Yield of upper endoscopy in dyspepsia; A multicenter database study

Upper endoscopy is used frequently for the investigation of patients with upper abdominal symptoms. Efficient utilization requires understanding of the potential yield in different clinical contexts. Five centers used a common "GI-Trac" database to document clinical facts before upper endoscopy procedures, and documented the findings. The data were shared without patient identifiers. Amongst 14,592 upper endoscopies documented, there were a total of 1,811 patients who underwent upper endoscopy for quot;dyspepsia/ abdominal pain", with no prior established upper gastrointestinal diagnosis. Endoscopy revealed a new diagnosis in no fewer than 78.5% of patients (including 1.5% tumors and 17.51 ulcers), with little variation between centers. To further investigate the effect of additional specific symptoms, patients were divided into three groups; group 1, dyspepsia/abdominal pain as sole symptoms; group 2, dyspepsia with "sinister" symptoms (anemia, bleeding, dysphagia, or weight loss); group 3, dyspepsia plus other symptoms (such as nausea, heartburn). The percentages in each group overall were 64%:12%:241 respectively. Final diagnoses were grouped into:tumors, ulcers, other upper GI diagnoses, no diagnosis. Table shows result: number of patients (percentages). Group 1 2 3 total 1151 227 433 tumors 14 (1.2) 7 (3) 7 (1.6) ulcers 178 (15.5) 50 (22) 90 (2.8) other GI diagnoses 684 (59.4) 133 (58.6) 258 (59.6) normal 275 (24) 37 (16) 78 (18) Although the yield of ulcers and tumors was significantly higher (p=.003) in group 2 patients with "sinister" symptoms, compared to those with dyspepsia alone (25% versus 16.7%), the difference was less impressive than anticipated. Conclusion: A common database enables multiple centers to share information, and to answer relevant practical questions.

Duke Scholars

Author Paul Simon Jowell Medicine, Gastroenterology