APPROACH TO THE PATIENT: Requirements for Specialized Nutritional Support INDICATIONS FOR SPECIALIZED NUTRITIONAL SUPPORT
Approximately one-fifth to one-quarter of patients in acute-care hospitals suffer from at least moderate protein-energy malnutrition (PEM), the defining features of which are malnutrition-induced weight loss and skeletal muscle atrophy. Usually, but not always, other features further compromise clinical responses; these features include a subnormal adipose tissue mass, with the accompanying adverse consequences of weakness, skin thinning, and breakdown; reduced ventilatory drive; ineffective cough; immunodeficiency; and impaired thermoregulation. Commonly, PEM is already present at the time of hospital admission and remains unimproved or worsens during the ensuing hospital stay. Common reasons for PEM worsening during hospitalization are refusal of food (because of anorexia, nausea, pain, or delirium), communication barriers, an unmet need for hand-feeding of patients with physical or sensory impairment, disordered or ineffective chewing or swallowing, and prolonged periods of physician-ordered fasting—all potentially taking place in a context of caregiver unawareness and inattention. Most patients who are suffering from in-hospital PEM do not, or ought not, to require SNS. A large proportion of these patients can be expected to improve with appropriate management of their primary disease. Others have a terminal disease whose downward course will not be altered by SNS. In yet other cases, the PEM is sufficiently mild that the benefits of SNS are exceeded by its risks. For patients who fall into this last category, the correct approach is to intensify and/or modify the patient’s oral nutrition as directed by the unit dietitian.
PEM is often classified as minimal, moderate, or severe on the basis of weight for height (body mass index, BMI) and percentage of body weight recently lost. As shown in Table 98e-1, the BMI (when corrected for abnormal extracellular fluid accumulation) is a crude but useful indicator of PEM severity. Note, however, that obesity does not preclude moderate or severe PEM, especially in older or bedridden patients; indeed, obesity can mask the presence of PEM if the patient’s muscle mass is not specifically examined.
The decision to implement SNS must be based on the determinations (1) that intensified or modified oral nutrition has failed or is impossible, impractical, or undesirable; and (2) that SNS will increase the patient’s rate and likelihood of recovery, reduce the risk of infection, improve healing, or otherwise shorten the hospital stay. In chronic-care situations, the decision to institute SNS is based on the likelihood that the intervention will extend the duration or quality of the patient’s life. An algorithm for determining when to use SNS is depicted in Fig. 98e-1.
The decision to enhance oral nutrition or—that attempt failing—to resort to SNS is based on the anticipated consequences of non-intervention. The mnemonic “in-in-in” (for inanition-inflammation-inactivity) can serve as a reminder of the three main factors that come into play when deciding whether or not it is acceptable to withhold SNS from a patient with PEM. Inanition
Key issues include whether normal food intake is likely to be impossible for a prolonged period and whether the patient can tolerate prolonged starvation. A previously well-nourished person can tolerate ∼7 days of starvation without harm, even in the presence of a moderate systemic response to inflammation (SRI), whereas the degree of tolerance to prolonged starvation is much less in patients whose skeletal muscle mass is already reduced, whether from PEM, from the muscle atrophy of old age (sarcopenia), or from muscle atrophy due to neuromuscular disease. Excess body fat does not exclude the possibility of coexisting muscle atrophy from any of these causes. In general, unintentional weight loss of >10% during the previous 6 months or a weight-to-height ratio that is <90% of standard, when associated with physiologic impairment, crudely predicts that the patient has moderate PEM. Weight loss >20% of usual or <80% of standard makes severe PEM more likely. Inflammation
The anorexia that invariably accompanies the SRI reduces the likelihood that a patient’s nutritional goals will be achieved by intensifying or modifying the diet, by providing counseling, or by hand-feeding. Furthermore, the protein-catabolic effects of the SRI accelerate skeletal muscle wasting and substantially block normal protein-sparing adaptation to protein and energy starvation. Inactivity
A nutritional red flag should be raised over every acutely ill patient who remains bedridden or inactive for a prolonged period. Such patients commonly manifest muscle atrophy (due to nutritional deficiencies and disuse) and anorexia with inadequate voluntary food intake.
Once it has been determined that a patient has significant—and, in particular, progressive—PEM despite meaningful efforts to reverse it by modifying the diet or the way food is provided, the next step is to decide whether SNS will have a net positive effect on the patient’s clinical outcome. The pathway to the end stage of most severe chronic diseases leads through PEM. In most patients with end-stage untreatable cancer or certain end-organ diseases, SNS will neither reverse PEM nor improve the quality of life. Provision of food and water is commonly regarded as an aspect of basic humane care; in contrast, enteral and parenteral SNS is a therapeutic intervention that can cause discomfort and pose risks. As with other life-support interventions, the discontinuation of enteral or parenteral SNS can be psychologically difficult for patients, their families, and their caregivers. Indeed, the difficulty can be greater with SNS than with other life-support interventions because the provision of food and water is often considered equivalent to comfort care. In such difficult, near end-of-life situations, it is prudent to explicitly state the treatment goals at the outset of a course of SNS therapy. Such clarity can smooth the way for subsequent appropriate discontinuation in those patients whose prognosis has become hopeless.
After the decision has been made that SNS is indeed appropriate, the next determinations are the route of delivery (enteral versus parenteral), timing, and calculation of the patient’s nutritional goals. Although enteral SNS is the default option, the choice of optimal route depends on the degree of gut function as well as on available technical resources.
Both the choice of route and the timing of SNS require an evaluation of the patient’s current nutritional status, the presence and extent of the SRI, and the anticipated clinical course. Severe SRI is identified on the basis of the standard clinical signs of leukocytosis, tachycardia, tachypnea, and temperature elevation or depression. Serum albumin is a negative acute-phase protein and hence a marker of the SRI. More severe hypoalbuminemia is a crude indicator of greater SRI severity, but this condition is almost certainly worsened by concurrent dietary protein deficiency. Despite the importance of adequate protein provision to patients with the SRI, no amount of SNS will raise serum albumin levels into the normal range as long as the SRI persists.
The SRI can be graded as mild, moderate, or severe. Examples of a severe SRI include (1) sepsis or other major inflammatory diseases (e.g., pancreatitis) that require care in the intensive care unit; (2) multiple trauma with an Injury Severity Score >20–25 or an Acute Physiology and Chronic Health Evaluation II (APACHE II) score >25; (3) closed head injury with a Glasgow Coma Scale <8; and (4) major third-degree burns of >40% of the body surface area. A moderate SRI occurs with less severe infections, injuries, or inflammatory conditions like pneumonia, uncomplicated major surgery, acute hepatic or renal injury, and exacerbations of ulcerative colitis or regional enteritis requiring hospitalization.
Patients with a severe SRI require the initiation of SNS within the first several days of care, for they are highly unlikely to consume an adequate amount of food voluntarily over the next 7 days. On the other hand, a moderate SRI, as is common during the period following major uncomplicated surgery without oral intake, may be tolerated for 5–7 days as long as the patient is initially well nourished. Patients awaiting elective major surgery benefit from preoperative nutritional repletion for 5–10 days but only in the presence of significant PEM. When adequate preoperative nutrition or SNS is impractical, early postoperative SNS is usually indicated. Furthermore, patients with a combination of a moderate SRI and moderate PEM are likely to benefit from early postoperative SNS. RISKS AND BENEFITS OF SPECIALIZED NUTRITIONAL SUPPORT
The risks of enteral SNS are determined primarily by the patient’s state of alertness and swallowing competence, the anatomy and function of the gastrointestinal tract, and the experience of the supervising clinical team. The safest and least costly approach is to avoid SNS by close attention to oral food intake; personal encouragement; dietary modifications; hand-feeding, when possible; and, often, the addition of an oral liquid supplement. For this reason, all patients at nutritional risk should be assessed and followed by a nutritionist. There is increasing interest in the use, under selected circumstances and when not contraindicated, of pharmacologic doses of anabolic steroids to stimulate appetite and promote muscle anabolism.
Nasogastric tube insertion is a bedside procedure, but many critically ill patients have impaired gastric emptying and a high risk of aspiration pneumonia. This risk can be reduced by placing the tip of the feeding tube in the jejunum beyond the ligament of Treitz, a procedure that usually requires fluoroscopic or endoscopic guidance. When a laparotomy is planned for a patient who has other surgical conditions likely to necessitate prolonged SNS, it is advantageous to place a jejunal feeding tube at the time of surgery.
A major disadvantage of enteral SNS is that the amounts of protein and calories provided to critically ill patients commonly fail to reach target goals within the first 7–14 days after SNS is initiated. This problem is compounded by the lack of enteral products that allow the provision of the recommended protein target of 1.5–2.0 g/kg without simultaneously inducing potentially harmful caloric overfeeding.
Enteral SNS is often required in patients with anorexia, impaired swallowing, or small-intestinal disease. The bowel and its associated digestive organs derive 70% of their required nutrients directly from nutritional substrates absorbed from the intestinal lumen. Enteral feeding also supports gut function by stimulating splanchnic blood flow, neuronal activity, IgA antibody release, and secretion of gastrointestinal hormones that stimulate gut trophic activity. These factors support the gut as an immunologic barrier against enteric pathogens. For these reasons, current evidence indicates that some luminal nutrition should be provided, even when PN is required to provide most of the nutritional support. The nonessential amino acids arginine and glutamine, short-chain fatty acids, long-chain omega 3 fatty acids, and nucleotides are available in some specialty enteral formulas and appear to have an important role in maintaining immune function. The addition of supplemental PN to enteral feeding (either by mouth or as SNS by enteral tube) may hasten the transition to full enteral feeding, which is usually successful when >50% of requirements can be met enterally. As long as protein and other essential nutrient requirements are met, substantial nutritional benefit can be achieved by providing ∼50% of energy needs for periods of up to 10 days. As a rule of thumb, dietary protein provision should be increased by ∼25–50% when energy intake is reduced by this amount, since negative energy balance reduces the efficiency of dietary protein retention. For longer periods and in patients who have a normal or increased body fat content, it may be preferable to provide only 75–80% of energy needs (together with increased protein), as the mild energy deficit improves gastrointestinal tolerance, makes glycemic control far easier, and avoids excess fluid administration.
The main risks of PN are related to the placement of a central venous catheter, with its complications of thrombosis and infection, and the relatively large intravenous volumes infused. Less often appreciated are the risks associated with the ease of inadvertently infusing excessive carbohydrate and lipid directly into the bloodstream. These risks include hyperglycemia, inadequate lipid clearance from the circulation, hepatic steatosis and inflammation, and even respiratory failure in patients with borderline pulmonary function. On the other hand, renal dysfunction does not reduce a patient’s requirement for protein or amino acids. In cases in which renal function is a limiting factor, appropriate renal replacement therapy must be provided along with SNS.
In the past, bowel rest through PN was the cornerstone of treatment for many severe gastrointestinal disorders. However, the value of providing even minimal amounts of enteral nutrition (EN) is now widely accepted. Protocols to facilitate more widespread use of EN include initiation within 24 h of ICU admission; aggressive use of the head-upright position; use of postpyloric and nasojejunal feeding tubes; use of prokinetic agents; more rapid increases in feeding rates; tolerance of higher gastric residuals; and adherence to nurse-directed algorithms for feeding progression. Parenteral SNS alone is generally necessary only for severe gut dysfunction due to prolonged ileus, intestinal obstruction, or severe hemorrhagic pancreatitis.
In critically ill patients, parenteral SNS can be commenced within the first 24 h of care, with the anticipation of a better clinical outcome and a lower mortality risk than those following delayed or inadequate enteral SNS; however, this point remains controversial. Some evidence suggests that early SNS is associated with a reduced risk of death but also with an increased risk of serious infection. More recent data, obtained in studies of moderately critically ill patients, suggest that early hypocaloric parenteral SNS lessens morbidity and mitigates muscle atrophy without an increased risk of infection, but also without a detectable reduction in mortality risk. Unfortunately, the current clinical-trial evidence fails to address several important unknowns. It is important to note that the level of protein substrate provided in published clinical trials generally falls well below the current recommendation, even in trials of supplemental parenteral SNS. Much of the increase in morbidity associated with parenteral and enteral SNS can be ascribed to hyperglycemia, which can be prevented by appropriately intensive insulin therapy. The level of glycemia necessary to prevent complications, whether <110 mg/dL or <150 mg/dL, remains unclear. Adequately fed surgical patients may benefit from the lower glucose range, but studies of intensive insulin therapy alone, without full feeding, suggest improved morbidity and mortality outcomes with looser control of glucose at <180 mg/dL.
In the early years of its use, PN was relatively expensive, but its components now are often less costly than specialty enteral formulas. Percutaneous placement of a central venous catheter into the subclavian vein or (less desirably) the internal jugular vein with advancement into the superior vena cava can be accomplished at the bedside by trained personnel using sterile techniques. Peripherally inserted central catheters (PICCs) can also be used, although they are usually more appropriate for non-ICU patients. Subclavian or internal jugular catheters carry the risks of pneumothorax or serious vascular damage but are generally well tolerated and, rather than requiring reinsertion, can be exchanged over a wire when catheter infection is suspected.
Although most SNS is delivered in hospitals, some patients require it on a long-term basis. At-home SNS requires a safe home environment, a stable clinical condition, and the patient’s ability and willingness to learn appropriate self-care techniques. Other important considerations in determining the appropriateness of at-home parenteral or enteral SNS are that the patient’s prognosis indicates survival for longer than several months and that the therapy enhances the patient’s quality of life. DISEASE-SPECIFIC NUTRITIONAL SUPPORT
The purpose of SNS is to correct and prevent malnutrition. Certain conditions require special modification of the SNS regimen. Protein intake may need to be limited in many stable patients with renal insufficiency or borderline liver function. In renal disease, except for brief periods, protein intakes should approach the required level for normal adults of at least 0.8 g/kg and should aim for 1.2 g/kg as long as severe azotemia does not occur. Patients with severe renal failure who require SNS need concurrent renal replacement therapy. In hepatic failure, protein intakes of 1.2–1.4 g/kg (up to 1.5 g/kg) should be provided as long as encephalopathy due to protein intolerance does not occur. In the presence of protein intolerance, formulas containing 33–50% branched-chain amino acids are available and can be provided at the 1.2- to 1.4-g/kg level. Cardiac patients and many other severely stressed patients often benefit from fluid and sodium restriction to 1000 mL of PN formula and 5–20 meq of sodium per day. In patients with severe chronic PEM characterized by severe weight loss, it is important to initiate PN gradually because of the profound antinatriuresis, antidiuresis, and intracellular accumulation of potassium, magnesium, and phosphorus that develop as a consequence of the resulting high insulin levels. This modification of parenteral SNS is usually accomplished by limiting daily fluid intake initially to ∼1000 mL; limiting carbohydrate intake to 10–20% dextrose; limiting sodium intake; and providing ample potassium, magnesium, and phosphorus, with careful daily assessment of fluid and electrolyte status. Protein need not be restricted.