William G. Simpson and Steven D. Heys
The importance of nutrition in all fields of clinical practice has become well recognised and none more so than in the management of patients who are undergoing surgery and/or are faced with critical illness. Despite this increasing understanding of the importance of nutrition in health, up to 40% of hospitalised patients can be classified as being malnourished and many of these patients are not recognised clinically as having this problem. In patients undergoing gastrointestinal surgery, for example, the prevalence of ‘mild’ and ‘moderate’ malnutrition has been estimated to be approximately 50% and 30%, respectively.
The clinical significance of this is vitally important because when patients are malnourished, disturbances in function at the organ and cellular level can manifest as the following:
The key point, therefore, is that as a result of these malnutrition-induced changes, patients have an increased risk of postoperative morbidity and mortality. To further complicate the situation with respect to nutrition, patients undergoing surgery will also be fasted for varying periods of time (preoperatively and/or postoperatively). Moreover, if patients then experience postoperative complications (e.g. sepsis), these effects may be further potentiated and the disturbances of cellular and organ function occurring in malnutrition then made even more complex.
In this chapter the following areas, which are important for surgical practice, will be outlined:
Metabolic response to feeding, trauma and sepsis
In order to maintain the health of cells, tissues and organs, the metabolism must adapt to changes in nutritional intake, trauma and sepsis.While a detailed knowledge of complex biochemical pathways is not necessary, it is important to understand the principles of these metabolic and biochemical changes, and the metabolic response when a patient experiences trauma, undergoes surgery or develops sepsis. This forms the basis for understanding nutrition and nutritional support in critically ill patients.
A major advance in understanding occurred more than 80 years ago when Sir David Cuthbertson described the loss of nitrogen from skeletal muscle that occurred following trauma.1Cuthbertson concluded that the response to injury could be considered as occurring in two phases (Fig. 17.1):
FIGURE 17.1 Diagrammatic representation of the ebb and flow phases in the metabolic response to injury. Reproduced from Broom J. Sepsis and trauma. In: Garrow JS, James WPT (eds) Human nutrition and dietetics, 9th edn. Edinburgh: Churchill Livingstone, 1993; pp. 456–64. With permission from Elsevier.
These changes result in the following:
If the changes of the ‘ebb phase’ are not replaced by the ‘flow phase’, then despite any advances in surgery, anaesthesia and intensive care support, death of the patient is the inevitable outcome.
The central nervous system and the neurohypophyseal axis play key roles in regulating these metabolic changes following trauma, utilising a range of hormones and cytokines. Afferent nerve impulses also stimulate the hypothalamus to secrete hypothalamic releasing factors that, in turn, stimulate the pituitary gland to release prolactin, arginine vasopressin (antidiuretic hormone, ADH), growth hormone and adrenocorticotrophic hormone (ACTH). The changes in hormone levels in plasma following trauma are outlined in Box 17.1, with the stress hormones (adrenaline, cortisol, glucagon) playing pivotal roles.
Box 17.1 Changes in hormone levels in plasma following trauma
Rapid increases in concentrations of adrenaline and noradrenaline within a few minutes of injury due to increased activity of sympathetic nervous system. Levels return to normal within 24 hours
Rises within a few hours; maximal levels 12–48 hours post-trauma
Initially plasma levels are low following trauma, but rise to above normal levels and reach a maximum several days after the injury
Rapid increase in cortisol (due to stimulation by ACTH), returning to normal 24–48 hours later; may remain elevated for up to several days. Has ‘permissive’ effects with other hormones such as catecholamines
Levels increased following trauma; usually return to normal levels within 24 hours
Following trauma, the biochemical features of ‘sick euthyroid syndrome’ may be present: thyroid-stimulating hormone (TSH) levels normal or low, levels of free thyroxine (T4) and tri-iodothyronine (T3) normal, whereas the total levels are altered because of changes in binding protein concentration. In addition, reverse T3 is generally high. These effects may be prolonged for some weeks
Other disturbances of thyroid function may, however, be present, including ‘transient hyperthyrotropinaemia of illness’ – a transiently raised TSH, not to be confused with hypothyroidism
Aldosterone levels increased after trauma, returning to normal within 12 hours. Its secretion is stimulated by renin, which in turn is produced in response to reduced renal perfusion
Plasma levels fall after trauma and may remain low for up to 7 days
Plasma levels rise following trauma and may remain elevated for several days
Secretion increased following trauma but function in trauma is unknown
Increased secretion of interleukin (IL)-2, IL-6, tumour necrosis factor, etc.; inter-relationship between these changes leads to differential responses seen in trauma and sepsis
Amino acids are required for:
In the well-fed state, proteins are synthesised at a rate exceeding breakdown, whereas in the fasting state breakdown predominates. Following prolonged fasting for 1–2 weeks, breakdown still predominates but at a lower rate as the metabolism adapts to starvation. In contrast, following trauma or sepsis, breakdown exceeds synthesis regardless of whether the patient is fed or fasted; this response is, however, impaired if the metabolism is already adapted to starvation.2 The magnitude of the nitrogen loss is proportional to the degree of operative trauma or the severity of the sepsis, and the major site of protein breakdown is skeletal muscle (contains 80% of the body's amino acid pool, with 60% being glutamine).3
Glucose is the main fuel used by many different tissues, being essential for some. In the well-fed state it is available for absorption from the gastrointestinal tract and, mostly driven by insulin, any excess is converted to glycogen (glycogenesis) in both liver and muscle, and to fatty acids (lipogenesis), the latter predominating when glycogen stores are replete. On fasting, insulin levels are lower, with an associated reduction in peripheral utilisation of glucose, and it is endogenously produced from glycogen (glycogenolysis) or other precursors (gluconeogenesis), e.g. amino acids and fatty acids. Initially, glycogenolysis predominates, but after a number of hours (dependent on demands), gluconeogenesis predominates (colloquially referred to as ‘getting your second wind’). Following trauma, there is an increase in hepatic glycogenolysis (caused by increased sympathetic activity),4 with these stores being substantially depleted within 24 hours.5 Insulin antagonists are also involved in this metabolic response (see Box 17.1), and the insulin resistance is accompanied by a rise in insulin concentration. The circulating insulin level usually reaches a maximum several days after the injury, before returning towards normal levels.
In general, the carbohydrate response is to produce hyperglycaemia both in the immediate ‘shock’ (‘ebb’ phase) and later ‘flow’ phase of the metabolic response. The origin of the increased glucose differs between these two phases – while reduced peripheral utilisation of glucose is common to both phases, the glycogenolysis of the ebb phase must be replaced by gluconeogenesis in the flow phase. In the critically ill patient the advent of hypoglycaemia is an indication of major problems – glycogenolysis has slowed with depletion of glycogen stores, but gluconeogenesis is not yet adequate.
In the healthy, resting, fed state, triacylglycerol, being energy dense, is used by the metabolism to efficiently store energy. When fasting, lipolysis of triglyceride releases free fatty acids, which can be used as respiratory fuel for most cells other than brain and red blood cells, and glycerol that can be converted to glucose by hepatic gluconeogenesis.6 Fatty acids are also metabolised in the liver to form ketone bodies, which are used as a preferential fuel source by many tissues (humans cannot use fatty acids for gluconeogenesis). Lipolysis is stimulated by glucagon during short-term fasting, by ACTH once the metabolism is adapted to starvation, or by adrenaline during exercise and stress. Following trauma, there is therefore an increase in the turnover of fatty acids and glycerol, although raised levels of lactate, for example in hypovolaemic shock, induce re-esterification leading to raised plasma triglyceride levels.
Mineral and micronutrient metabolism
Changes in fluid compartments, minerals and micronutrients (micronutrients are broadly defined as substances required in amounts of < 1 g daily) are beyond the scope of this chapter, but it is worth re-emphasising that measured serum concentrations rarely reflect body status, and this is even more pronounced in starvation and illness; for example, hyponatraemia is more often associated with an excess of water than with a deficiency of sodium; hypocalcaemia does not indicate a deficiency of calcium, but can suggest a deficiency of magnesium. It is therefore essential to consider the effect of illness before trying to interpret laboratory results.7
The metabolic response to sepsis is also characterised by alterations in protein, carbohydrate and fat metabolism, but the following are key differences:8
A significant abnormality in the patient with sepsis is the disruption of the microstructure of the hepatocyte mitochondria, particularly of the inner membrane. There is a block in energy transduction pathways, with consequent reduction in the aerobic metabolism of both glucose and fatty acids. The body therefore depends on the anaerobic metabolism of glucose, which also results in lactate production. It is essential, therefore, that there is an adequate supply of glucose from gluconeogenic pathways. If this is impaired or inadequate, then hypoglycaemia (and death) may ensue. The development of hypoglycaemia during sepsis is an indicator of an extremely poor prognosis and is usually associated with inevitable mortality.
Proteins And Amino Acids
Protein is required for the maintenance of normal health and cellular function. Proteins have many functions, including being essential components of cellular structure. They are required for the synthesis of a variety of secretory proteins produced by many organs. The average daily intake of protein is approximately 80 g in the UK, with a recommended daily intake of 0.8 g/kg body weight and with nitrogen comprising approximately 16% of its weight. However, more than 50% of the world's population exist on less!
Conventionally, amino acids have been classified as either ‘essential’ or ‘non-essential’. The ‘essential’ amino acids cannot be synthesised endogenously and are required in the diet. Paradoxically, the so-called ‘non-essential’ amino acids are actually so metabolically important that humans have retained the ability to synthesise them. Both groups of amino acids are necessary for normal tissue growth and metabolism.
Dietary intake and endogenous synthesis of amino acids in the body maintain the relevant pool of amino acids, replacing those that have been lost by excretion in the urine, losses from the skin and gastrointestinal tract, utilisation as precursors for non-protein synthetic pathways, irreversible modification and irreducible oxidation.
Under certain circumstances (e.g. sepsis, trauma, growth) endogenous synthesis of some amino acids normally considered to be ‘non-essential’ is inadequate; therefore, these amino acids are described as ‘conditionally essential’: L-alanine, L-glutamate and L-aspartate, which are produced by a simple transamination reaction. These are the three most important amino acids in times of starvation:
Energy transduction is accomplished by the breakdown of carbohydrate, fat and proteins. The energy available from various common nutrients is:
The principal carbohydrates in the diet are polysaccharides (starch and dietary fibre), dextrins and free sugars (monosaccharides), disaccharides, oligosaccharides and sugar alcohols. Dietary fat includes triacylglycerol, containing long-chain fatty acids (C16–C18 triacylglycerols) and medium-chain fatty acids (C6–C12 triacylglycerols) and cholesterol.
If the energy intake of an individual is greater than energy expenditure, extra carbohydrate intake, on reaching the liver via the portal vein, will be channelled into synthesis of glycogen or fat. Glycogenesis dominates until hepatic glycogen stores are replete; thereafter, fat synthesis dominates. Additional fat intake will be stored in adipose tissue as triacylglycerol. In contrast, if there is a negative energy balance, then glycogenolysis dominates until glycogen stores are depleted, then fat and protein will be broken down to provide energy.
Total daily energy expenditure comprises the following:
Normally, approximately 25–30 kcal/kg (105–125 kJ/g) are required daily; the magnitudes of changes in requirements in some common conditions are given in Box 17.2.
Box 17.2 Additional energy requirements in disease states
Trauma: 0.3 × RME
Elective surgery: 0.1 × RME
Sepsis: up to 0.5 × RME
Severe sepsis: up to 0.6 × RME
Massive burns: 1 × RME
Minerals And Micronutrients
A number of specific organic compounds (vitamins) and inorganic elements are essential for tissue growth and repair, and for maintenance of body function, playing key roles in metabolism, including the processing of macronutrients (protein, carbohydrate and fat). For many micronutrients, specific deficiency diseases have been described. Details of individual substances are beyond the scope of this chapter and can be found elsewhere,9 but some examples are given in Box 17.3.
Box 17.3 Functions of some micronutrients important in surgical practice
Stabilises epithelial cell membranes; necessary for fibroblast differentiation and collagen secretion
Role in calcium and phosphate regulation
Immunostimulant and free radical scavenger
Required for liver synthesis of clotting factors
Important in synthesis of proteins and nucleic acids
Important in hydroxylation (e.g. collagen synthesis) and energy transduction
Necessary for carbohydrate metabolism and ATP synthesis
Antioxidant; protection against peroxidation processes occurring in tissue damage and repair
Cofactor in numerous enzymes; necessary for wound healing
In general, micronutrients are classified into:
The exact requirement for micronutrients during trauma and sepsis is unclear and may alter depending on the type of metabolic support provided.
It should be remembered that micronutrients, if given at high doses, can have toxic effects on tissues. In particular, toxicity can be a problem with excesses of vitamin A, iron, selenium, zinc and copper; vitamin D toxicity is no longer thought to be a significant problem. Care must be taken when these micronutrients are provided for a prolonged period to ensure that toxicity does not occur. It is also unusual to find isolated deficiencies, so identification of one micronutrient deficiency should stimulate consideration of other deficiencies.
Identification of patients who are malnourished
It is important to assess the nutritional status in all patients undergoing surgery and identify those who are malnourished, or who are at risk of becoming so. Measurements used previously in clinical practice include:
Height and weight
Height and weight are two commonly used indices of nutritional status.10 Body weight on its own takes no account of frame size but body mass index (BMI; weight divided by the square of the height) is a good anthropometric indicator of total body fat in adults.
Loss of body weight has been used as an indicator of nutritional status. This is determined by subtracting current weight from recall weight when the patient was ‘well’, or from the ‘ideal’ weight, obtained from published tables. The loss of more than 10% of body weight, or more than 4.5 kg of recall weight, is associated with a significant increase in postoperative mortality. The shorter the period of weight loss, the more significant this is in predicting increased postoperative complications. Malnutrition can be defined as a BMI of less than the 10th percentile with a weight loss of 5% or more.11
Various techniques for assessing the body's different compartments (e.g. fat, fat-free mass, total body nitrogen and total body mineral contents) have become available but many require specialised equipment and may not be readily applicable to clinical practice. Relatively simple techniques, such as skinfold thickness and bioelectrical impedance, can be used clinically, although even these tend to be used more for research or clinical audits of nutrition and nutritional support.
Subcutaneous fat thickness
Skinfold thickness has been used as an index of total body fat (50% of total body fat is subcutaneous, depending on age, sex and fat pad). Triceps skinfold thickness is most commonly measured but assessment of skinfolds at multiple sites is better and correlates with total body fat. Regression equations for the estimation of total body fat from these measurements are available.12 However, skinfold thickness measurements are susceptible to intra-observer and inter-observer variability, which limits clinical use.
This entails the passage of an alternating electrical current between electrodes attached to the hand and foot. The current passes through the water and electrolyte compartment of lean tissues and the drop in voltage between the electrodes is measured.13 This change in voltage gives an estimation of total body resistance, which depends principally on total body water and electrolyte content (i.e. lean body mass). This estimate can give an accurate measure of body composition in stable subjects; it becomes less reliable in patients with oedema and electrolyte shifts, and so the value of bioelectrical impedance in critically ill patients remains unclear.14
Albumin is the major protein in serum and the relationship between serum proteins and malnutrition was recognised over 150 years ago.14 Low serum albumin levels are associated with increased risk of complications in patients undergoing surgery.15 In experimental starvation, however, serum albumin levels may not fall for several weeks16 because, although synthesis decreases, only 30% of the total exchangeable albumin is in the intravascular space, with the remainder being extravascular. In addition, albumin has a relatively long half-life of approximately 21 days. The flux of albumin between the intravascular and extravascular compartments is about 10 times the rate of albumin synthesis, although this varies greatly depending on capillary permeability.17
Importantly, serum albumin acts as a negative marker of the acute-phase response, and so is lowered in malignancy, trauma and sepsis, even in the presence of an adequate intake. Serum albumin should therefore not be used as an assessment of nutritional state, although low levels point to the increased nutritional risk associated with underlying disease, and indeed the implied reduction of gut absorption may indicate that the parenteral route may be preferred for provision of nutrition.
Alternatives to using albumin as a marker of nutritional status by measuring other serum protein concentrations, including transferrin (half-life 7 days), retinol-binding protein (half-life 1–2 hours) and pre-albumin (half-life 2 days), have been suggested. The serum levels of these proteins are, however, also altered in stress, sepsis and cancer, and so, as for albumin, they are not useful for assessing nutritional status in routine clinical practice.
Most of the nitrogen lost from the body is excreted in urine, mainly as urea (approximately 80% of total urinary nitrogen). Urea alone may be measured as an approximate indicator of losses, or total urinary nitrogen may be measured, although this latter technique is not widely available. In addition, there are also losses of nitrogen from the skin and in stool of approximately 2–4 g per day. One equation used for balance studies is:
Although nitrogen balance has not been shown to be a prognostic indicator, it is a useful way of assessing a patient's nutritional requirements and the response to provision of nutritional support.
Tests Of Function
In malnutrition there is a reduction in total circulating lymphocyte count and impairment in immune functions, e.g. decreased skin reactivity to mumps, Candida and tuberculin (assuming prior exposure), and reduced lymphocyte responsiveness to mitogens in vitro.18,19
A correlation between depressed immune function and postoperative morbidity and mortality has been demonstrated, and depression of total circulating lymphocyte count is associated with a poorer prognosis in surgical patients.20 However, these alterations in immune function are non-specific and affected by trauma, surgery, anaesthetic and sedative drugs, pain and psychological stress,21 and are not generally applicable to clinical practice.
Skeletal Muscle: Various aspects of skeletal muscle structure and function are deranged in malnutrition. In patients undergoing surgery, handgrip strength (cheap and easy to perform) may predict patients who develop postoperative complications (sensitivity > 90%). However, grip strength is influenced by factors such as patients' motivation and cooperation. Furthermore, such tests may be difficult to apply to critically ill patients. Alternatively, stimulation of the ulnar nerve at the wrist with a variable electrical stimulus results in contraction of the adductor pollicis muscle, the force of which reflects nutritional intake.22
Respiratory Muscle: The function of the respiratory muscles is impaired by malnutrition and can be detected by deterioration in respiratory function tests, in particular vital capacity.23Measurements of inspiratory muscle strength have the advantage that they can be performed in patients who are intubated.
Nutrition risk index
A nutrition risk index is an index of nutritional status based on combinations of variables. Although several indices exist, one that is commonly used depends on serum albumin, current weight and the patient's usual weight:
The score obtained is used to categorise nutritional risk: < 83.5, ‘severe’ risk; 83.5–97.5, ‘mild’ risk; 97.5–100, ‘borderline’ risk. Note that this is a prognostic index and bears little relationship to the patient's nutritional status.
How Should Nutritional Status Be Assessed In Clinical Practice?
Although the various techniques outlined above can help to predict the risks of complications, there is at present no reliable technique for assessing nutritional status. There is, however, increasing support for using the following techniques, which are applicable to clinical practice.
The Malnutrition Universal Screening Tool (MUST)
This simple, yet effective, tool was developed by the British Association for Parenteral and Enteral Nutrition (BAPEN). Details are available from the website, which we recommend you read carefully (www.BAPEN.org.uk). This tool has been endorsed by external organisations and its routine use is recommended for all hospital admissions in the UK. As a minimum standard in clinical practice, the MUST tool should be used to assess nutritional risk. Essentially, MUST consists of a series of five steps:
Significance Of The Resultant Score And Clinical Management:
The MUST tool should be used routinely to assess nutritional risk for all hospital admissions (www.BAPEN.org.uk).
Subjective global assessment
Useful indicators for bedside assessment of nutritional status applicable to clinical practice have been identified,24 and include estimation of protein and energy balance, assessment of body composition and evaluation of physiological function.
Assessment Of Protein And Energy Balance: Protein and energy balance can be assessed either by a dietician or by a clinician, who determines the frequency and size of meals eaten. This information is compared with the patient's rate of loss of body weight and BMI.
Assessment Of Body Composition: Loss of body fat can be determined by observing the physical appearance of the patient (loss of body contours) and feeling the patient's skinfolds between finger and thumb. In particular, if the dermis can be felt on pinching the biceps and triceps skinfolds, then considerable weight loss has occurred.
The stores of protein in the body can be assessed from various muscle groups, including the temporalis, deltoid, suprascapular, infrascapular, biceps and triceps, and the interossei of the hands. When tendons of the muscles are prominent and bony protruberances of the scapula are obvious, greater than 30% of the total body protein stores have been lost.
Assessment Of Physiological Function: Assessments of function are made by observing the patient's activities. Grip strength is determined by asking patients to squeeze the clinician's index and middle fingers for at least 10 seconds, and respiratory function by asking them to blow hard on a strip of paper held 10 cm from the patient's lips. The measurement of metabolic expenditure requires specialised equipment, but additional metabolic stresses on the patient can be determined from clinical examination. Extra metabolic stresses will occur if trauma or surgery has taken place or there is evidence of significant sepsis (elevated temperature and/or white blood cell counts, tachycardia, tachypnoea, positive blood cultures) or active inflammatory bowel disease. In addition, patients should be asked about their ability to heal wounds, changes in exercise tolerance and their ‘tiredness’.
Re-Feeding Syndrome: Once a patient's need for nutritional support has been identified, it is important to consider whether the patient is at risk of re-feeding syndrome. This is described in detail elsewhere, but in essence it is the inability of a patient's metabolism to handle macronutrients. After approximately 10 days without nutritional intake, the metabolism adapts to the state of starvation. Re-feeding with full ‘normal’ required amounts of macronutrients will induce a sudden reversal of this adaptation, with an anabolic drive that may result in catastrophic depletion of available potassium, phosphate and magnesium. Before re-feeding, the serum biochemistry may appear ‘normal’, and so the possibility of re-feeding must be anticipated on history alone. The other essential nutrient liable to become depleted in this situation is thiamine, a cofactor of pyruvate kinase, which is required for glucose to undergo oxidative phosphorylation, and without which glucose is metabolised to lactic acid. Thiamine must therefore be replenished before feeding is commenced in the starved patient to prevent development of Wernicke–Korsakoff syndrome. The potential for this is considerably higher in patients with a history of chronic excessive ethanol intake, and so even greater caution is required.
Thiamine deficiency must always be considered in patients who have had no nutritional intake for more than 1 week, who have a history of excessive alcohol intake, or in the presence of an unexplained metabolic acidosis.
Nutritional support in surgical practice
Route Of Nutritional Support
The preferred route of administration of nutritional support is through the gastrointestinal tract (enteral), with intravenous (parenteral) nutrient delivery reserved for patients with intestinal failure.
Enteral Nutritional Support
If there is an intact and functioning gastrointestinal tract, enteral feeding should be used if oral intake is insufficient. Enteral feeding is contraindicated to various degrees in patients with intestinal obstruction, paralytic ileus, vomiting and diarrhoea, high-output intestinal fistulas or in the presence of major intra-abdominal sepsis.
The importance of enteral nutrition
Studies in animals have shown that in the absence of nutrients into the intestinal lumen, changes occur in the intestinal mucosa. There is loss of height of villi, reduction in cellular proliferation and the mucosa becomes atrophic.25,26 Activities of enzymes found in association with the mucosa are reduced and permeability of the mucosa to macromolecules increased.27Stimulation of the intestinal tract by nutrients is important for release of many gut- related hormones, including those responsible for gut motility and stimulation of secretions necessary for normal maintenance of the mucosa. The gut acts as a barrier to bacteria, both physically and by release of chemical and immunological substances. There is evidence to suggest that atrophy of the intestinal mucosa is associated with loss of intercellular adhesion and opening of intercellular channels. This predisposes to increased translocation of bacteria and endotoxin from the gut lumen into portal venous and lymphatic systems.28 Loss of gut integrity may account for a substantial proportion of septicaemic events in severely ill patients. However, the extent to which it contributes to sepsis in patients is not fully understood.
Routes of access for enteral nutritional support
Nasoenteric Tubes: Nasogastric feeding via fine-bore tubes (polyvinyl chloride or polyurethane) may be used in patients who require nutritional support for a short period of time. There has been considerable debate as to whether positioning the feeding tube beyond the pylorus into the duodenum will result in reduction in the risks of regurgitation of gastric contents and pulmonary aspiration (occurs in up to 30% of patients fed this way). This is most likely in patients with impaired gastric motility. In the latter patients, the fine-bore tube can be manipulated through the pylorus into the duodenum, reducing the risk of gastric aspiration. Other complications associated with the use of nasoenteric tubes include:
More recently, double-lumen tubes have been used – one lumen resides in the stomach and is used to aspirate gastric contents, while the distal lumen is placed in the jejunum for feeding, thus reducing risks of aspiration. This can be successful even in patients with relatively high gastric aspirates, previously thought to be a contraindication for feeding via the enteral route.
Gastrostomy Tubes: A gastrostomy tube can be placed into the stomach at laparotomy, although percutaneous endoscopic or percutaneous fluoroscopic techniques are preferred. Details of how these are performed can be found in standard texts.
The establishment and use of a gastrostomy has certain disadvantages and there is a recognised morbidity:
The overall mortality rate for a gastrostomy is 1–2%, with major and minor complications occurring in up to 15% of patients. Mechanical complications associated with the tube include blockage, fracture and displacement. Furthermore, ‘dumping’ and diarrhoea are more common when the tip of the tube lies in the duodenum or jejunum.29
Jejunostomy Tubes: A feeding jejunostomy is usually carried out at the time of laparotomy if it is envisaged that a patient will need nutritional support for a longer period. Details of the operative technique are also in standard operative texts and the smaller needle-catheter tubes are to be preferred. Advantages of a feeding jejunostomy compared with a gastrostomy are:
Nutrient solutions available for enteral nutrition
A range of nutrient solutions are available for use in enteral nutritional support and examples can be found in specialised texts. However, there are four main categories of enteral diet.
Polymeric Diets: Polymeric diets are ‘nutritionally complete’ diets and provided to patients with inadequate oral intake, but whose intestinal function is good. They contain whole protein as the source of nitrogen, and energy is provided as complex carbohydrates and fat. They also contain vitamins, trace elements and electrolytes in standard amounts.
Elemental Diets: Elemental diets are required if the patient is unable to produce an adequate amount of digestive enzymes or has a reduced area for absorption (e.g. severe pancreatic insufficiency or short-bowel syndrome). Elemental diets contain nitrogen as oligopeptides (free amino acids are not as easily absorbed as dipeptide and tripeptide mixtures). The energy source is provided as glucose polymers and medium-chain triacylglycerols. Each oligopeptide molecule contributes as much to the osmolarity of the solution as one molecule of intact protein, and it can be difficult to provide complete requirements without producing side-effects associated with an osmotic load, e.g. ‘dumping’ and diarrhoea.
Special Formulations: Special formulations have been developed for patients with particular diseases. Examples of such diets include: (i) those with increased concentrations of branched-chain amino acids and low in aromatic amino acids for patients with hepatic encephalopathy; (ii) those with a higher fat but lower glucose energy content for patients who are artificially ventilated; and (iii) diets containing key nutrients that modulate the immune response (see later).
Modular Diets: Modular diets are not commonly used but allow provision of a diet rich in a particular nutrient for specific patients. For example, the diet may be enriched in protein if the patient is protein deficient or in sodium if sodium deficient. These modular diets can be used to supplement other enteral regimens or oral intake.
Enteral nutrition delivery and complications
Previously, when starting an enteral nutrition feeding regimen, patients received either a reduced rate of infusion or a lower strength formula for the first 2 or 3 days to reduce gastrointestinal complications. Recent studies have demonstrated this is not required and nutritional support can commence using full-strength feeds at the desired rate in those not at risk of developing ‘re-feeding syndrome’. Cyclical feeding (e.g. 16 hours feeding with a post-absorptive period of 8 hours) is optimal and more closely mimics the natural feeding cycle than other types of feeding regimens.30
Enteral nutrition should be administered through a volumetric pump. If not available, then it is possible to use a gravity drip flow but care should be taken to reduce the risk of a large bolus being administered. In patients whose conscious level is impaired or confined to bed, the head of the bed should be elevated by 25° to reduce risks of pulmonary aspiration. Some clinicians prefer patients to be sitting upright when receiving enteral nutrition. The stomach contents should be aspirated every 4 hours during feeding and if a residual volume of more than 100 mL is found, enteral nutrition is temporarily discontinued.
The aspirate is checked again after 2 hours, and when satisfactory volumes are aspirated (< 100 mL) feeding is re-instituted. If more than 400 mL per 24 hours is aspirated, then feeding is discontinued. Gastric emptying may be improved by the administration of cisapride or erythromycin, which may allow feeding to be continued.
Metabolic disturbances are less likely with enteral feeding. The other complications of enteral nutrition are those associated with the route of access to the gastrointestinal tract (Box 17.4).
Box 17.4 Complications of enteral nutrition
Diarrhoea, nausea, vomiting, abdominal discomfort and bloating, regurgitation and aspiration of feed/stomach contents
Dislodgement of the feeding tube, blockage of the tube, leakage of stomach/small intestine contents onto the skin with the use of jejunostomies or gastrostomies
Excess or deficiency of glucose, electrolytes, minerals or trace elements. Some of these will be noted through routine testing protocols, e.g hyperkalaemia, but others such as hypophosphataemia may be missed if not specifically anticipated
Local effects (e.g. diarrhoea, vomiting) or systemic effects (e.g. pyrexia, malaise)
Parenteral Nutritional Support
Patients who require nutritional support but with enteral feeding contraindicated will require parenteral nutrition. These include:
Detailed guidance for parenteral nutrition (PN) in patients has been published by the American Society for Parenteral and Enteral Nutrition (ASPEN).31 Current ASPEN guidelines are available on their website (www.nutritioncare.org), as are those of the European Society for Clinical Nutrition and Metabolism (www.espen.org/).
Parenteral routes of access
Central Venous Access: Central venous access is obtained by positioning a catheter into the superior vena cava through subclavian or internal jugular veins. The catheter either emerges through the skin (usually after being tunnelled in the subcutaneous fat) or is connected to a port placed in the subcutaneous fat of the anterior chest wall. A variety of techniques for insertion of central venous lines are used. For example, catheters may be introduced into the internal jugular or subclavian vein directly by ‘blind’ percutaneous puncture, using small hand-held ultrasound imaging, by ‘cut-down’ techniques utilising the cephalic vein to access the subclavian vein, or under fluoroscopic control. Details of these techniques, their advantages and disadvantages can be found elsewhere.32–34 However, it is important that whoever inserts a central venous line is expert, well practised and carries out the procedure under full aseptic techniques.
Technical Aspects Of Feeding Lines: Central lines are manufactured from polyurethane or silicone. Both of these materials are tolerated well with low thrombogenic potential. However, polyurethane does have advantages:
Catheter manufacturers have attempted to reduce risks of bacterial colonisation of the line by bonding antiseptics (e.g. chlorhexidene) and antibiotics (e.g. silver sulphadiazine) into the catheter's fabric. Some catheters have an antimicrobial cuff, usually made of Dacron, around their external surface. This acts as a barrier to micro-organisms, which may migrate from subcutaneous tissues along the external aspect of the catheter to its tip. Although studies have suggested that risks of septicaemia are reduced by using a cuff around the catheter, this makes positioning of the catheter more difficult technically. Complications of central venous catheters are shown in Box 17.5.
Box 17.5 Complications of central venous catheter placement and incidence of occurrence
Catheter-related sepsis: variable, but reported in up to 40% of catheters
Thrombosis of central vein: variable, but reported in up to 20% of catheters
Pleural space damage: pneumothorax (5–10%), haemothorax (2%)
Major arterial damage: subclavian artery (1–2%)
Catheter problems: thrombosis (1–2%), embolism (< 1%), air embolism (< 1%)
Miscellaneous problems: brachial plexus (< 1%), thoracic duct damage (< 1%)
Catheter Care: Appropriate dressings of the catheter are essential. The dressing should be changed weekly with strict aseptic technique, and the skin exit site cleaned with chlorhexidene. A variety of dressings have been used at the skin exit, but a transparent adherent type of dressing has the advantage of allowing a visible check on the puncture site for inflammation or pus.
Infection of the catheter tip is the most serious type of infection. The patient usually is pyrexial and may have systemic signs of sepsis. This may be diagnosed by blood (at least three cultures 1 hour apart) and catheter cultures.35 Antibiotic therapy may result in recovery, but in some the feeding line has to be removed to eradicate the infection. However, less serious infection may occur in the skin at the exit site of the catheter. This is recognised by skin erythema, possibly associated with fluid exudate and pus.
Peripheral Venous Access: Peripheral venous cannulation, using a sterile technique, may be used to supply nutrients intravenously, avoiding complications associated with central venous catheters. Peripheral intravenous nutrition is likely to be used in patients who do not require nutritional support for long enough to justify risks of central vein cannulation or in whom central vein cannulation is contraindicated (e.g. central line insertion sites are traumatised, increased risks of infective complications, thrombosis of the central veins or significant clotting defects).
Problems associated with the delivery of intravenous nutrition using the peripheral route include:
The lifespan of a peripheral intravenous cannula can be prolonged by treating it as if it is a central line with regard to aseptic care, and by using a narrow-gauge cannula giving better mixing and flow characteristics of the nutrient solution. Risks of phlebitis can be reduced by frequent changes of infusion site, ultrafine-bore catheters or using a vasodilator patch over the cannulation site (e.g. transdermal glyceryl trinitrate). Furthermore, peripheral intravenous nutrition can only be used where fat emulsion is part of the single-phase administration of nutrients to avoid thrombophlebitis.
Nutrients used in parenteral feeding solutions
Various nutrient solutions (amino acids, glucose and fat) are available and a complete list is given in the British National Formulary (www.bnf.org/bnf/). There are also available a variety of pre-mixed bags containing various concentrations of amino acids and glucose, with or without fat, which are suitable for different clinical situations. These mixtures do not usually contain vitamins or trace elements, which must be given in addition to avoid development of metabolic complications. Care should be taken that patients receive sufficient electrolytes and minerals to satisfy requirements.
Nitrogen Sources: Nitrogen sources are solutions of crystalline L-amino acids containing all essential and a balanced mixture of the non-essential amino acids required. Amino acids that are relatively insoluble (e.g. L-glutamine, L-arginine, L-taurine, L-tyrosine, L-methionine) may be absent or present in inadequate amounts.
Attention has focused on the provision of L-glutamine because of its key roles in metabolism. Despite being one of the most abundant amino acids, its use in PN fluids has been limited by instability. It can, however, be supplied as N-acetylglutamine (hydrolysed in the renal tubule to free L-glutamine) or as L-glutamine dipeptides such as alanylglutamine (broken down to release free L-glutamine). Recent evidence, however, questions the need for enrichment of PN with glutamine.36
Energy Sources: Energy is supplied as a balanced combination of dextrose and fat. Glucose is the primary carbohydrate source and the main form of energy supply to the majority of tissues. During critical illness the body's preferred calorie source is fat (fasted or fed states).37,38 There are controversies as to the utilisation of fat in sepsis because of defects in energy substrate metabolism at the oxidative level.
Glucose utilisation may be impaired in certain patients and glucose is then metabolised through other pathways. This results in increased production and oxidation of fatty acids, resulting in increased carbon dioxide (excreted through the lungs). In addition, if glucose is the only energy source, patients may develop essential fatty acid (linolenic, linoleic) deficiency.
Fat (e.g. soyabean oil emulsions) provides a more concentrated energy source. Usually, approximately 30–50% of the total calories are given as fat, with non-protein calorie to nitrogen ratio varying from 150:1 to 200:1 (lower in hypercatabolic conditions). The provision of exogenous lipids has also been associated with problems. Intravenous fat emulsions can impair lung function, inhibit the reticuloendothelial system and modulate neutrophil function; recent interest has also focused on the use of fish oils as a source of fat rich in omega-3 polyunsaturated fatty acids, as this appears to be associated with reduced incidence of hepatic dysfunction.39,40
Other Nutrients: Commercially available preparations of trace elements (e.g. Additrace®) and vitamins, water soluble (e.g. Solivito®) and fat soluble (e.g. Vitlipid®), supply daily requirements. Larger amounts, particularly of the water-soluble vitamins, may be required initially if recent nutritional intake has been inadequate. Additionally, total fluid volume and amounts of electrolytes can be modified daily to meet particular requirements.
Delivery and administration of PN
In practice, commercially available solutions for parenteral infusion are mixed under sterile conditions in laminar flow facilities. The feeding regimen is made up in an inert 3- to 4-litre bag (ethyl vinyl acetate), comprising all nutrients and stored for up to 1 week, although compatibility between different constituents must be ensured. No additions of drugs should be made as this could make the emulsion unstable, affect the bioavailability of the drug or compromise sterility. Advantages of pre-mixed bags include:
Pre-prepared bags are available where the fat emulsion is stored separately from the aqueous solution and is mixed by bag rupture immediately prior to administration, conferring the advantage of a longer shelf life.
Complications of parenteral nutritional support
Instant availability of nutrients provided by the intravenous route can lead to metabolic complications if the composition or flow rate is inappropriate. Rapid infusion of high concentrations of glucose can precipitate hyperglycaemia, which may be further complicated by lactic acidosis. Electrolyte disturbances may present problems, not least because the intravenous feeding regimen is usually prescribed in advance for 24 hours. Prediction of the patient's nutrient requirements must be complemented by frequent monitoring. The provision of nutrients may lead to further electrolyte abnormalities when potassium, magnesium and phosphate enter the intracellular compartment. This is particularly noticeable in patients whose previous nutrient intake was especially poor, as highlighted previously. Others complications of PN are shown in Box 17.6.
Box 17.6 Metabolic complications of parenteral nutrition
Hyperglycaemia: excessive administration of glucose inadequate insulin, sepsis
Hypoglycaemia: rebound hypoglycaemia occurs if glucose is stopped abruptly but insulin levels remain high
Hyperlipidaemia: directly through excess administration of lipid, or indirectly through excess calories that will be converted to fat or reduced metabolism (e.g. renal failure, liver failure)
Fatty acid deficiency: essential fatty acid deficiency leads to hair loss, dry skin, impaired wound healing
Hyperammonaemia: occurs if deficiency of L-arginine, L-ornithine, L-aspartate or L-glutamate in infusion. Also occurs in liver diseases
Metabolic acidosis: caused by excessive amounts of chloride and monochloride amino acids
Hyperkalaemia: excessive potassium administration or reduced losses
Hypokalaemia: inadequate potassium administration or excessive loss
Hypocalcaemia: inadequate calcium replacement, losses in pancreatitis, hypoalbuminaemia
Hypophosphataemia: inadequate phosphorus supplementation, also tissue compartment fluxes
Elevations in aspartate aminotransferase, alkaline phosphatase and γ-glutamyltransferase may occur because of enzyme induction secondary to amino acid imbalances or deposition of fat and/or glycogen in liver
If excessive amounts of glucose are given, the increased production of CO2 may precipitate ventilatory failure in non-ventilated patients
Monitoring Patients Receiving Nutritional Support
Patients receiving nutritional support should be monitored by accurate recording of fluid balance and daily weighing. Daily intake of calories and nitrogen should be documented. Biochemical assessments include daily measurements of renal and liver function, with twice-weekly checks of phosphate, calcium, magnesium, albumin and protein levels, and haematological indices (haemoglobin, white blood cell count, haematocrit), until the patient is stabilised. Then, weekly or fortnightly measurements are necessary. Patients receiving PN require urinalysis daily initially in case glycosuria occurs, as this induces further fluid and electrolyte losses. If glycosuria occurs, it may be necessary to commence intravenous insulin on a sliding scale with hourly blood glucose monitoring. It is important to note that if the PN fluid is stopped, insulin requirements will reduce immediately; it is safest to discontinue the insulin at the same time as the PN, reviewing the sliding scale with a view to giving intravenous glucose if required.
Routes of access (enteral or parenteral) should be regularly examined to ensure that the catheter is correctly positioned and mechanically satisfactory.
When feeding is prolonged, other assessments, e.g. muscle function, nitrogen balance, measurement of trace elements and vitamins, may be performed regularly to ascertain patient progress (see nutritional assessment section above).
Nutritional Support Teams
It is clear that for optimal provision of nutritional support, a multidisciplinary nutritional support team is required. This may comprise a clinician with a special interest in nutritional support and understanding of metabolic pathways, a biochemist, a pharmacist, a dietician and a nursing specialist.
The provision of nutritional support by such a team results in the most cost-effective use of nutritional support and the least risk of infective, metabolic and feeding-line complications.41
Nutritional support in defined clinical situations
Nutritional Support In The Perioperative Period
There is debate as to which patients require preoperative and/or postoperative nutritional support. Many studies have evaluated the effects of nutritional support in the perioperative period; clinical benefit with supplemental nutrition has not been a consistent finding. This may be because the studies were small, with many different end-points (e.g. morbidity, mortality), frequently without proper randomisation or allowance for malnutrition prior to the study commencing.
A meta-analysis has examined 27 randomised controlled trials (almost 3000 patients) of nutritional support in the perioperative period.42 The results are important and provide a basis for the rational use of nutritional support in this situation. The key findings are detailed in Table 17.1. When PN was given in the preoperative period there was a reduction in complication rates (relative risk 0.52, 95% confidence interval (CI) 0.30–0.91) in malnourished patients but not when nutritional state was adequate. However, there was no difference in mortality. Analysis of patients in the postoperative period indicated no reduction in complications (relative risk 1.08, 95% CI 0.81–1.43) or mortality in patients receiving PN. Subgroup analyses indicated that nutritional support in the preoperative period be considered for:
Effect of perioperative nutritional support on morbidity and mortality in surgical patients
Complications (RR and 95% CI)
Mortality (RR and 95% CI)
CI, confidence interval; RR, relative risk; TPN, total parenteral nutrition.
Nutritional support should be given to malnourished patients for at least 7–10 days preoperatively where possible to reduce postoperative morbidity.42 Nutritional support in the postoperative period should be considered for:
As discussed, the enteral route is the preferred route except in specific circumstances where not possible (e.g. intestinal obstruction, ileus, intestinal ischaemia, etc.) or used in combination with PN if the nutritional requirements cannot be provided by the enteral route alone. A recent analysis of
The ESPEN guidance is summarised as follows.43 Patients who should receive perioperative nutritional support:
Patients who should receive preoperative enteral nutritional support:
Patients who should receive postoperative nutritional support:
studies of enteral nutrition given to patients in the perioperative period has been carried out and published as European Society of Parenteral and Enteral Nutrition (ESPEN) guidelines43(www.espen.org/Education/documents/ENSurgery.pdf).
Nutritional Support In Patients With Acute Pancreatitis (See Also Chapter 8)
Severe pancreatitis produces a major catabolic stress with rapid loss of muscle proteins. The daily nitrogen requirements of such patients are high, reaching 1.2–2.0 g protein/kg body weight (0.2–0.3 g of nitrogen/kg). Daily energy requirements also increase with disease severity to be 28–35 kcal/kg. Previously, patients with pancreatitis were fasted in order to avoid pancreatic stimulation. However, views have now changed44,45 and a recent systematic review46 has shown that patients with acute severe pancreatitis should commence enteral support early (within 5 days). This resulted in better outcomes with reduced infectious complications and hospital stay, but without effect on mortality.44–46
Patients with severe acute pancreatitis should commence early enteral nutritional support as this is associated with a better outcome.46
Nutritional Supplementation In Inflammatory Bowel Disease
A significant number of patients with Crohn's disease and ulcerative colitis become malnourished. The reasons for this include decreased nutrient intake, malabsorption by the small intestine (decreased length, bacterial overgrowth, protein-losing enteropathy) and increased calorie/nitrogen requirements in those with coexistent sepsis. There may be deficiencies of specific vitamins and trace elements.
Nutritional support, therefore, may have different roles: (i) to provide nutritional requirements and correct nutritional deficiencies the patient may have; (ii) the possibility that provision of PN with bowel rest in Crohn's disease may be therapeutically beneficial. The results of studies addressing this latter point are inconclusive,47,48 suggesting that PN itself does not have a therapeutic effect in inflammatory bowel disease. Furthermore, there is evidence showing that enteral nutrition is as effective as PN in these patients.49 This has the added benefits of maintaining gut mucosa integrity in addition to stimulating production of gut hormones necessary for function.
While systematic reviews have shown some potential for enteral nutrition in this regard, further studies are required to clarify the use of nutrition in this way.50,51
Nutritional Support In Enterocutaneous Fistulas
Nutritional support has an important role to play in management of patients with enterocutaneous fistulas as up to 50% are malnourished. The importance of adequate nutritional support was demonstrated by Chapman et al.,52 who found that if patients with fistulas received nutritional support with PN and enteral feeding (> 3000 kcal (12.6 MJ) daily), spontaneous fistula healing with a reduced mortality occurred compared with patients with fistulas receiving less than 1000 kcal (4.1 MJ) daily. The management of such patients commences with correction of fluid and electrolyte deficits and elimination of septic foci. Nutritional support is required to correct any nutritional deficits and provides maintenance requirements when the patient is stabilised. Whether PN or enteral nutrition is more effective is unknown. Other techniques for providing nutritional support have included collecting the intestinal output from the proximal end of the fistula and re-infusing it into the distal part of the small intestine or by giving enteral nutrition via the fistula. If the fistula output is low, enteral nutritional support should be considered because of the benefits.53
Enteral nutrition has theoretical benefits due to its effects on gut mucosa and case series have suggested that healing rates with enteral nutrition are comparable to those of parenteral nutrition.53
Nutritional Support In Patients With Burns
Major burns induce severe hypermetabolic and hypercatabolic states. There is increased skeletal muscle breakdown, nitrogen losses of 15 g daily or more, and up to a doubling of metabolic rate. In patients with burns of greater than 20% of their body surface area, nutritional support is required, orally or by nasoenteric feeding. There may be clinical benefits by introducing feeding early, and with glutamine supplementation.54,55 This appears to be associated with reduced infectious complications and better wound healing, and is recommended by the ESPEN guidelines.56
Glutamine supplementation should be given to patients with substantial burns to reduce their complications and improve healing. If enteral nutritional support is not possible, e.g. with gastric stasis, ileus or other coexistent injuries, parenteral nutrition is required.
Several formulae exist for calculating the protein and calorie requirements.57 However, up to 20–25 g of nitrogen per day may be required initially, with a non-protein calorie to nitrogen ratio of 100–200. Energy is provided as carbohydrate and lipids, with the calorie requirement being 35–50% as lipid.
Nutritional supplementation with key nutrients: application to clinical practice
Certain nutrients can have effects on cellular and tissue function. Some of these nutrients modulate immune and inflammatory responses if given in excess of normal intake or requirements. The use of nutrients (‘nutriceuticals’) in this way has been termed ‘nutritional pharmacology’. Examples and specific effects include:
The clinical benefits of supplementation with key nutrients have, however, been difficult to demonstrate.
Combinations Of These Nutrients And Their Place In Practice
Several studies have evaluated the use of combinations of key nutrients in clinical practice in patients with critical illnesses (trauma, surgery for malignant disease, burns), but particularly in upper gastrointestinal cancer. A combination of L-arginine, n-3 essential fatty acids and ribonucleic acid is commercially available (Impact; Sandoz Nutrition, Minneapolis, MN, USA) and has been used in many trials. The supplemented nutrition has been given in the postoperative period (nasoenteric tube or feeding jejunostomy), starting within 12–48 hours of the critical events and continued for several days.
The first meta-analysis of the studies that have compared supplemented nutritional versus standard nutritional diets (Figs 17.2 and 17.3) showed that supplemented nutrition had clinical benefits:66
FIGURE 17.2 Effect of immune-enhancing diets on the incidence of major infective complications (wound infections, intra-abdominal abscesses, pneumonia, septicaemia). Expt, patients receiving immune-enhancing diets; Ctrl, patients receiving standard nutrition; n, number of events; N, number of patients in each group on an intention-to-treat basis; OR, odds ratio; CI, confidence interval. (Study sources are given in Heys et al.60). Reproduced from Heys SD, Walker LG, Smith IC et al. Enteral nutritional supplementation with key nutrients in patients with critical illness and cancer. A meta-analysis of randomised controlled clinical trials. Ann Surg 1999; 229:467–77. With permission from Wolter Kluwers Health.
FIGURE 17.3 Effect of immune-enhancing diets on the length of hospital stay. WMD, weighted mean difference; CI, confidence interval. (Study sources are given in Heys et al.66) Reproduced from Heys SD, Walker LG, Smith IC et al. Enteral nutritional supplementation with key nutrients in patients with critical illness and cancer. A meta-analysis of randomised controlled clinical trials. Ann Surg 1999; 229:467–77. With permission from Wolter Kluwers Health.
However, there was no significant difference in mortality. A subsequent meta-analysis of 17 trials has confirmed this benefit.67
Many of these studies had methodological limitations but, nevertheless, the role of immunonutrition in critically ill patients was further investigated by ESPEN56(www.espen.org/Education/documents/ENICU.pdf). The conclusion drawn from the consensus based on the available evidence was that an immune-modulating nutrition (enriched with arginine, nucleotides and omega-3 fatty acids) was beneficial and recommended for the following:
In addition, there were situations identified where immunonutrition should not be given due to potentially adverse effects:
Immune-modulating nutrition is associated with a reduction in septic complications and a reduced hospital stay. It should be considered in patients with mild sepsis (APACHE II score < 15), patients undergoing elective major intra-abdominal surgery for cancer and in patients with ARDS.56,66,67