underfeeding suggests a lower nutritional intake (e.g., calories, proteins, and micronutrients) is acceptable. Hypocaloric feeding implies that only energy intake is lower. Trophic feeding has no clear definition, although it is accepted to represent an enteral intake of 10–20 mL/h or 500–1000 kcal/day.

Recent trials on trophic, permissive underfeeding and full nutritional support did not demonstrate benefits of either strategy, suggesting that trophic feeding or permissive underfeeding (PERMIT trial)could be sufficient; however, gastrointestinal tolerance was better in the trophic feeding arm in the EDEN trial. Most studies included relatively young, well-nourished (high BMI) patients with low nutritional risk (NUTRIC score <5). In the only trial with a higher nutritional risk, trophic feeding was associated with more infections. Functional outcomes in all studies were not investigated, although long-term outcomes have been shown to be associated with feeding adequacy.

Energy-dense versus routine enteral nutrition

In 3957 patients undergoing mechanical ventilation, the rate of survival at 90 days associated with the use of an energy-dense formulation (1.5 kcal/mL) for enteral delivery of nutrition was not higher than that with routine EN (1.0 kcal/mL). Higher calorie delivery did not affect survival time, receipt of organ support, number of days alive and out of the ICU and hospital or free of organ support, or the incidence of infective complications or adverse events.

Stoppages

Interruptions in enteral feeding for various reasons are frequent, resulting in intakes less than the prescribed amounts. Administering target volumes over 20 hours to circumvent this problem or to increase the infusion rate after stoppages can be recommended.

Gastric residual volume

The increased gastric residual volume (GRV) amount aspirated from the stomach after administration of feed represents feeding intolerance. Risk factors for delayed gastric emptying include gastroparesis, diabetes mellitus, gastric outlet obstruction, (postoperative) ileus, trauma, or sepsis, and medications affecting gastric motor function (e.g., opioids). Aspiration of gastric contents should be prevented, although the GRV thresholds are non–evidence based. Recommendations have gradually increased up to 500 mL in 6 hours and GRV can be reinfused or discarded. Discarded GRV reduces feeding efficacy.

Proton pump inhibitors reduce gastric secretions. Prokinetic medications are used to improve gastric emptying, such as metoclopramide (e.g., 4 × 10 mg and erythromycin 2 × 200 mg) either alone or combined. They are safe when QTC intervals are monitored and used for up to 7 days, and tachyphylaxis is common. Two trials demonstrated that the omission of GRV measurement does not increase the aspiration incidence. Moreover, allowing a large GRV increases the amount of feeding administered. To abandon the GRV measurement is debated, and if all strategies fail, a postpyloric feeding tube can be inserted as often as small intestinal functions are normal.

Postpyloric feeding

Immediate postpyloric feeding may facilitate the early increase of EN, although evidence of superiority is lacking. However, reduced aspiration risk has been clearly demonstrated. Postpyloric tubes are placed by duodenal endoscopy or disposable tubes with fish-eye lenses, using electromagnetic tube placement or self-advancing nasal-jejunal devices. Selection of modality, timing, and success of positioning are related to patient factors, operator experience, and logistics. Criteria to monitor postpyloric feeding are provided ( Box 41.1 ), as monitoring jejunal feeding is complex.