Proximal Intestinal Loss

When mainly proximal intestine is lost, which is unusual in pediatric SBS, clinical impact on digestion is generally small. In some patients, increased gastric fluid output offsets the usual contribution of the small bowel to the total volume secreted. Gastric acid secretion increases in proportion to the magnitude of the resection, possibly because of a parallel reduction in release of enteric hormones, such as somatostatin, that inhibit acid production. In some patients, gastric hyperacidity may cause transient malabsorption by inactivating pancreatic enzymes and precipitating bile acids. However, diarrhea is rarely substantial after proximal intestinal resection, because the more distal jejunum, ileum, and colon have sufficient functional reserve to increase fluid uptake three- to fivefold. Similarly, protein and carbohydrate absorption are minimally affected, because their complete assimilation requires only about one-third of normal small bowel length. Only lipids require the entire small bowel for normal uptake, so intestinal resection from any region may produce nutritionally significant lipid malabsorption.

Distal Intestinal Loss

Resection of more distal small intestine, particularly ileum, generally reduces nutrient, fluid, and electrolyte absorption more than resection of an equivalent length of proximal jejunum. The concept that loss of ileum should significantly impair nutrient assimilation is counterintuitive, since most macronutrients are assimilated in the upper small bowel. However, loss of all or most of this region of the small intestine has two consequences in addition to loss of length per se .

First, in contrast with the duodenum and jejunum, only ileum actively reabsorbs bile acids. If around one-third or more of the ileum is lost, about 100 cm in adults and 50 cm in children, the compensatory increase in hepatic bile acid synthesis will not keep pace with increased fecal bile acid loss. In that case, the proximal intestinal lumen bile salt concentration will be inadequate for efficient lipid emulsification, thereby contributing to fat malabsorption. Furthermore, fat malabsorption increases colon fluid loss, both because long-chain fatty acids hydroxylated by colonic bacteria stimulate colonocyte electrolyte (and thereby water) secretion, particularly potassium and bicarbonate, and because long-chain fatty acids are themselves impermeable to colon epithelium, further increasing the total solute concentration in the colon lumen. A lesser amount of ileal loss may produce bile acid malabsorption without causing total body bile acid depletion or fat malabsorption. However, diarrhea may still result, since bile acids as well as long-chain fatty acids may stimulate colonic water secretion.

Second, loss of ileum and probably the ileocecal valve and colon adversely affect motility of more proximal gut. Ileal loss, that is, removal of the “ileal brake,” accelerates gastric emptying of liquids and increases proximal small bowel motility directly, thereby shortening total intestinal transit time independent of that resulting from reduction in intestinal length per se . The result is a further reduction in contact between luminal contents and the mucosal surface, adding to the aggregate reduction in nutrient, fluid, and electrolyte assimilation. Hormones normally secreted by the distal ileal and colonic mucosa, including peptide YY (PYY) and glucagon-like peptide 1 (GLP-1), probably mediate the ileal brake. The relative contributions of the distal ileum, ileocecal valve, and proximal colon in slowing proximal motility remain incompletely defined.

Colon Loss

The colon normally absorbs only a small fraction of water reclaimed during digestion, about 10% to 15% of the total. The colon also normally plays only a secondary role in digestion; at most, 20% of complex dietary starch and even less (<5%) dietary nitrogen and lipid escape small intestinal absorption. Starches that escape absorption in the small intestine and also soluble fibers are salvaged in the colon with fermentation to bioavailable short-chain fatty acids—primarily acetic, propionic, and butyric acid—by resident anaerobic bacteria. An increased load of carbohydrates (and proteins) enters the colon following massive small intestinal resection, thereby increasing substrate available for fermentation; up to 1100 kcal may be recovered daily in this fashion. Uptake of short-chain fatty acid molecules in the colon also creates an osmotic gradient that enhances water absorption, thereby limiting total fecal fluid loss.

When all or most of the colon is lost, outcome depends on the magnitude of concurrent small intestinal loss. Patients who have relatively preserved small bowel that ends in an ileostomy do not generally lose enough calories to affect nutrition, and it is possible to maintain positive fluid and electrolyte balance with increased enteral intake because of the absorptive reserve of distal small bowel. In contrast, loss of most of the ileum in addition to the colon, resulting in an end-jejunostomy, generally produces negative fluid balance that is driven by the secretory character of proximal small bowel. Attempted compensation with increased enteral fluid intake disproportionately increases fluid loss from the stoma, and in this situation, only intravenous fluid therapy can maintain a positive fluid balance.

Small Intestinal Length and Absorptive Function

The small intestine approximately doubles in length during the last trimester of a normal pregnancy, the mean being 100 cm at 27 weeks of gestation and 200 cm at 40 weeks of gestation. By age 3 years, mean small intestinal length is 350 cm, which approaches small bowel length in normal adults. Length of small bowel that remains after a resection is the single most important factor determining whether SBS shall develop and if so, whether PN dependence shall be permanent ( Figure 35-1 ); natural variation makes the length of bowel resected (versus length remaining) less important. Because of the marked increase in small bowel length that occurs late in gestation and because of the high frequency of premature births in infants with NEC and congenital malformations, remnant small bowel length is of maximal prognostic value for SBS when expressed as a fraction of that expected for gestational age. Thus, premature infants have a greater probability of ending PN compared to full-term infants with equivalent lengths of remnant small bowel.

Theoretical relationship between probability of autonomy from parenteral nutrition (PN) and residual measured small bowel length.

(From Figure 2 in Andorsky.)

In adults, PN is likely when total enteral macronutrient absorption by the remnant gastrointestinal tract is less than one-third of that ingested or about 84% of the basal metabolic rate. Medical and nutritional interventions short of PN in adults are typical when nutrient assimilation ranges between one-third and two-thirds of that ingested. Similarly, parenteral fluid therapy can be expected when net assimilation falls below 1.4 kg per day. The implication is that adult patients with an extreme intestinal loss must become profoundly hyperphagic, increasing caloric intake up to threefold to avoid PN. Comparable data have not been definitively established in children. However, given the metabolic demands of growth and development, children probably need to tolerate an enteral intake twofold greater than normal, that is, assimilate at least 50% of calories consumed, to avoid PN.

The Impact of Colon Anatomy

Normal colon length at birth is about 40 to 50 cm, and outcome of SBS is influenced by colon resection. Because the ileocecal valve (ICV) is rarely removed without some colon and ileum, prognostic significance of ICV loss per se is difficult to judge. In adult-onset SBS, and by implication child- and adolescent-onset SBS, a remnant small bowel that is shorter than 100 to 115 cm and ends as a stoma results in life-long PN dependence in most affected patients. In these older patients, preservation of at least 60 cm of proximal jejunum anastomosed to some length of colon (no ICV) yields an 85% to 90% probability of ending PN, whereas no more than half can end PN if remnant jejunal length is only 30 to 35 cm. In contrast, if the same length of remnant small bowel (30 to 35 cm) is in continuity with the ICV and entire colon, which implies at least some ileal preservation, then PN can usually be discontinued eventually. In infant-onset SBS, PN is usually permanent if remnant small bowel is less than 70 cm and ends with an abdominal wall stoma. Infants with 20 to 25 cm or less of small bowel often end PN, particularly if the remnant small bowel retains continuity with an intact colon including the ICV. Indefinite PN is rare when a remnant jejunoileum of at least 40 cm is in continuity with an ICV and intact colon. The combined impact of remnant small bowel length and presence or absence of an ICV is depicted in Figure 35-2 .

Interaction between remnant small bowel length and presence or absence of ileocecal valve (ICV) on duration of PN.

(From Figure 3 in Goulet.)

Function of Remnant Gut

Lengths of small intestine and colon remaining after surgical resection are not the sole determinants of PN permanency in SBS. In addition to gut anatomy, functional integrity of remaining bowel undoubtedly has an impact, since non-necrotic bowel retained after extensive resection may be sub-lethally but permanently damaged by the original event. Bowel function after resection is difficult to quantify directly in infants, and early tolerance of enteral nutrition (EN) can be a surrogate measure. Thus, in infants with 25 cm of small bowel remaining after neonatal resection, uncomplicated delivery of 75% of calories via the gastrointestinal tract by age 3 months has predicted a 90% probability of ending PN. Conversely, tolerance of only 25% of daily calories via the gastrointestinal tract with 25 cm of remnant small bowel at age 3 months predicts only a 50% chance of ending PN ( Figure 35-3 ). Most studies demonstrate that the etiology of SBS is not an independent predictor of outcome, although gastroschisis may have a poorer prognosis than other congenital malformations, including volvulus associated with malrotation and jejunoileal atresia. In particular, the 10% to 25% of infants with gastroschisis associated with atretic or stenotic segments often require extended PN due to dysmotility of remnant small bowel of seemingly adequate length.

Three plots derived from the Cox proportional hazards equation for three groups of patients with short-bowel syndrome. Patients who receive (A) 75%, (B) 50%, and (C) 25% of their daily calories by the enteral route at 3 months of adjusted age. The vertical axis shows expected percentage of patients who depend on PN at any age (months). Patients with three different residual intestinal lengths after initial surgery are shown (25 cm, 75 cm, and 120 cm residual intestine, respectively). By using the Cox proportional hazards equations, survival curves such as these can be generated for any combination of the two variables. Care must be taken in applying these plots to patients whose medical management differs significantly from that described for subjects of this study. Variation of the model will require a prospective evaluation of its accuracy in more patients with neonatal intestinal resection.

(From Figure 2 of Sondheimer.)

Intestinal Adaptation

Alimentary tract function gradually improves following massive intestinal resection, thereby diminishing impact of gut loss to a variable degree through compensatory events collectively referred to as “adaptation.” Central to the concept of adaptation is that global absorptive function of a segment of adapted remnant small bowel is greater than that of an equivalent length of bowel immediately after resection. Most information about adaptation is derived from adult laboratory animal models of massive intestinal resection. Applicability of this information to humans in general, and to small infants in particular, is uncertain. Apart from species-specific differences in physiology, animal models usually do not account for congenital or acquired abnormalities of remnant bowel that might undermine subsequent reparative processes. Nonetheless, animal data are a useful framework upon which SBS may be understood and physiologic patient care practices developed. Following massive small intestinal loss, generalized cellular hyperplasia , most notably affecting myocytes and enterocytes, ensues under the stimulus of enteral nutrients. Mucosal hypoplasia occurs in the absence of enteral feeding, although this phenomenon may not be nearly as pronounced in humans as in experimental animals. Complex nutrients are most effective in stimulating cellular hyperplasia, intact proteins more than free amino acids or peptides, and long-chain triglycerides more than medium-chain triglycerides. Increased numbers of enterocytes lengthen the villus-crypt unit, increase the total surface area available for absorption, and increase the total number of the various enzyme and transport molecules per crypt, although the absolute number of these molecules in individual enterocytes decreases. Expansion of enterocyte and myocyte mass presumably contributes to expected dilation of remnant bowel , further increasing total absorptive surface area. Recent experimental evidence supports the concept that adaptive potential of remnant ileum is greater than jejunum.

Although most research concerning adaptation has focused on responses by remnant small bowel to massive small bowel resection, recent experimental data indicate that remnant colon also demonstrates a hyperplastic response to partial colon resection that is a common clinical feature of infant SBS. Like adaptation of the small bowel, colon adaptation appears most intense under the stimulation of a polymeric diet, particularly a diet enriched with fish oil compared to one incorporating long-chain lipids obtained only from vegetable sources.

The time-frame over which hyperplasia occurs following resection is not precisely established in humans. One indication that enterocyte hyperplasia continues for months after resection is the rise in plasma citrulline concentration that frequently accompanies successful withdrawal of PN. Enterocytes are the sole source of plasma citrulline, the concentration of which varies in proportion to the length of small intestine remaining after resection. Either an initial postresection plasma citrulline concentration of about 12- to 15 µmol/L or a rise to this level or greater following initiation of EN appears to predict successful PN withdrawal. Clinical tolerance of enteral feeding improves for about 2 years after resection in adults, but for a longer period after resection in infancy of at least 3 to 4 years, as would be predicted by the natural linear growth of the intestinal tract during this period. Patients who continue to require PN after these intervals often need PN indefinitely.

In the clinical setting, apparent adaptation may continue over time without obvious change in small intestinal structure or function. Increasing efficiency of bacterial fermentation in the colon may be one explanation. Increases in colonic bacterial fermentation may be gradual and clinically indistinguishable from improving nutrient assimilation by the small intestine; this phenomenon may contribute to the reported lack of relationship between remnant colon length and time required to end PN. Evidence concerning systemic metabolic adjustments to massive intestinal resection is scant. Limited data in children suggest that there is no significant alteration in total energy expenditure in SBS. In contrast, adults may discontinue PN despite aggregate nutrient absorption well below the predicted basal metabolic rate, suggesting that metabolic compensation for severe malabsorption does occur following massive intestinal resection. In infancy and childhood, energy expenditures, expressed per kilogram of total bodyweight or lean body mass, fall with advancing chronologic age, which may also contribute to apparent adaptation, since the magnitude of increase in calories required to sustain growth gradually falls over time.

The Early Postoperative Phase

The period of postoperative ileus ordinarily lasts less than one week if ongoing abdominal sepsis or other complications do not occur. At this time, fecal output is low as upper gastrointestinal secretions are drained, usually with a nasogastric or concurrently placed gastrostomy tube. Recovery of motility permits discontinuation of upper gastrointestinal tract decompression following which fecal output may variably increase. The clinical challenge is appropriate replacement of ongoing losses to maintain fluid and electrolyte balance. Proper water and electrolyte replacement is most difficult when the small bowel ends as a proximal jejunostomy. Given the fundamentally secretory character of proximal small bowel, a high enterostomy is associated with an output that approximates 30 to 50 mL/kg per day even while the patient remains nil per os , in effect, a secretory diarrhea. Gastric hypersecretion may contribute to early, high fluid loss from an enterostomy. Histamine-2 receptor antagonists and proton pump inhibitors delivered intravenously have been advocated to reduce gastric hypersecretion during the first months after massive intestinal resection, although evidence that supports the efficacy of this practice in infants and children is lacking. Sodium and chloride concentrations in proximal jejunostomy effluent are relatively high, both up to 120 mEq/L, and these electrolytes must also be replaced in addition to fluid to prevent hyponatremia and hypochloremia. In contrast, potassium and bicarbonate losses are low. Progressively distal placement of an enterostomy reduces fluid loss as additional, predominantly absorptive small bowel modifies the fecal stream. Retention of a substantial length of colon generally precludes significant secretory fluid loss; however, potassium and bicarbonate requirements tend to climb as requirements for sodium and chloride generally fall.

PN should begin following establishment of stable fluid and electrolyte status. When remnant bowel anatomy predicts prolonged if not indefinite PN, a semi-permanent, cuffed catheter (Broviac, Hickman) can be placed at the time of original surgery or soon thereafter. Alternatively, parenteral feeding may be initiated via a peripherally inserted central catheter (“PICC line”). During the early period of PN in the hospital, it is usually useful to deliver replacement fluids separate from PN, since volume and composition of replacement fluid are likely to change more frequently than PN, especially after EN is initiated.

Initiation of Feeding

Criteria to start EN are resolution of postoperative ileus, cessation of upper gastrointestinal tract decompression, and achievement of stable metabolic, fluid, and electrolyte status with PN and ancillary fluid and electrolyte support. In all ages, proprietary liquid diets are usually used. Most attention concerning EN in pediatric SBS has focused on young infants; most SBS is diagnosed at this stage. Breast milk is used when available, because it appears to promote intestinal adaptation. As an alternative to breast milk that is commonly limited in availability, most favor the use of amino acid–based formula over those based on intact protein or short peptides (protein hydrolysates) because of an apparently reduced tendency to precipitate hypersensitivity reactions. In contrast, adult patients with SBS, including those with terminal enterostomies, do not benefit from elemental or peptide formulas in comparison with diets and formulas containing intact proteins, and it is logical to infer the same in older children and adolescents. Continuous intragastric infusion is standard, but infusion of formula directly into the proximal small bowel, usually via trans-pyloric feeding tube, may circumvent regurgitation associated with small bowel dilation that results from obstruction in utero or in chronic lung disease.

Impact of Intestinal Anatomy

A temporary end-enterostomy is commonly placed at the time of initial resection. It is generally highly desirable to re-establish continuity of remnant small bowel and colon at the earliest practical time, because total fluid and nutrient absorption usually improve afterward for the reasons noted earlier. Gut anatomy also influences formula selection, particularly carbohydrate and lipid composition ( Box 35-2 ). Extreme resections that require a permanent enterostomy are rare in infants and children, but patients who retain an enterostomy or whose remaining small bowel is in continuity with a short length of rectosigmoid colon do best with diets or formulas with relatively high, lipid content. The reason is that enterocyte lipid absorption, being nonsaturable, is a fixed fraction of the total ingested; the greater the amount ingested, the greater the amount absorbed. Because dietary lipids do not contribute substantially to osmolarity of enteric succus, lipid malabsorption, unlike carbohydrate malabsorption, does not markedly increase water loss in patients with an enterostomy. In these patients, substitution of some dietary long-chain triglycerides with medium-chain triglycerides offers no significant absorptive advantage and may actually be deleterious, because medium-chain triglycerides may stimulate adaptive enterocyte hyperplasia less than long-chain triglycerides. Furthermore, medium-chain triglycerides are only useful as an energy source and are not incorporated into structural lipids essential for tissue growth.

Patients with substantial colon length in continuity with small bowel benefit from a diet that includes medium-chain triglycerides, because medium-chain fatty acids, which are more water-soluble than long-chain fatty acids, can be assimilated by the colon if not absorbed by the small bowel. Consequently, medium-chain fatty acids may be less prone to stimulate colonic salt and water secretion than malabsorbed long-chain fatty acids. Because bacterial production of short-chain fatty acids occurs predominantly in the colon, patients with enterocolonic continuity benefit more from diets enriched with complex carbohydrates, including both starches and soluble fibers, than do patients with enterostomies.

Advancement of EN and Reduction in PN

The primary objective in managing the patient with SBS is prevention of intestinal failure, that is, elimination of PN and removal of the central venous catheter required to deliver it. In practice, PN is gradually curtailed, as enteral feeding, which initially consists predominantly of a liquid formula delivered by tube, is increased. Enteral feeding is increased based on two interrelated criteria: the ability to increase body mass appropriately as parenteral calories are withdrawn and the ability to maintain hydration as PN volume is curtailed.

Enteral feeds are generally initially increased by 0.5 to 1.0 cal/kg every 1 to 3 days; continued and preferably accelerated weight gain justifies intermittent reduction in parenteral calories. Decreases in PN are usually quantitatively less than concomitant increases in EN given the malabsorption inherent in SBS. Aggregate enteral absorption of nitrogen, carbohydrate, and lipid calories is not routinely measured directly, but fractional intestinal absorption probably varies between 33% and 67% of the total ingested depending on the length and region(s) of bowel remaining. Fractional enteric absorption can be estimated as the difference between estimated total energy expenditure and delivered PN calories divided by the quantity of calories actually infused into the gastrointestinal tract. Total energy expenditure should be equivalent to parenteral caloric intake before initiation of EN under the key assumption that PN has established appropriate body growth. However, a tendency to overestimate energy needs during the period of total or near-total PN has been described, resulting in disproportionately high weight gain relative to body length with attendant cardiovascular and metabolic stresses. In contrast, failure to consider the reduced efficiency of EN compared to PN can subsequently lead to reduced fat mass and, potentially, growth impairment. Malnutrition is especially likely during predictable periods of increased total energy expenditure, particularly adolescence, and may require increased PN or its temporary resumption to sustain growth. Continuing estimation of enteric absorption can guide future PN weaning, quantify adaptation of remnant bowel, and help to identify reductions in bowel function that suggest development of complications such as small intestinal bacterial overgrowth (see “ Complications of SBS and PN ”). A constant or preferably increasing percentage of enteral nutrient absorption over time suggests that adaptation is occurring as expected. Macronutrient absorption is often superior to that of fluid and electrolytes, so PN calories may be weaned by reducing the concentrations of macronutrients in PN rather than by reducing PN volume.

PN volume as well as parenteral calories can be reduced if stable hydration is maintained after increasing EN as demonstrated by physical examination, estimated stool output, and relevant laboratory data, including urea nitrogen, hematocrit, and albumin. Estimating hydration is often challenging, because subcutaneous fluid may be difficult to distinguish from body fat. Fecal volume may also be difficult to estimate, because liquid stools blend with urine in diapers of infants without stomas. Useful indicators of a moderate stool pattern include the finding of diapers that contain urine only several times a day, less than 6 to 8 stools daily, absence of diaper rash attributable to liquid feces and, in the case of infants with an enterostomy or colostomy, outputs less than 50 mL/kg per day. It is important to resist the tendency to increase PN volume to prevent or reverse dehydration in response to increased enteral feeding unless the increased enteral feeding confers the benefit of a reduced need for PN calories.

As PN is curtailed, it becomes increasingly necessary to deliver electrolytes, trace elements, and vitamins via the gastrointestinal tract that were formerly delivered intravenously; failure to do so based on routine monitoring results in deficiency states ( Box 35-3 ). Patients who retain a substantial length of colon often need supplemental potassium and bicarbonate. Those with predominantly small bowel resection may need extra calcium, magnesium, and zinc. When intravenous vitamin therapy ends, patients lacking all or most of the ileum are particularly likely to develop deficiencies of vitamin D and vitamin E due to bile acid malabsorption. The amount of remnant terminal ileum necessary to permit adequate vitamin B ₁₂ absorption may be as little as 15 cm. Blood levels of vitamin B ₁₂ should be checked every 3 to 6 months; however, an elevated plasma methylmalonic acid concentration may be more sensitive than serum vitamin B ₁₂ concentration in detecting early deficiency. Hypophosphatemia is unusual and usually indicates vitamin D deficiency, possible calcium deficiency, and secondary hyperparathyroidism. Despite the rarity of duodenal resection, iron supplementation is needed in one-third of patients, possibly because of chronic gastrointestinal blood losses.

Because ending of PN is the first nutritional priority for patients with SBS, use of continuous rather than bolus infusion of formula beyond the initial feeding period is traditional, since this strategy is most likely to achieve maximal retention of calories and fluid delivered into the gastrointestinal tract. The desirability of delivering EN continuously over the longest period possible must be weighed against the desirability of permitting some oral feeding for the purpose of preserving and promoting feeding skills. Maintaining feeding skills is especially important for those patients who have a high probability of ending PN, since eventual achievement of a normal to near-normal lifestyle includes eating by mouth, a skill that is quickly lost during infancy if all oral intake is forbidden. If fluid balance is tenuous, consumption of limited quantities of glucose-electrolyte solution in lieu of formula may minimize stool output while preserving the ability to swallow. When all or most of the colon is present, proprietary solutions containing 50 mEq/L of sodium (Pedialyte) are satisfactory, but when there is only a short segment of colon in continuity with proximal jejunum or there is no colon, customized solutions containing as much as 90 to 120 mEq/L of sodium are needed for optimal water absorption. Solids rich in complex carbohydrates, particularly those that are high in soluble fiber such as cereals, unsweetened fruits, and lean meats, are best tolerated in those retaining a significant length of colon. Patients with little or no colon may experience few if any problems following ingestion of items with high fat content. Given the high degree of individual variation, some trial and error with feeding is often necessary. The combination of formula infusion and oral solids while awake may produce smaller and more formed stools than formula infusion alone, which may be a function of a net increase in viscosity of the diet. Reducing the rate of infusion during sleep may be justified, because infrequent diaper changes during the night may promote skin breakdown.

An additional test of tolerance to EN is the ability to deliver PN intermittently, that is, to “cycle” PN. One goal of PN cycling is the periodic curtailment of reactive hyperinsulinemia that may contribute to liver dysfunction associated with PN. The second goal of PN cycling is to obtain a developmental benefit from the improved physical mobility allowed by disconnection from an intravenous infusion pump. Because no prospective data establish optimal duration of the “off” period, this interval is based largely on the ability to maintain stable hydration and metabolic status, especially blood glucose level. Typical periods of interruption range from 4 hours in young infants to 12 hours in older children and adolescents. As increasingly larger quantities of enteral feeding are tolerated, duration of the “off” period can also be increased. Timing of the PN infusion is based largely on caretaker convenience. Adults with SBS customarily receive PN at night, because freedom from intravenous infusion during the daytime usually outweighs the inconvenience of frequent nocturnal urination. Nighttime infusion may be less advantageous to incontinent children, since associated increases in urine production may further increase the risk of local skin breakdown. The ultimate test of successful PN withdrawal is the ability to sustain the growth trajectory established during the period of PN therapy. Patients are most likely to fail a trial of PN withdrawal when small intestinal anatomy is marginal, that is, the length is around 40 cm or less, with absent ICV and partial colon resection. Despite an initial failure, a later trial of PN cessation may be successful as metabolic requirements relative to body size decrease.

Home PN in SBS

Pediatric as well as adult patients with SBS fare better when receiving PN in the home than in the institutional setting. The benefit to patient life quality and family social structure is obvious. PN at home is relatively safe, and cost may be reduced by as much as 25% to 50% compared to the expense of continued hospitalization. Planning for discharge to home should begin soon after surgery, once patient survival appears to be a reasonable certainty. Teaching should emphasize the primary responsibility of parents and/or other family providers for all aspects of intravenous fluid delivery and routine central venous catheter care. After discharge, a multidisciplinary team that includes physician, nursing, dietary, social work, and pharmacy components directs care. Professional nursing care should be provided in the home only as necessary to facilitate a smooth transition from the hospital. Formal mechanisms for regular communication between family care providers and the team, and delivery of fluids, drugs, and other supplies to the home and ambulatory laboratory support are established. This organization permits weaning of PN and advancement of EN as rapidly as possible, as caregivers assess response to dietary interventions, evaluate hydration, and obtain scheduled and unscheduled laboratory testing, all in telephone, e-mail, and office consultation with the team.