Infectious implications in the porcine model of natural orifice transluminal endoscopic surgery (NOTES) with PEG-tube closure: a quantitative bacteriologic study.

BACKGROUND: Obtaining reliable closure of transvisceral defects currently limits natural orifice transluminal endoscopic surgery (NOTES). PEG tubes are potential means of managing NOTES gastrotomies.

OBJECTIVE: To determine the efficacy of PEG closure after NOTES.

DESIGN: An experimental animal study.

SETTING: A laboratory.

INTERVENTIONS: The pigs received gastric lavage with saline solution, chloramphenicol, or no lavage, and then underwent transgastric NOTES peritoneoscopy. Cultures were obtained by endoscopy during the surgery. A sterile foreign body was left in the peritoneal cavity. The gastrotomy was closed with a 20F PEG tube. The animals were observed for 14 days and underwent sterile laparotomy. Cultures of the foreign body and the peritoneal cavity were obtained.

MAIN OUTCOME MEASUREMENTS: Abscess rate, peritoneal quantitative cultures.

RESULTS: After 1 exclusion for anesthetic complications, 19 animals underwent NOTES; 18 (94.7%) survived the entire postoperative period. One animal died on postoperative day 2 after the PEG tube dislodged. At 14 days, 5 animals (27.8%) had intra-abdominal abscesses, 8 (44.4%) had positive peritoneal cultures, and 9 (50%) foreign bodies were contaminated on culture. Infectious complications were not altered by the type of gastric lavage or peritoneal bacterial inoculum introduced at the time of surgery.

LIMITATION: An animal model.

CONCLUSIONS: PEG closure of a NOTES gastrotomy is associated with subclinical intra-abdominal abscess formation and can result in death when the tube is dislodged during the early postoperative period. Preprocedural gastric lavage does not alter the intra-abdominal bacterial burden introduced at the time of surgery or subsequent infectious outcomes in the porcine model. These concerning findings necessitate additional studies to determine if porcine models are appropriate and applicable to human subjects in the NOTES setting.