Does the urea breath test tell us what we need to know?

[The authors respond:]

The point Julius Wroblewski has brought up is a good one. We would certainly like to detect lesions early, rather than at a point when they are no longer treatable. However, we must recognize that a definitive diagnosis is not necessarily required for adequate treatment and that serious diagnoses are rare in patients who present with gastrointestinal symptoms. We would also like to point out that the context for using the urea breath test in the "test and treat" approach for adult patients with dyspepsia is primary care.

Dyspepsia is extremely common, affecting 7% of patients presenting to a general practitioner's office, and it occurs with moderate severity in approximately 29% of Canadians.[1, 2] It is obviously not feasible nor necessary for close to 30% of the Canadian population to undergo endoscopy. We should try to perform this procedure in the patients who would most benefit. In fact, if all patients with dyspepsia who present to a primary care physician were to have a gastroscopic examination the waiting list for this procedure would become enormous, potentially resulting in a delay in diagnosis for those patients with symptoms suggesting more significant pathology. Hence, we have to find ways to determine which patients may have significant pathology. Alarm features (vomiting, bleeding, anemia, abdominal mass, dysphagia and weight loss) and advanced age suggest a higher risk of pathology. Performing endoscopy on these individuals and simply performing a test for and treating Helicobacter pylori infection in those that do not have these risk factors may reduce the waiting lists for endoscopy and hence potentially increase the detection of early lesions.

Furthermore, once the urea breath test becomes more widely available, testing for and treating H. pylori infection would not result in a significant delay for further investigation if the patient were not to respond to treatment. A urea breath test result can be faxed within 24-48 hours and the course of treatment is only 1 week. In addition, a Canadian randomized controlled trial recently showed significant improvement in symptoms with the "test and treat" approach compared with placebo3 and another study found this strategy to be significantly more cost-effective, without detrimental outcome, than a strategy using endoscopy first.4