Is there a case for routine duodenal biopsy at gastroscopy or is a targeted approach adequate?

Abstract

Introduction: United Kingdom recommendations advocate duodenal biopsy at the time of gastroscopy if there is evidence of certain indications. Replacing this with routine duodenal biopsy has been suggested, as low duodenal biopsy rates have been reported even when these indications are present. We aimed to compare the diagnostic yield of routine duodenal biopsy against a policy of targeted duodenal biopsy using the indications of anemia, weight loss, or diarrhea. Methods: We retrospectively examined our endoscopy database from January 2003 to January 2004. The numbers of gastroscopies and duodenal biopsies performed were noted (retrospective group, group 1). Subsequently, adult patients referred for gastroscopy were prospectively recruited from January 2004 to April 2006 (prospective group, group 2). In this group, routine duodenal biopsy was performed and the subgroup of patients with target indications was identified. We noted the number of new diagnoses made as a result of duodenal biopsy in all groups. Results: In group 1, 5979 gastroscopies were performed, and duodenal biopsy was performed in 1464 (24.5%). In group 2, 2133 gastroscopies were performed and 739 patients had target indications. A significantly higher prevalence of giardiasis and celiac disease was shown in group 2 (P<0.01). Comparing all the patients within group 2 to just those within group 2 who had target duodenal biopsy indications, no significant difference was seen in the prevalence of celiac disease (3.9% versus 3.6%). However, there would have been a significant 63% reduction in duodenal biopsy rate in group 2 if the target indication approach had been adopted (P <0.0001). Conclusion: The diagnostic yield for celiac disease and giardiasis is significantly increased by performing routine duodenal biopsy. However, by accurately targeting patients, the number of duodenal biopsies performed can be significantly reduced without compromising the detection rate of small-bowel disease.