BACKGROUND: Routine duodenal biopsies during upper gastrointestinal endoscopy (UGE) have been suggested to be useful in detecting coeliac disease (CD). However results from previous studies are not conclusive. The aim of this study is to investigate the diagnostic yield and cost-effectiveness of routine duodenal biopsy during UGE. METHODS: In this retrospective single-centre study, we studied 6442 patients undergoing first-time UGE at the Rijnstate Hospital, Arnhem, the Netherlands, from January 2009 to December 2010. All UGE reports were analysed for indication, duodenal intubation, and endoscopic aspect of duodenal mucosa. Endomysium and tissue transglutaminase antibody titre, when present, were scored as positive or negative. CD was defined as Marsh 3a or higher. Costs of duodenal biopsies and pathology analysis were calculated. Comparisons were done with T-tests for continuous data and Chi-square tests for categorical data. RESULTS: Forty-one patients had newly diagnosed CD; 34 of these 41 patients had definite indications for biopsy prior to UGE, e.g. positive serology or symptoms. Thus, routine duodenal biopsies identified seven patients as having CD, who otherwise would not have been biopsied. The number needed to biopsy was therefore 577, spending more than v 30,000 per case. CONCLUSIONS: We do not recommend routine duodenal biopsy to screen for coeliac disease because of the high number needed to biopsy as well as high costs.
BACKGROUND: Routine duodenal biopsies during upper gastrointestinal endoscopy (UGE) have been suggested to be useful in detecting coeliac disease (CD). However results from previous studies are not conclusive. The aim of this study is to investigate the diagnostic yield and cost-effectiveness of routine duodenal biopsy during UGE. METHODS: In this retrospective single-centre study, we studied 6442 patients undergoing first-time UGE at the Rijnstate Hospital, Arnhem, the Netherlands, from January 2009 to December 2010. All UGE reports were analysed for indication, duodenal intubation, and endoscopic aspect of duodenal mucosa. Endomysium and tissue transglutaminase antibody titre, when present, were scored as positive or negative. CD was defined as Marsh 3a or higher. Costs of duodenal biopsies and pathology analysis were calculated. Comparisons were done with T-tests for continuous data and Chi-square tests for categorical data. RESULTS: Forty-one patients had newly diagnosed CD; 34 of these 41 patients had definite indications for biopsy prior to UGE, e.g. positive serology or symptoms. Thus, routine duodenal biopsies identified seven patients as having CD, who otherwise would not have been biopsied. The number needed to biopsy was therefore 577, spending more than v 30,000 per case. CONCLUSIONS: We do not recommend routine duodenal biopsy to screen for coeliac disease because of the high number needed to biopsy as well as high costs.