Background: We studied the etiology, treatment, and outcome of enterocutaneous fistulas in 106 patients to evaluate our current practice and the impact of newer therapies-octreotide, wound vacuum-assisted closure (VAC), and fibrin glue-on clinical outcomes. Review of the literature and our own 1990 study indicate a mortality rate of 5% to 20% for enterocutaneous fistula, and a healing rate of 75% to 85% after definitive surgery.
Methods: We reviewed all cases of gastrointestinal-cutaneous fistula from 1997 to 2005 at 2 large teaching hospitals. We identified 106 patients with enterocutaneous fistula; patients with irritable bowel disease and anorectal fistulas were excluded.
Results: The origin of the fistula was the small bowel in 67 patients, colon in 26, stomach in 8, and duodenum in 5. The etiology of the fistula was previous operation in 81 patients, trauma in 15, hernia mesh erosion in 6, diverticulitis in 2, and radiation in 2. Of the 106 patients in the study, 31 had a high output fistula (greater than 200 mL/day), 44 had a low output fistula, and, in 31 patients, the fistula output was low but there was no record of volume. Initial treatment was nonoperative except for patients with an abscess who needed urgent drainage. In 24 patients, the effect of octreotide was monitored: in 8 patients, fistula output declined; in 16 patients, octreotide was of no benefit. Fibrin glue was used in 8 patients and was of benefit to 1. The wound VAC was used in 13 patients: 12 patients still required operative repair of the fistula, whereas the fistula was healed in 1 patient. The main benefit of the VAC system was improved wound care in all patients before definitive surgery. Total parenteral nutrition was used in most patients to provide nutritional support. Operative repair was performed in 77 patients and was successful in 69 (89%), failing in 6 patients with persistent cancer or infection. Nonoperative treatment was used in 29 patients and resulted in healing in 60%. Of 106 patients, 7 (7%) died of fistula complications. The cause of death was persistence or recurrence of cancer in 4 patients and persistent sepsis in 3.
Conclusion: Enterocutaneous fistula continues to be a serious surgical problem. The wound VAC and fibrin glue had anecdotal successes (n = 2), and one-third of patients responded to octreotide. We believe that octreotide should be tried in most patients and that the wound VAC has a role in selected patients. Although 7% overall mortality is lower than in previous studies, the number managed without operation (27%) remains the same. In addition to early control of sepsis, nutritional support, and wound care, a well-timed operation was the most effective treatment.