Disorders of the small bowel are the most common indication for parenteral nutritional support. Among a series of patients requiring total parenteral nutrition (TPN) at the Massachusetts General Hospital in the period 1 November 1988 to 1 November 1989, prolonged postoperative ileus or intestinal obstruction, sometimes associated with intra-abdominal sepsis or anastomotic leakage, formed the single largest group of patients requiring TPN (Table 1) 321. Enterocutaneous fistula was a relatively uncommon indication for TPN. Other conditions of the small bowel necessitating TPN were mesenteric infarction, massive small bowel resection, malabsorption syndromes, and inflammatory bowel disease.

Four questions need to be addressed when assessing a patient for nutritional support. First, the degree of malnutrition must be estimated: the more malnourished the patient, the more urgent nutritional support becomes. Second, the prognosis or outcome of the disease must be assessed. The need for nutritional support is greater if the underlying condition is unlikely to resolve quickly. Third, an assessment of the patient's calorie, protein, and fluid and electrolyte requirements is a prerequisite to appropriate nutritional support. Finally, the route of administration of nutritional support must be assessed. Parenteral nutrition should only be administered if nutrition cannot safely be administered by the gastrointestinal tract.

Malnutrition may pre-exist in patients presenting for surgery, and the higher risks associated with surgery in such patients are well known. These may not be preventable by preoperative TPN. Alternatively, malnutrition may develop insidiously because of delay in the institution of oral feeding after surgery. The development of malnutrition under these circumstances may be difficult to recognize.

Many methods for documenting nutritional depletion have been described. The simplest is based on the change in usual body weight and the serum albumin estimation. Patients who have lost more than 10 per cent of their usual body weight as a result of illness in less than 2 months may be considered malnourished. A loss of 25 per cent or more of the usual body weight indicates severe malnutrition. The serum albumin estimation below 35 g/l indicates malnutrition if there is no protein losing enteropathy and no nephrotic syndrome. Anthropometric measurements such as arm muscle circumference and triceps skinfold, estimation of the serum transferrin, and skin tests of delayed hypersensitivity have limited application in routine clinical assessment of nutritional status.

Patients undergoing major abdominal surgery suffer few ill effects from a 3- to 5-day period of starvation while awaiting resolution of the paralytic ileus. In such patients, the risks of administering parenteral nutritional support are greater than the benefits. However, in the small minority of patients who develop postoperative complications which prevent the restitution of oral intake, parenteral nutritional support should be considered after 7 days, and even earlier if sepsis develops.

The administration of insufficient calories to meet the needs of the patient will fail to reverse catabolism and the main goals of nutritional support, wound healing, maintenance of skin integrity, and the prevention of loss of lean body mass will not be achieved. Overfeeding may also be hazardous. The over-administration of protein may precipitate the requirement for dialysis in patients with borderline renal failure. Hyperglycaemia resulting from infusion may be difficult to control with insulin. Overfeeding increases CO&sub2; production and may make it difficult to wean patients from ventilatory support. Hyperalimentation is associated with fatty change of the liver. Finally, patients on long-term TPN may become obese if overfed.

The Harris–Benedict equations (Table 2) 322 are one of the earliest methods for determining the calorie requirements at rest. These formulae take into account the height, weight, age, and sex of the patient. Tall heavy young men require almost three times as many calories as small elderly thin women. The basal energy expenditure that this equation calculates must then be multiplied by a stress factor (Table 3) 323. For patients who are severely catabolic due to burns or septicaemia, this stress factor is 2. This method of assessing calorie requirements has been compared to the results of indirect calorimetry and found to be on the generous side.

Having assessed the patient's energy needs, an attempt at estimating the protein requirement should also be made. Patients who are severely catabolic will require up to 2.5 g of protein/kg body weight/day while a patient who is well will only require 1 g protein/kg.day. Patients in renal failure in whom it is hoped to avoid the requirement for dialysis, should have their protein restricted to 40g/day, mainly in the form of essential amino acids. Patients with liver failure may respond to over-administration of protein with a rise in the serum urea and a worsening of hepatic encephalopathy.

The protein component of TPN is administered in the form of amino acid solutions. There is little to choose between the different types of amino acid solutions for patients with disorders of the small bowel except that glutamine, as yet not routinely available, may be an important substrate for maintenance of the gastrointestinal mucosa. In enteral feeds, the source of nitrogen in polymeric diets is whole protein derived from caseinates, soy isolate, and egg white. Predigested or elemental enteral diets are also available, in which the protein is provided as free amino acids or oligopeptides. Elemental diets are expensive and relatively unpalatable and should be reserved for patients who are not able to digest and absorb polymeric diets.

The remaining calorie requirement may be made up by the administration of a combination of dextrose solutions and fat emulsions. Up to 60 per cent of total calorie intake, but not more than 2 g/kg body weight, may be made up of intralipid. Since fat has a respiratory quotient of 0.7, compared with 1.0 for carbohydrates, the use of intralipid as a main source of calories in ventilator dependent patients will result in production of less CO&sub2; than is the case if the calories are made up predominantly with dextrose, and this may aid weaning from the ventilator. For other patients, 500 ml of 10 per cent intralipid twice a week will meet the requirements for essential fatty acids, and the remaining calories may be made up with dextrose. In enteral feeds, fat is derived commonly from corn or soybean oil and carbohydrates from hydrolysed corn starch. The majority of enteral diets are lactose free; some patients develop transient lactose intolerance postoperatively.

The average patient requires between 2 and 3 litres of free water a day with between 90 and 150 mmol/l sodium, and between 60 and 100 mmol/l potassium. For a patient maintained on TPN, a 2000-kcal, 80-g protein diet is made up of a 4 per cent amino acid and 25 per cent dextrose solution, administered via an intravenous catheter placed in a central vein. Losses from the gastrointestinal tract, for example from a high output fistula or nasogastric aspiration, must be replaced by increasing the free water and electrolyte composition of the TPN solution as appropriate. The majority of enteral diets are made up to 1 kcal/ml, a 2000-kcal diet being provided in a volume of 2 litres.

Most TPN regimens include sufficient vitamins to meet the patient's daily requirements, but these will not make up a deficit in the vitamin depleted patient. While it is possible to measure the levels of most vitamins, it is often easier to treat empirically any suspected deficiencies.

Parenteral administration is more expensive, prone to a greater complication rate and may be less effective at reversing nutritional loss than enteral nutrition. Parenteral nutrition also results in atrophy of gastrointestinal mucosa, while enteral nutrition reduces bacterial translocation and maintains mucosal integrity. There is, therefore, every reason to administer nutrition enterally rather than parenterally, if at all feasible.

Ileus following abdominal surgery affects the stomach and colon rather than the small bowel, whose motility and function is often normal in the immediate postoperative period. Nutrients can be administered into the jejunum in the immediate postoperative period, but this is associated with a high incidence of nausea, bloating, vomiting, abdominal pain, and diarrhoea. There is also a risk of aspiration of enteral feed if vomiting occurs, especially in drowsy patients. Carefully selected patients with prolonged ileus, for example, following pylorus-preserving pancreatoduodenectomy, may be successfully nourished by jejunostomy feeding. Enteral nutrition is contraindicated in patients with mechanical intestinal obstruction, intra-abdominal sepsis, and anastomotic leakage. In practical terms, any patient whose postoperative ileus has not resolved within 7 days after surgery should be considered for TPN. The development of intra-abdominal sepsis, anastomotic leakage or Gram-negative septicaemia results in severe catabolism and rapid loss of lean body mass. In such patients, TPN should be given early.

Enterocutaneous fistulae develop as a result of Crohn's disease, radiation enteritis, or following surgery, and may develop more commonly in malnourished patients than in those with a normal nutritional status. Intestinal enterocutaneous fistulae may be divided into two types; high output fistulae, which produce more than 500 ml/day, and low output fistulae. High output fistulae usually arise from the small bowel while low output fistulae commonly communicate with the colon, and both types may be associated with intra-abdominal abscess. Enterocutaneous fistulae will close provided that there is no distal obstruction to the gastrointestinal tract, that intra-abdominal sepsis is adequately drained, and that the nutritional status of the patient is good.

Nutritional problems are common in patients with enterocutaneous fistulae. In the early stages, fluid and electrolyte balance is often unstable and care must be taken to replace electrolytes, including magnesium. Continuing intra-abdominal sepsis may result in a requirement for more calories than expected. Later, specific deficiencies, especially in calcium, the fat-soluble vitamins A, D, E, and K, and vitamin B&sub1;&sub2; may occur, and supplementation in addition to the vitamin preparation routinely added to TPN may be required. Once sepsis has resolved, selected patients may be managed by enteral feeding. Patients with low output fistulae may be managed on low residue enteral feeds. For patients with high output fistulae, enteral feeding may be possible if access to the distal gastrointestinal tract is available.

Patients with inflammatory bowel disease are often malnourished. This is manifested by loss of weight, hypoalbuminaemia, and growth retardation in children. In addition, patients with inflammatory bowel disease may be malnourished as a result of the complications of surgery, including enterocutaneous fistulae and short bowel syndrome. Remission of disease has been reported with bowel rest and total parenteral nutrition, but oral nutritional supplementation may be equally effective. Enteral nutritional supplementation with elemental diets such as Vivonex or Vital HN are useful if the small bowel is extensively affected by Crohn's disease. TPN should probably be reserved for patients with intra-abdominal sepsis, fistulae, or short bowel syndrome.

Specific nutritional deficiencies occur in patients with inflammatory bowel disease. The predilection of Crohn's disease for the terminal ileum may cause vitamin B&sub1;&sub2; depletion. Fat malabsorption related to both Crohn's disease and a lack of bile acids in the enterohepatic circulation may lead to deficiency in the fat-soluble vitamins A, D, E, and K. Folate deficiency may occur in association with the administration of sulphasalazine.

Resection of more than 50 per cent of the small bowel, as a result of mesenteric infarction or multiple resections for Crohn's disease, may result in an inability to maintain nutritional status, although some patients with as little as 25 per cent of the small bowel remaining may eventually adapt to enteral nutrition. Much depends on whether the colon is in continuity with the residual small bowel. Terminal ileal resection has a greater effect on the nutritional status since the jejunum is not able to adapt to absorb bile salts and vitamin B&sub1;&sub2;, while the ileum will adapt to absorb nutrients after jejunal resection. The nutritional deficit which results from massive small bowel resection is compounded by fluid and electrolyte loss. This may be exacerbated by diarrhoea, secondary loss of bile salts, and fatty acids into the colon.

TPN should be administered while the transition to oral fluids and nutrients is made gradually: adaptation of the residual bowel may continue for up to 2 years, and oral intake should be instituted slowly. Some patients require life-long TPN supplementation. Fluid and electrolyte depletion secondary to diarrhoea or high ileostomy output may be helped by glucose–electrolyte replacement solutions or by adding salt to the diet. Many patients are able to tolerate a normal diet but many require enteral supplementation since more calories are needed to compensate for malabsorption. The standard polymeric diets are often more useful than elemental diets, whose high osmolarity may exacerbate intestinal fluid loss. Special attention must be paid to replacement of calcium and vitamin D, the fat-soluble vitamins A, E, and K whose absorption may be impaired along with dietary fats, and vitamin B&sub1;&sub2;, whose specific absorption mechanism may have resected.

The progress of patients on TPN must be monitored according to clinical indications. Daily estimation of fluid balance, weight, and plasma electrolytes and glucose are essential in the early stages of instituting nutritional support. Weekly estimation of calcium, liver function, albumin level, the haematological profile, and trace element level are also important. A decline in weight, the development of muscle wasting, the failure of wounds to heal, or the development of pressure sores should indicate that nutritional support needs to be reviewed.

Alexander JW, Nutrition and translocation. J Parenteral Enteral Nutr 1990; 14 (suppl): 170–4S.

Benedict FG. Lectures on Nutrition. Philadelphia: W B Saunders & Co, 1925: 31–54.

Daly JM, Bonau R, Stofberg P, Bloch A, Jeevanandam M, Morse M. Immediate postoperative jejunostomy feeding: clinical and metabolic results in a prospective trial. Am J Surg 1987; 153: 198–204.

Detsky AS, Baker JP, O'Rourke K, Goel V. Perioperative parenteral nutrition: a meta-analysis. Ann Intern Med 1987; 107: 195–203.

Greenberg GR, Fleming CR, Jeejeebhoy KN, Rosenberg IH, Sales D, Tremaine WJ. Controlled trial of bowel rest and nutritional support in the management of Crohn's disease. Gut 1988; 29: 1309–15.

Harris JA, Benedict FG. A Biometric Study of Basal Metabolism in Men. Publication No. 279. Washington DC: Carnegie Institute of Washington, 1919.

Jeejeebhoy KN. Energy metabolism in the critically ill. London: John Libbey & Co Ltd; 1985: 93–101.

Mann S, Westenskow DR, Houtchens BA. Measured and predicted caloric expenditure in the acutely ill. Crit Care Med 1985; 13: 173–7.

Matuchansky C. Parenteral nutrition in inflammatory bowel disease. Gut 1986; 27: 81–4.

Michel L, Serrano A, Malt RA. Nutritional support of hospitalized patients. N Engl J Med 1981; 304: 1147–52.

Rombeau JL, Rolandelli RH. Enteral and parenteral nutrition in patients with enteric fistulas and short bowel syndrome. Surg Clin N Am 1987; 67: 551–71.

Souba WW, Herskowitz K, Salloum RM, Chen MK, Austgen TR. Gut glutamine metabolism. J Parenteral Enteral Nutr 1990; 14 (suppl). 45–50S.

Studley HO. Percentage of weight loss; a basic indicator of surgical risk in patients with chronic peptic ulcer. JAMA 1973; 106: 458–60.

Veterans Affairs total parenteral nutrition cooperative study group. Perioperative total parenteral nutrition in surgical patients. N Engl J Med 1991; 325: 525–32.

Wells C, Tinckler L, Rawlinson K, Jones H, Saunders J. Postoperative gastrointestinal motility. Lancet 1964; i: 4–10.