How should Helicobacter pylori infection be diagnosed in the elderly?

Purpose. - The real prevalence of Helicobacter pylori (H. pylori) infection is difficult to determine in the elderly because of the frequency of drug intake (antibiotics or anti-secretory drugs). The aim of this study was to evaluate the diagnostic performance of five tests in the elderly. Methods. - The study population consisted of consecutive patients undergoin g a routine endoscopy between August 1998 and December 1999. We evaluated t he diagnostic performance of four tests in all of the included patients: cu lture and histology of biopsy specimens, serology (ELISA) and urea breath t est (C-13-UBT). Detection of H. pylori antigens in stool samples (HpSA) was realized in a subgroup. Patients were considered H. pylori + when result f or culture was positive or when two tests were positive. Results. - One hundred and sixty-seven patients were included in this study (55 men, 112 women; mean age: 85.6 +/- 5.1 years). Only 38 (22.8%) patient s were H. pylori +. Test performances showed the following results: serolog y sensitivity: 90.9% (IC 95%: 75.6-98.1) versus 86.9% (IC 95%. 63.6-96.9) f or culture versus 77.8% (IC 95%: 60.8-89.9) for histology and 74.3% (IC 95% : 56.7-87.5) for C-13-UBT. Eighty-nine (53.3%) took antibiotics or anti-sec retory drugs, only C-13-UBT performances decreased significantly (sensitivi ty: 94.4% [72.7-99.8] versus 52.9% [27.8-77]; P < 10(-6)). When gastric or duodenal ulcer were endoscopically diagnosed in older patients, both histol ogy and C-13-UBT could nor improve the diagnosis of H. pylori infection. Hp SA was realized in 107 patients (sensitivity: 74.1%, specificity: 98.7%). W e showed no statistical difference between HpSA performances and drug intak e. Conclusion, - Diagnostic performances decreased in older patients especiall y because of drug intake. (C) 2001 Editions scientifiques et medicales Else vier SAS.