A comparison of upper gastrointestinal endoscopy and radiography

One hundred symptomatic patients were evaluated independently with upper gastrointestinal radiography and fiberoptic endoscopy, and the results were compared. Of the two endoscopists sequentially examining the same patient, one was informed of available clinical and radiographic details and the other was not. Knowledge of the x-ray examination by the informed endoscopist did not improve his accuracy. Each endoscopist made four errors of interpretation. The endoscopic and x-ray findings agreed in 46 of the 100 patients, most often (68%) in esophagus, least often (29%) in the stomach and half the time (45%) in the duodenum. Ulcer craters seen endoscopically were detected radiographically in 36% of patients. We conclude that: 1) knowledge of results of prior upper gastrointestinal radiography did not alter endoscopic results; 2) experienced endoscopists are accurate but make mistakes; and 3) endoscopic findings would have been unaltered had radiography not been performed.