Primary conservative management of external small-bowel fistulas. Changing composition of fistula series?

A primary conservative approach was used in treatment of 42 patients with 45 external small-bowel fistulas. Closure of fistula was achieved in 29 patients but failed in 13 (10 died and 3 discharged with open fistula). There was no time-related improvement in closure rate or mortality, probably because of a general shift towards more serious cases. To evaluate this concept more specifically, groups of simple (19) or complicated (23) fistula were distinguished. The 23 patients with complicated fistula were more severely ill, had more sepsis, e.g. intra-abdominal abscess, and were more often treated in the intensive care unit (p less than 0.01) than the 19 with simple fistula. The respective mortality rates were 7/23 and 3/19. All six patients who died of sepsis had complicated fistula. Mortality in the simple fistula group was related to the primary disease. Especially in cases of complicated fistula, further improvements in management of septic complications should be aimed for, to permit success in delayed definitive surgery.

Reinfusion of secretions from high-output proximal stomas or fistulas.

Patients with proximal stomas or high fistulas and defunctionalized intestine who are receiving total parenteral nutrition (TPN) often develop hepatic enzyme abnormalities and hyperbilirubinemia. A technique was developed to collect intestinal secretions from proximal stoma and to reinfuse these secretions into the distal part of the intestine. This technique was applied in eight patients with a disrupted intestinal tract. A significant decrease (p less than 0.05) in elevated serum bilirubin, alkaline phosphatase and gamma-glutamyl transpeptidase levels was observed. Alanine aminotransferase and aspartate aminotransferase levels did not change significantly. The plasma sodium levels, slightly subnormal before reinfusion (131.0 +/- 4.6 millimolar per liter), despite enormous supplementation, normalized during reinfusion (137.0 +/- 4.0 millimolar per liter). TPN was continued during this infusion. This suggests that TPN by itself does not cause intrahepatic cholestasis. Neither could it be explained by an effect of secondary bile acids because these were most likely not produced as bile did not reach the distal defunctionalized intestine. Three possible mechanisms are suggested. Restoration of passage in the distal intestine may diminish bacterial overgrowth, endotoxin production and absorption. Enlargement of the bile acid pool may diminish the susceptibility of the liver to the deleterious effects of endotoxins. We advocate this reinfusion technique to overcome the metabolic disturbances occurring in those patients with high-output stomas or fistulas arising from the proximal parts of the small intestine.