RESUMO
BACKGROUND AND AIMS: The originality of this retrospective study relies on the evaluation of the effectiveness of the endoscopic internal drainage (EID) according to the type of fistula. METHODS: The type of fistula was classified initially according to a CT scan with oral opacification: fistula without a communicating abscess (type I), fistula with a communicating abscess (type II), and fistula with an abscessed sub- and sus-diaphragmatic communicating collection (type III). Treatment algorithm consisted of the insertion of a nasojejunal feeding tube (NJFT) for type I fistulas and the placement of a NJFT with EID with or without surgical drainage for types II and III. RESULTS: Forty-nine patients were included. The clinical success rate with fistula healing was 100% in group I, 96% in group II, and 12% for group III (p = 0.001). Mean time for diagnosis of the fistula was significantly higher in type III (p = 0.04). The mean estimated size of the defect was higher in type II, 11.2 mm and III, 10 mm versus type I, 2.8 mm (p = 0.001). The average number of scheduled endoscopic sessions were 2, 2.7, and 5.2 for types I, II, and III, respectively (p = 0.001). The number of unscheduled reinterventions was also significantly higher in type III (p = 0.03). The NJFT was left in place for a significantly longer duration in type III (136 days) compared to types I (3, 13) and II (49) p = 0.001. CONCLUSION: This study shows that proper characterization of the type of fistula before the endoscopic treatment of post-sleeve fistulas improves the efficacy of the endoscopic treatment.