A study of the safety and clinical efficacy of esophagogastroduodenoscopy after esophageal, gastric, or duodenal surgery in 60 patients.

Objectives : Our purpose was to evaluate the safety of esophagogastroduodenoscopy (EGD) after esophageal, gastric, or duodenal surgery, a previously unstudied subject. EGD could promote suture breakdown at sites of anastomoses, ostia, or repair by abrasion or stretch from endoscopic intubation, torque, or insufflation. Methods : Risks versus benefits of EGD performed within 24 days of upper gastrointestinal surgery were analyzed at five medical centers in 60 study patients and were compared with 120 controls matched for age, sex, and EGD indication. Results : Of 13,170 hospitalized patients undergoing EGD, 60 patients (0.46%) had EGD within 24 days of upper gastrointestinal surgery. EGD indications in study patients included gross upper gastrointestinal bleeding in 24, nausea and vomiting in 23, bright red blood per rectum in four, dysphagia in three, abdominal pain in three, and other in three. EGD was performed on average 14.7 ± 7.2 (SD) days after surgery. EGD provided the diagnosis in 45 (75%) study patients (control rate = 77%, not significantly different, X 2 ). In particular, postoperative EGD provided the diagnosis in 20 (83%) of 24 patients with upper gastrointestinal bleeding (control rate = 92%, not significantly different, Fisher's exact test). Thirteen study patients (22%) had upper gastrointestinal lesions directly related to surgery including : anastomotic ulcers in six, anastomotic erosions in six, and an excessively tight Nissen fundoplication in one. One additional patient at EGD had extrinsic duodenal compression attributed to postoperative adhesions. EGD led to surgery in nine and to successful endoscopic therapy in six. No study patient experienced an endoscopic complication. The control group had one endoscopic complication (not significantly different rates, Fisher's exact test). Conclusion : In this study, the benefits outweighed the risks of postoperative EGD when performed more than 7 days after surgery : the diagnostic yield was high and the complication rate was low. Indeed, the yield and risks were similar to that of other EGDs. Postoperative EGD identified a surgical complication in about 20% of cases. Clinicians should use discretion and perform postoperative EGD for clinically important indications. EGD is contraindicated when wound dehiscence or bowel perforation is suspected.