Complications after Percutaneous Endoscopic Gastrostomy Placement in a Gastroenterology Fellowship Program

Abstract

Purpose: To assess the safety of percutaneous endoscopic gastrostomy (PEG) placement in a gastroenterology fellowship program. Methods: Design: Retrospective review of all PEG placements performed at a tertiary medical center from January 2005 through December 2005. Review of the electronic medical record was done to determine fellow involvement in case, indication, and follow up. Results: Retrospective chart review identified a total of 86 patients who underwent PEG placement. Two patients were excluded because of pre-existing PEG. A standard guide wire pull through technique was used for all procedures using 20 French gastrostomy tube following endoscopic localization of site. Endoscopic confirmation of gastrostomy was at the discretion of the endoscopist. Fellows assited in 75 PEG placements. The remaining 9 were completed by endoscopist alone (8) or with assistance by a surgical resident (1). Indication for PEG placement included CVA in 35 patients (41.6%), head and neck cancer in 28 patients (33.3%), progressive neurological disease in 11 patients (13.1%), respiratory failure-ventilatory dependent in 3 patients (3.6%), and miscellaneous causes in 7 patients (8.3%). The complications observed included 4 bacterial infections (4.8%), 2 yeast infection (2.3%), 2 cases of excessive bumper tightness (2.3%), 1 peristent gastrostomy leakage (1.1%), and 1 buried bumper requiring endoscopic removal (1.1%). There was 1 case of anterior abdominal wall separation from anterior gastric wall requiring surgical repair (1.1%). 1 post PEG gastrointestinal bleed occured related to gastric ulcer successfully treated endoscopically (1.1%). There were no deaths related to PEG placement. However, 24 of the 84 patients (28.6%) were deceased at the time of the retrospective study review. The number of gastrostomy tubes placed by each fellow was not an accurate reflection of volume because the study reviewed PEG placement over a 12 month cycle which did not correspond with the academic training year. For this reason the complication rate determined by year of fellowship training was not calculated. Conclusions: To date no good published data exists specifically looking at complication rates for PEG placement within a gastroenterololgy fellowship training program. Our data suggest that overall the incidence of PEG related complications in this setting is relatively low. The most commonly observed complication was site infection requiring oral antibiotic therapy.