to the editor: I agreed with the authors' recommendations for antibiotic selection and duration in the Helicobacter pylori article.¹ I also strongly agree that treatment should be individualized, and that patients with risk factors for malignancy or complicated disease should probably undergo early endoscopy.

However, I feel that the authors fail to appreciate the value of H. pylori serology in the evaluation of the outpatient with dyspepsia. They state, “Because of the high prevalence of H. pylori infection in persons over the age of 50 and the absence of ulcer disease in most of these persons, patients over the age of 50 should have documented ulcer disease before anti-Helicobacter therapy is initiated.” While this statement is true for persons in the general population, the population that should be considered in this context is patients who present to their family physician with dyspepsia.

The table illustrates the relationship between ulcer and H. pylori infection in patients with dyspepsia (based on previous studies) and assumes a 45 percent prevalence of H. pylori infection in patients who are dyspeptic.²

The table shows that a positive H. pylori serology is a reasonably good test for ulcer: 44 percent of H. pylori–positive patients have ulcer, compared with about 4 percent of those with negative serology. The 45 percent overall rate of H. pylori–positive serology among dyspeptic patients is typical of an older population; younger groups (as the authors note) have a lower prevalence, making the test even better for distinguishing ulcer from nonulcer dyspepsia. Studies have confirmed this finding; one found that of 120 H. pylori–negative dyspeptic patients with no recent history of non-steroidal anti-inflammatory drug (NSAID) use, none had an ulcer on endoscopy.³ Thus, serology is a useful tool for identifying patients with little risk of ulcer (and probably little risk of malignancy as well) who do not require immediate endoscopy.

in reply: The diagnostic challenge in dyspepsia is that a wide array of pathophysiology in the upper digestive tract can present with similar symptoms. Potential etiologies for dyspepsia include gastroesophageal reflux disease, peptic ulcer disease, malignancy, pancreaticobiliary disorders, and functional etiologies such as irritable bowel syndrome or nonulcer dyspepsia. To date, H. pylori has been shown to bear a significant relationship to peptic ulcer disease, gastric carcinomas and lymphomas. Given the high prevalence of H. pylori infection in older individuals, the utility of serology in identifying patients with significant organic disease is markedly diminished in patients over the age of 50. We would disagree that H. pylori positivity is a “reasonably good test” for ulcer. In the table noted by Dr. Ebell, only 44 percent of H. pylori–positive patients actually have an ulcer; 56 percent of positives do not. This makes a coin toss slightly more predictive than serology.

We agree that the absence of infection with H. pylori is strongly predictive of the absence of ulcer disease, especially when the use of NSAIDs has been rigorously excluded. Whether malignancy can be reliably excluded based simply on negative serology is controversial. The Fraser study¹ noted above found only one patient with gastric cancer and did not identify any patients with lymphoma, even among the H. pylori–positive subset. No conclusions can be drawn regarding risk of malignancy in seronegative subjects using such small sample sizes. Serology is of no benefit in identifying dyspeptic patients with symptoms based on gastroesophageal reflux. Thirty-six percent of seronegative subjects in the Fraser study had endoscopic evidence of esophagitis.

Finally, prompt endoscopy has been shown to be cost-effective in the evaluation of the patient with dyspepsia. In a study randomizing 414 patients with dyspepsia to either prompt endoscopy or empiric medical therapy (with later endoscopy only for medical failures), the prompt endoscopy group had significantly lower overall costs, primarily because of lower drug costs, fewer office visits and less time off from work.²