HOME
Return to Table of Contents
EFFECTIVENESS OF ENDOSCOPIC AND NON-SURGICAL MANAGEMENT OF GASTROINTESTINAL PERFORATIONS CAUSED DURING ENDOSCOPIC PRODEDURES
WP Pais, G Hammoud, P Tiwari, C Bartalos, M Miller, A Diaz-Arias, D Antillon, ML Bechtold, JA Ibdah, MR Antillon
Division of Gastroenterology, University of Missouri, Columbia, MO
BACKGROUND: Gastrointestinal perforations are unplanned events during endoscopy. Perforations are more common with therapeutic interventions than diagnostic procedures; therefore, interventional endoscopists are more likely to encounter perforations. Generally, surgical intervention with sealing of the perforation is considered the gold standard. However, the leak-proof sealing by surgical approach carries significant morality and morbidity. Furthermore, certain surgically-managed colonic perforations may require colostomy. We evaluated the effectiveness of managing perforations by endoscopic and non-surgical therapy.
OBJECTIVE: Our purpose was to assess the safety and effectiveness of managing gastrointestinal perforation using various endoscopic clips, stents, and antibiotic therapy as opposed to surgery.
METHODS: Retrospective analysis from chart audits of patients who underwent therapeutic endoscopy at our large tertiary-care university hospital.
RESULTS: We reviewed the therapeutic endoscopies performed from June 2006 to September 2007. Six perforations were noted in which all perforations were recognized endoscopically. These included one esophageal, two duodenal, and three colonic perforations. The esophageal perforation occurred during the attempt at therapeutic endoscopic retrograde cholangiopancreatography (ERCP). The remainder were during ESD. Five out of six perforations were managed by endoscopic and medical therapy. One colonic perforation required surgery and colostomy. The esophageal perforation was managed by placement of a Boston-Scientific POLYFLEX esophageal stent. A chest tube was also required for the pneumothorax. The duodenal and colonic perforations were managed with endoscopic clips. The mean number of clips required to close the perforations (n=4) was 10.25 (4 to 18 clips). The mean hospital stay for endoscopically managed patients was 6 days (3 to 10 days). The hospital stay for the surgically managed patient was five days. At this time, no delayed adverse events have been reported in any of these six perforations.
CONCLUSIONS: Gastrointestinal perforations can be effectively managed with endoscopic techniques and medical management when recognized at the time of the endoscopic procedure. Significance and effectiveness of leak-proof sealing in perforations managed by endoscopic techniques and medical management is unknown and additional studies are needed.