Rate and yield of repeat upper endoscopy in patients with dyspepsia

To determine the rate and yield of repeat esophagogastroduodenoscopy (EGD) for dyspepsia in clinical practice, whether second opinions drive its use, and whether it is performed at the expense of colorectal cancer screening.

We performed a retrospective cohort study of all patients who underwent repeat EGD for dyspepsia from 1996 to 2006 at the University of California, San Francisco endoscopy service.

Of 24,780 EGDs, 5460 (22%) were performed for dyspepsia in 4873 patients. Of these, 451 patients (9.3%) underwent repeat EGD for dyspepsia at a median 1.7 (interquartile range, 0.8-3.1) years after initial EGD. Significant findings possibly related to dyspepsia were more likely at initial (29%) vs repeat EGD (18%) [odds ratio (OR), 1.45; 95% confidence interval (CI): 1.20-1.75, P < 0.0001], and at repeat EGD if the initial EGD had reported such findings (26%) than if it had not (14%) (OR, 1.32; 95% CI: 1.08-1.62, P = 0.0015). The same endoscopist performed the repeat and initial EGD in 77% of cases. Of patients aged 50 years or older, 286/311 (92%) underwent lower endoscopy.

Repeat EGD for dyspepsia occurred at a low but substantial rate, with lower yield than initial EGD. Optimizing endoscopy use remains a public health priority.