EDUCATIONAL REVIEW

Nutrition in the critically ill patient

Introduction Malnutrition in critically ill patients General goals of nutritional support

Route of administration Specific considerations Nutritional additives for special purposes Conclusion References

Critically ill patients are hypermetabolic and have increased nutrient requirements. Although it is assumed that nutritional support is beneficial in this group of patients there are no well designed clinical trials to test this hypothesis. The rationale for nutritional support, therefore, is based upon clinical judgement. Although it is not known how long a critically ill patient can tolerate what is effectively starvation, the loss of lean tissue which occurs in catabolic patients (20-40g nitrogen/day) suggests that depletion to a critical level may occur after 14 days. Total parenteral nutrition given to malnourished patients with gastrointestinal cancer for 7 to 10 days before surgery has been shown to decrease complications by about 10%. Wound healing and normal immune responses are dependent upon adequate nutrient intake, and it seems reasonable, therefore, to commence feeding as soon as possible. Earlier feeding may decrease length of stay and complication rates in both critically ill patients and following surgery. It has been shown in randomised controlled trials of both enteral and parenteral feeding in the critically ill, that current regimens are sub-optimal. Calorie intake was shown to be often considerably in excess of metabolic requirements alone, and variable study design has made comparisons impossible. Despite the lack of knowledge regarding even the most simple of nutritional requirements in these patients, the administration of specific micronutrients and specialised supplements has attracted attention. Again, many of these more recent studies are limited because of poor study design.

Keywords: Enteral nutrition, glutamine, nutritional antioxidants, parenteral nutrition

J.R.Coll.Surg.Edinb., 45, December 2000, 373-379 

INTRODUCTION

Over the past 30 to 40 years advances in enteral feeding techniques, venous access and enteral and parenteral nutritional formulations have made the provision of nutritional support possible for most critically ill patients. Specialised formulations and additives providing either specific micronutrients and other so-called nutritional support agents are now available. However, although there is widespread use of nutritional therapy and a wealth of medical literature on this subject, many areas of nutritional support remain controversial. Discussion on this topic is hampered by the relative lack of well controlled randomised trials and much of the literature is in the form of case series or meta-analyses.

It is assumed that nutrient depletion is associated with increased morbidity and mortality, and that correction or prevention of such nutrient depletion can eliminate this excess morbidity and mortality associated with malnutrition. There are many factors that determine outcome from critical care and indeed any hospital treatment, and nutritional status clearly has an effect. However, the interaction between malnutrition and the other factors which influence outcome, makes it very difficult to isolate the independent contribution of nutritional status. Consequently, there are few data that conclusively support this generally assumed interaction.

Nutritional support has become a routine part of the care of critically ill patients and it is now widely accepted for the treatment and prevention of malnutrition and specific nutrient deficiencies. It is also generally assumed to improve outcome. As with most other forms of treatment there are, however, concomitant adverse effects.

A number of comprehensive textbooks have been written on the subject of nutritional support in the critically ill, and this review, therefore, will focus on the current 'state of the art' relating to nutritional approaches in the critically ill adult. The review will cover most of the information required for the Diploma examination candidate to demonstrate adequate knowledge in this topic. In addition, recent and possibly controversial information will be presented to enable the reader to formulate an opinion on the relative merits of specific nutrient supplementation and the role of very early enteral nutrition following surgery or trauma.

MALNUTRITION IN CRITICALLY-ILL PATIENTS

The adaptive changes which occur during starvation and as part of the stress response after injury are now well known and are directed at conserving body protein and maintaining normoglycaemia, whilst ensuring that the body is well placed to fight infection and undergo subsequent necessary healing processes.¹ The term 'malnutrition' refers to a generalised disordering of body composition such that either macro- or micro- nutrient intake is sub-optimal. In effect, malnutrition would be better referred to as a nutritional deficit. Severe deficits can result in organ dysfunction, abnormalities in blood biochemistry and reduced body mass, and may also be associated with altered outcome following hospital treatment. Although a clear definition of malnutrition has caused some difficulty, it is thought that most critically ill patients suffer to some degree, particularly those patients who have been treated on the intensive care unit (ICU) for more than a few days.

Assessment of the state of nutrition, or more accurately the level of nutritional deficit, of a patient is clearly the first step in deciding the degree of nutritional support required. General loss of body weight associated with a catabolic process leading to loss of muscle mass as a result of protein and calorie deficits, provides a gross indicator of inadequate nutritional intake. However, in the critically ill patient, oedema formation occurs frequently and accurate weighing is often impossible, such that nutritional deficits are not always obvious. In an ideal situation, a measure of body mass index (BMI): weight (kg) divided by height (m) squared is a useful indicator of nutritional state.

Body mass index = Weight(kg)/Height (m)²

Unintentional weight loss greater than 10% in the six months before hospital treatment has been shown to reflect clinical outcome in a general surgical population.² Anthropometry is widely used in some hospital patients to determine skin fold thickness and skeletal muscle mass as an index of nutritional state, but is of limited value in the ICU setting. Muscle function tests such as hand grip strength can also be performed, but again they are of little use since an awake, alert and co-operative patient is usually required. In addition, the interpretation of such data is complicated by metabolic derangements such as hypercapnoea, hypoxia, hypophosphataemia and the effect of various medications. However, pre-operative handgrip dynamometry and respiratory muscle strength assessment (maximal voluntary ventilation and vital capacity and maximal airway pressures) have been shown to predict postoperative complications.³

Aspects of immune function are altered by nutritional depletion and this can be assessed by measures such as delayed hypersensitivity responses to a variety of antigens -again this has been little used in the critically-ill patient. Plasma protein concentrations - usually albumin or less frequently transferrin, retinol binding protein or prealbumin have also been used to monitor nutritional status. These biochemical indices are generally fairly insensitive in terms of reflecting nutritional deficit since production may be influenced by factors other than nutritional status and plasma half-lives also vary. In addition, the normal rate of albumin exchange between intra- and extra-vascular compartments is more than ten times the rate of either synthesis or degradation. In view of this, even relatively minor variations between these compartments can have profound effects on circulating albumin levels. Plasma albumin concentrations are usually not affected by nutritional intake and will not increase in metabolically stressed patients until the cause of the stress is removed. In fact in chronic nutritional depletion, plasma albumin levels may even increase due to a combination of dehydration, decreased protein degradation and movement of extra-vascular albumin into the intra-vascular compartment.⁴ Several body composition analysers are available and although some have been used in critically ill patients⁵ at present these remain research tools. In summary, there is no good indicator of nutritional status in the critically ill patient (Table 1).

Table 1: Tests for nutritional deficit

Why is it that critically ill patients become nutritionally depleted? Various disease processes, common to ICU and high dependency units (HDU), cause changes in substrate metabolism which lead to changes in body composition and, ultimately, relative deficiencies of some nutrients. The body has reserves from which it can liberate calories for the production of energy - essentially these are as carbohydrates and fats. Some tissues have an absolute requirement for glucose and small amounts of glucose are also required for metabolism of fat. Although reserves of glucose are relatively small (glucose in the circulation and glycogen in liver and muscle), glucose can also be formed from certain amino acids by a process called gluconeogenesis. During starvation, i.e. nutritional deficit due to inadequate intake of all nutrients, increased oxidation of fat provides the main source of energy, and nitrogen losses are reduced by enhanced fat mobilisation. In the early phase of starvation, both fat and protein are consumed but protein loss is minimised and only limited muscle breakdown occurs to enable gluconeogenesis. However, severe muscle wasting can occur late in starvation when fat reserves are depleted. In conditions such as sepsis, where hypermetabolism occurs, there is accelerated protein breakdown both to provide energy by gluconeogenesis and also to liberate the amino acids required for increased protein synthesis. However, the situation is obviously not that simple since the provision of even a great excess of calories above and beyond requirements, does not eliminate the protein breakdown seen in most long stay ICU patients.

GENERAL GOALS OF NUTRITIONAL SUPPORT

The overall aim of nutritional support is to provide patients with their general nutrient requirements. However, it should be remembered that requirements will need to be modified according to individual patient needs and specific disease processes. Requirements for particular nutrients depend upon both utilisation and rate of loss. Nutritional support is often considered in terms of macronutrients and micronutrients. As a first step to working out what form of nutrition the patient needs it is often best to start with estimating total fluid requirements. As a rule of thumb this will be between 30 and 40 ml/kg/day or 1 ml water per calorie for an adult, but this will need to be supplemented should fluid losses be excessive. These losses can be either overt (excess urine, upper gastrointestinal losses or diarrhoea) and insensible - it is only possible to guess what these latter losses are. In ICU and HDU patients, excessive insensible losses can occur as a result of pyrexia or through the use of high loss air mattresses where evaporative loss from the skin is increased.

Total calorie requirements can be either estimated or calculated. Although calorie overload has been shown to be harmful and should be avoided, precisely balancing calorie input with energy expenditure has not been shown to improve outcome. To measure energy expenditure it is usual to measure oxygen consumption using a metabolic cart. Energy requirement can also be calculated by measuring nitrogen balance using 24 h urinary urea and in relation to blood urea and albumin. Knowing the nitrogen loss it is then possible to calculate calorie requirement since each gram of nitrogen produced uses 100-150kcal of energy. However, more usually used techniques estimate rather than calculate, calorie requirement. The Harris-Benedict equation can be used or even more simply, requirements can be guessed - 25 total kilocalories /kg/day is the estimate most often used. Calories can be given in three forms:

• Carbohydrate: 30 to 70% of the total calories can be given as carbohydrate. This is usually given as glucose but fructose and sorbitol are used in some countries. Insulin may be required to maintain blood glucose concentration within normal limits, especially since insulin resistance is often seen as part of the response to stress.
• Fat: 20 to 50% of the total calories can be given as fat. Critically ill patients often utilize fat better than carbohydrate as an energy source and although our normal diet contains around 30% fat (much higher in some western diets), it is often advantageous to provide more than this to patients on ICU or HDU. Provision of fat also provides essential fatty acids required as normal constituents of cells. Omega-6-polyunsaturated fatty acid (PUFA) triglycerides should be provided to prevent essential fatty acid deficiency - at least 7% of total calories. Medium and long chain triglycerides can also be used as a source of calories and the precise ratio of medium to long chain triglycerides is dependent on which product is used and route of administration. There is no clear evidence to confirm the superiority of any particular triglyceride regimen over any other in the typical critically ill patient.
• Protein: 15 to 20% of the total calories per day can be administered as protein or amino acids depending on route of administration.

Finally, micronutrient requirements should be considered. Approximately 1mmol/kg of both sodium and potassium are usually given but this figure will need to be altered when there are excessive losses - this is particularly common from excess sweating and gastrointestinal losses. Other electrolytes, e.g. magnesium, iron, copper, zinc and selenium are also necessary, but in much smaller amounts. Patients on long-term nutritional supplementation will need to have levels of these electrolytes checked periodically. An often forgotten electrolyte is phosphate and this is important since it is required for normal metabolic processes resulting in the formation of ATP. Muscle weakness associated with a long-term requirement for ventilatory support, and failure in weaning from ventilation is associated with hypophosphataemia.

Other micronutrients include fat (vitamins A, carotene) and water-soluble vitamins (B, C, D, E). The precise requirements for specific vitamins remains unclear, although several studies have shown extremely low circulating concentrations.⁶

The Harris-Benedict equation: used to calculate resting energy expenditure, REE, along with the usual multiplication factors to provide adeqaute calorie intake

Women: REE = 655 + (9.6 X weight in kg) + (1.7 X height in cm) - (4.7 X age in years)

Men: REE = 66 + (13.7 X weight in kg) + (5.0 X height in cm) - (6.8 X age in years)

Calorie requirements/day = 1.25 X REE (for each 1°C above 37 add 10% extra allowance

ROUTE OF ADMINISTRATION

Enteral Route

This is the preferred route for nutritional support since the gut barrier and immune functions are preserved and systemic infections and other complications are reduced.⁷ Many studies support the implementation of enteral nutrition as soon as possible after resuscitation. Of course, enteral feeding requires adequate gastric motility and a gastric residual volume in excess of 150 ml will usually require feeding solutions to be administered slowly. However, it is not necessary to have bowel sounds for successful enteral nutrition. Persistently inadequate gastric emptying may dictate small bowel feeding techniques (duodenal or jejunal) or supplemental intravenous nutritional support. The main disadvantage with enteral feeding is that if feeding needs to be stopped periodically to allow gastric emptying, then the calorie target will not be met. Indeed, it has recently been suggested that the combination of enteral and parenteral may be appropriate for most critically ill patients, with enteral feeding increasing as tolerance to the feed increases.⁸

Agents which promote gastric and intestinal motility such as metoclopramide and erythromycin can be used although cisapride has recently been withdrawn for use in the USA and UK due to it's ability to cause serious dysrythmias. Cisapride selectively enhances cholinergic motor activity throughout the gastro-intestinal tract; erythromycin increases motilin, a substance which enhances contractile activity of the gastric antrum and duodenum; metoclopramide is a selective dopamine-2-receptor antagonist which increases peristaltic contractility of the oesophagus, gastric antrum and jejunum. Neither erythormycin nor metoclopramide have effects that may affect healing in the large bowel. A single enteral dose of any of these agents promotes gastric emptying in critically ill patients with gastric motility dysfunction but further studies are required to assess the benefit in terms of increasing tolerance to enteral feeding.^9,10

If factors affecting the ability to tolerate early postoperative enteral feeding can be identified, perhaps these factors can be minimised. In a recent study, 200 patients undergoing colon resection received early post-operative feeding comprising clear fluids on the second post-operative day and a normal diet on day 3. Only 27 (13.5%) of patients failed to tolerate early feeding, the majority of whom were men. Importantly, there were no anastomotic leaks or abdominal abscesses. There was no effect of metoclopramide on the tolerance to early feeding in a sub-group of patients, and no increase in anastomosis breakdown. These data suggest that early post-operative feeding is safe and effective, with reduced hospital stays.¹¹

The absence of bowel sounds or passage of flatus do not preclude enteral feeding, especially when this is instituted distal to the pylorus. Small bowel feeding is now generally advocated in mild or resolving pancreatitis and low output enterocutaneous fistulae (<500ml/day). Two problems, however, may limit the use of this and other forms of enteral nutrition - abdominal distension (which may be sufficient to result in compromised respiratory function) and diarrhoea (which although usually secretory and associated with a high osmotic load to the colon may be associated with Clostridium difficile enterocolitis).

There is some evidence that enteral nutrition may be better than total parenteral nutrition (TPN) in terms of infection rate.¹² TPN may of course be associated with a higher infection rate rather than enteral feeding reducing the incidence of infections. Enteral feeding preserves the gastrointestinal barrier function that can prevent or decrease translocation of bacteria across the bowel wall which may decrease nosocomial infections. However, enteral feeding does increase the risk of ventilator associated pneumonia,¹³ probably because it increases gastric pH and promotes gastric colonisation. In addition, the gastric tube compromises lower oesophageal sphincter function.

Parenteral Route

This should only be used when the enteral route is either not possible or cannot provide sufficient nutrient input. Complications are more frequent with the parenteral route and are usually related to catheter insertion and infection. It should be remembered that TPN can be delivered into peripheral veins via very fine-bore catheters, as well as through central lines. Lines used for TPN should be inserted under full aseptic technique and TPN should be delivered through a dedicated catheter. Interruptions and reconnections should be kept to a minimum and again carried out under aseptic conditions. A TPN team, responsible for inserting the feeding catheters, prescribing the feed required, adjusting the nutritional regimen as necessary in individual patients and monitoring the occurrence of line-associated sepsis, is used in many hospitals.

Although it is generally assumed that TPN may provide targeted nutritional support for critically ill patients, it has been shown that in many patients calorie and nitrogen target intakes were not reached, and underfeeding may be a genuine problem in some patients.¹⁴

The use of TPN in the critically ill has been summarised in a meta-analysis which included 26 randomised trials involving 2211 patients, although only 6 of the studies included were based in ICU¹⁵. There was no effect on mortality although there was a slight but not statistically significant decrease in major complications in those patients given TPN. The authors suggested that TPN should only be used in those critically ill patients who cannot tolerate enteral nutrition. However, it is not clear how many critically ill patients cannot really tolerate enteral nutrition. The authors, perhaps, should have suggested that TPN be used only for the pre-operative treatment of nutritionally depleted surgical patients who cannot tolerate feeds given by the enteral route.¹⁶

SPECIFIC CONSIDERATIONS

Sepsis

Sepsis is characterised by a catabolic state such that total calorie requirements are increased and rapid net protein breakdown occurs. Increases in calorie intakes of about 10-20% and higher nitrogen intakes may be required. Micronutrient and electrolyte requirement are also increased and electrolyte concentrations should be monitored frequently. Hyperglycaemia may be present and there may be a requirement for intravenous infusion of insulin. Hypertriglyceridaemia with lipaemic serum may occur and fat intake may need to be decreased. The sedative agent propofol is prepared in a fat emulsion and may cause excess fat intake in those patients receiving propofol infusions.

LIVER FAILURE

Patients with liver failure may have marked pre-existing electrolyte abnormalities and may have been treated by fluid restriction in an attempt to decrease the formation of ascites. Such patients commonly have increased potassium, magnesium and zinc losses and may be quite severely hyponatraemic despite having high total body sodium concentrations. Rapid correction of plasma sodium concentration should be avoided since total body sodium is often abnormally high and central pontine myelinolysis may result.

Acute liver failure is often accompanied by encephalopathy that may be due to accumulation of ammonia as a result of defects in the urea cycle in the damaged liver. Administration of certain amino acids can exacerbate encephalopathy and nitrogen sources with increased branched chain amino acids and reduced aromatic sulphur containing amino acids are available.

RESPIRATORY FAILURE

Overfeeding is the major problem associated with feeding patients with compromised respiratory function. The ratio of carbon dioxide production to oxygen consumption is called the respiratory quotient - R/Q - and this is normally between 0.85 and 0.90. Metabolism of fat is associated with an R/Q of 0.7, whilst glucose metabolism gives an R/Q of 1.0. Patients with respiratory failure may have difficulty in eliminating carbon dioxide and a feed regimen associated with lower R/Q has been suggested. In practice, this has little impact while patients are undergoing mechanical ventilation, but may be important when patients are weaned from the ventilator.¹⁷ It is probably more important to ensure that calorie intake is the minimum required and an R/Q greater than 1.0 (which usually indicates overfeeding) should be avoided.

Although, in theory, administration of protein is associated with an increase in oxygen consumption this is of little practical importance since small increases in inspired oxygen tension overcomes the problem.

RENAL FAILURE

Nutritional support in patients with renal failure depends both on the frequency and type of renal support therapy and whether the renal failure is acute or chronic. Most patients are sensitive to excess fluid loads and care, therefore, is required with electrolyte administration, particularly potassium, magnesium and phosphate. Low volume/low sodium feeds are available for enteral nutrition in patients with renal compromise. Nitrogen intake may also need to be reduced to between 0.5 and 0.8g N₂/kg/day in patients with chronic renal failure.

Amino acids will be lost during haemodialysis or haemofiltration and intake may require increasing, although there is no advantage in giving only the essential amino acids. Fluids used in peritoneal dialysis usually contain glucose providing a 'hidden' source of calorie intake in such patients.

ACUTE PANCREATITIS

Eating during an episode of acute pancreatitis causes pain and increased pancreatic enzyme release and this has led to many studies that have evaluated various alternative feeding regimens. TPN has little stimulatory effects on pancreatic secretion and intravenous lipids are usually well tolerated. The effect of enteral feeding is unclear. Some studies have shown no effect on pancreatic secretion during jejunal feeding while others show a stimulatory effect. Jejunal feeding in patients with mild to moderate pancreatitis was well tolerated and less expensive, as shown in a recent study. ¹⁸ Although only 82% of enterally fed patients received their target calorie intake, compared with 96% of patients given TPN, clinical outcome was the same. There are no randomised controlled trials which have evaluated the clinical efficacy of nutritional support by any particular route in patients with severe pancreatitis.

NUTRITIONAL ADDITIVES FOR SPECIAL PURPOSES

Branched chain amino acids

The branched chain amino acids leucine, isoleucine and valine are essential amino acids required for synthesis of proteins. The clinical benefit of preparations containing these amino acids has been evaluated in several small randomised controlled trials in patients with liver disease and these have been summarised in a meta-analysis.¹⁹

Although a small statistically significant improvement in recovery from encephalopathy in the short-term was reported, some of the original publications may have inadvertently favoured the patients receiving the branched-chain amino acids. The benefit of branched chain amino acids in improving outcome generally in ICU and HDU patients is unproven. There is some evidence that the respiratory centre is stimulated and muscle function is improved, leading to their use in patients who are difficult to wean from ventilator support.

Peptides and elemental diets for enteral nutrition

Protein absorption and tolerance of enteral nutrition may be improved by the use of feeds containing peptides rather than whole proteins. The role of such formulations in improving outcome in the critically ill has yet to be determined, although these feeds are helpful in specific malabsorption conditions such as Crohn's disease.

Immuno-nutrients

Arginine is an amino acid required for variety of metabolic functions including urea synthesis, lymphocyte proliferation and wound healing. Arginine also stimulates the release of hormones including insulin, prolactin and growth hormone and is needed for nitric oxide production, which regulates blood flow. These effects may be beneficial in, for example, the gastrointestinal tract.^20,21 Arginine has been given postoperatively in doses of up to 30 g/day to adult patients but the benefit remains unclear.

Glutamine has also been used as a nutritional supplement. It is a so-called conditionally essential amino acid, since although dietary intake is normally adequate, under certain conditions such as serious illness, requirements are increased and supplementation may be needed. It is one of the three amino acids in the important antioxidant compound glutathione and it is used as a substrate for metabolism by leucocytes and enterocytes. Leucocytes can derive almost as much energy from the metabolism of glutamine as glucose, and glutamine, therefore, can augment lymphocyte and macrophage function. Omega-3 polyunsaturated fatty acids are currently being evaluated as both immune modulators and antiinflammatory agents and doses of up to 5 g/day have been used in critically ill patients with sepsis.

The effect of combinations of immunonutrients on outcome has been investigated. Enteral feeds supplemented with omega-3 fatty acids, arginine and ribonucleic acids, and also glutamine in some studies, have been compared with standard enteral feeds in critically ill patients, including medical, surgical and trauma patients. ^22-26

A trend towards higher early mortality in immunonutrienttreated patients was seen in one study: less multi-organ failure, fewer infection-related complications, and reduced requirements for mechanical ventilation have been reported.

In one study of parenteral feeding, patients were followed up for 6 months and survival was better in patients who had received glutamine supplemented TPN.²⁷ A meta-analysis of 11 randomised trials of supplemented enteral feeds revealed a significant decrease in the risk of developing major infectious complications in the immunonutrition groups and decreased hospital stay.²⁸ However, there was also an increased risk of death although this did not reach statistical significance.

Nucleic acids

The role of specific replacement of nucleic acids in feeding regimens remains to be determined and a meta-analysis concluded that little evidence supports the use of such supplementation in humans despite the theoretical effects on cell-mediated immunity and lymphocyte function.²⁹

Fibre

Fibre can be added to enteral nutrition preparations and has beneficial effects on colonocyte function. Studies in rats have shown that fibre or glutamine-supplemented enteral feeds promote healing of experimental colonic anastomoses.³⁰

Growth hormone

Growth hormone is released endogenously during critical illness as part of the stress response, but administration of exogenous growth hormone has been used for its anabolic effects. These include increased fat oxidation and protein synthesis and improved immune responsiveness. Although there have been many individual case reports and small case series showing benefit, particularly in burns patients, a recent large multi-centre study, however, has shown significantly higher mortality in long stay ICU patients treated with growth hormone.³¹ However, growth hormone may still have a role in patients with burns, particularly children.

CONCLUSION

For both enteral and parenteral nutritional support, it is clear that many critically ill patients do not receive their target intake, due to elective interruption as a result of intolerance, interruption due to other therapeutic interventions, and practical factors such as inadequate stock of feeding solutions. Early enteral feeding has been suggested to be beneficial in the critically ill, but dogged adherence to the enteral route alone might mean that many patients are inadequately fed. The mode of delivery of nutritional support should be determined by needs and for many patients this will mean combinations of enteral and parenteral nutrition, adjusted as tolerance to enteral feeding increases.

REFERENCES

1. Cahill GF, Herrera MG, Morgan AP et al. Hormone-fuel interrelationships during fasting. J Clin Invest 1996; 45: 1751-69
2. DeWys WD, Begg C, Lavin PT et al. Prognostic effect of weight loss prior to chemotherapy in cancer patients. Am J Med 1980; 69: 491-7
3. Windsor JA, Hill GL. Weight loss with physiologic impairment: a basic indicator of surgical risk. Ann Surg 1988; 207: 290-6
4. Klein S. The myth of serum albumin as a measure of nutritional status. Gastroenterol 1990; 99: 1845-6
5. Roos AN, Westendorp RGJ, Brand R, Souverijn JHM, Frolich M, Meinders AE. Predictive value of tetrapolar body impedance measurements for hydration status in critically ill patients. Intensive Care Med 1995; 21: 125-31
6. Goode HF, Cowley HC, Walker BE, Howdle PD, Webster NR. Decreased antioxidant status and increased lipid peroxidation in patients with septic shock and secondary organ dysfunction. Crit Care Med 1995; 23: 646-51
7. Carr CS, Ling KD, Boulos P, Singer M. Randomised trial of safety and efficacy of immediate postoperative enteral feeding in patients undergoing gastrointestinal resection. Br Med J 1996; 312: 869-71
8. MacFie J. Enteral versus parenteral nutrition. Br J Surg 2000; 87: 1121-2
9. MacLaren R, Kuhl DA, Gervasio JM et al. Sequential single doses of cisapride, erythromycin, and metoclopramide in critically ill patients intolerant to enteral nutrition: a randomized, placebo-controlled, crossover study. Crit Care Med 2000; 28: 438-44
10. Jooste CA, Mustoe J, Collee G. Metoclopramide improves gastric motility in critically ill patients. Intensive Care Med 1999; 25: 464-8
11. DiFronzo LA, Cymerman J, O'Connell TX. Factors affecting early postoperative feeding following elective open colon resection. Arch Surg 1999; 134: 941-6
12. Moore FA, Feliciano DV, Andrassy RJ et al. Early enteral feeding compared with parenteral, reduces postoperative septic complications. The results of a meta-analysis. Ann Surg 1992; 216: 172-83
13. Drakulovic MB, Torres A, Bauer TT, Nicolas JM, Nogue S, Ferrer M. Supine body position as a risk factor for nosocomial pneumonia in mechanically ventilated patients: a randomised trial. Lancet 1999; 354: 1851-8
14. Green CJ, Campbell IT, McCleland P et al. Energy and nitrogen balance and changes in mid upper-arm-circumference with multiple organ failure. Nutrition 1995; 11: 739-46
15. Heyland DK, MacDonald S, Keefe L, Drover JW. Total parenteral nutrition in the critically ill patient: a meta-analysis. J Am Med Assoc 1998; 280: 2013-9
16. Kennedy BC, Hall GM. Metabolic support of critically ill patients: parenteral nutrition to immunonutrition. Br J Anaesth 2000; 85: 185-7
17. Van den Berg B, Bogaard JM, Hop WCJ. High fat, low carbohydrate, enteral feeding in patients weaning from the ventilator. Intensive Care Med 1994; 20: 470-5
18. McClave SA, Greene LM, Snider HL et al. Comparison of the safety of early enteral vs parenteral nutrition in mild acute pancreatitis. J Parent Ent Nutr 1997; 21: 14-20
19. Naylor CD, O'Rourke K, Detsky AS et al. Parenteral nutrition with branched-chain amino acids in hepatic encephalopathy. A meta-analysis. Gastroenterol 1989; 97: 1033-42
20. Chang DW, DeSanti L, Demling RH. Anticatabolic and anabolic strategies in critical illness: a review of current treatment modalities. Shock 1998; 10: 150-60
21. Gianotti L, Alexander JW, Pyles T, Fukushima R. Arginine-supplemented diets improve survival in gutderived sepsis and peritonitis by modulating bacterial clearance. The role of nitric oxide. Ann Surg 1993; 217: 644-53
22. Bower RH, Cerra FB, Bershadsky B et al. Early enteral administration of a formula (Impact) supplemented with arginine, nucleotides, and fish oil in intensive care unit patients: Results of a multicenter, prospective, randomized clinical trail. Crit Care Med 1995; 23: 436-49
23. Moore FA, Moore EE, Kudsk KA et al. Clinical benefits of an immune-enhancing diet for early postinjury enteral feeding. J Trauma 1994; 37: 607-15
24. Kudsk KA, Minard G, Croce MA et al. A randomized trial of isonitrogenous enteral diets after severe trauma. Ann Surg 1996; 224: 531-43
25. Atkinson S, Sieffert E, Bihari D, on behalf of the Guy's Hospital Intensive Care Group. A prospective, randomized, double-blind, controlled clinical trial of enteral immuno-nutrition in the critically ill. Crit Care Med 1998; 26: 1164-72
26. Weimann A, Bastian L, Bischoff WE et al. Influence of arginine, omega-3 fatty acids and nucleotide supplemented enteral support on systemic inflammatory response syndrome and multiple organ failure in patients after severe trauma. Nutrition 1997; 14: 165-72
27. Griffiths RD, Jones C, Allan Palmer TE. Six-month outcome of critically ill patients given glutamine-supple-mented parenteral nutrition. Nutrition 1997; 13: 285-302
28. Heys SD, Walker LG, Smith I, Eremin O. Enteral nutritional supplementation with key nutrients in patients with critical illness and cancer: a meta-analysis of randomized controlled clinical trials. Ann Surg 1999; 229: 467-77
29. Heyland DK, Cook DJ, Guyatt GH. Does the formulation of enteral feeding products influence the infectious morbidity and mortality rates in the critically ill patient? A critical review of the evidence. Crit Care Med 1995; 22: 1192-1202
30. Demetriades H, Botsios D, Kazantzidou D et al. Effect of early postoperative enteral feeding on the healing of colonic anastomoses in rats. Comparison of three different enteral diets. Eur Surg Res 1999; 31: 57-63
31. Takala J, Ruokonen E, Webster NR et al. Increased mortality associated with growth hormone treatment in critically ill adults. N Engl J Med 1999; 341: 785-92

Copyright date: 9th October 2000

Correspondence: Professor N.R. Webster, Anaesthesia and Intensive Care, Institute of Medical Sciences, Foresterhill, Aberdeen AB25 2ZD, U.K.

Email: n.r.webster@abdn.ac.uk

©2000 The Royal College of Surgeons of Edinburgh, J.R.Coll.Surg.Edinb. 373-379