Diagnostic yield and clinical implications of push enteroscopy: results from a nonspecialized center.

BACKGROUND Push enteroscopy is increasingly used as an investigative tool for the evaluation of gastrointestinal bleeding, and studies from specialized centers have shown an overall diagnostic yield of push enteroscopy in such patients ranging from 38% to 75%. The aim of our study was to characterize the yield and clinical effect of push enteroscopy to determine the applicability of prior observations to other academic centers. STUDY We retrospectively studied patients who underwent push enteroscopy between January 1995 and December 2000 at our institution. Detailed clinical history, endoscopic findings, endoscopic therapy, and subsequent medical treatment were obtained through review of medical records and our endoscopic database. Medications prescribed after enteroscopy and whether medical management was affected by the findings of push enteroscopy were also recorded. RESULTS Over the 6-year study period, 126 patients (48% men; mean age, 62 years; range, 15-91 years) underwent push enteroscopy. The most common indications for push enteroscopy were gastrointestinal bleeding in 57 patients (45%) and iron-deficiency anemia in 32 (25%). The results of push enteroscopy were normal in 44 patients (35%), and the most frequent endoscopic lesions were angiectasias in 24 patients (19%), gastric erosions in 10 (8%), gastric ulcer in four (3%), jejunal ulcer in three (2%), and esophagitis in three (2%). The identified lesions (n = 89) were within reach of a standard upper endoscope in 42 patients (47%). Endoscopic therapy was performed in 12 patients (13%), and the management of 50 patients (40%) was changed based on findings at push enteroscopy. CONCLUSIONS Push enteroscopy has a high diagnostic yield, similar to reports from specialized centers suggesting the potential clinical benefit of more widespread use.