Current issues of ACP Journal Club are published in Annals of Internal Medicine


Eradication of Helicobacter pylori with omeprazole and antibiotics improved symptoms in nonulcer dyspepsia

ACP J Club. 1999 May-June;130:65. doi:10.7326/ACPJC-1999-130-3-065

Related Content in this Issue
• Companion Abstract and Commentary: Eradication of Helicobacter pylori did not relieve symptoms of nonulcer dyspepsia but increased healing of gastritis

Source Citation

McColl K, Murray L, El-Omar E, et al. Symptomatic benefit from eradicating Helicobacter pylori infection in patients with nonulcer dyspepsia. N Engl J Med. 1998 Dec 24;339:1869-74.



In patients with nonulcer (or functional) dyspepsia, does eradication of Helicobacter pylori relieve symptoms?


Randomized, double-blind, placebo-controlled trial with 1-year follow-up.


A dyspepsia clinic in Glasgow, Scotland, United Kingdom.


318 patients (mean age 42 y, 51% women) referred by primary care physicians for dyspepsia (pain or discomfort in the upper abdomen or lower chest, heartburn, nausea, feeling of postprandial fullness, or symptoms related to the upper gastrointestinal tract) that persisted for ≥ 4 months and who had H. pylori infection with no endoscopic evidence of peptic ulcer disease. Exclusion criteria were peptic ulcer disease, endoscopic evidence of esophagitis, use of nonsteroidal anti-inflammatory drugs, previous gastric resection, pregnancy, or previous treatment for H. pylori infection. Follow-up was 97%.


Patients were allocated to omeprazole and antibiotics (n = 160) or omeprazole and placebo (n = 158) for 2 weeks. Omeprazole, 20 mg, was given twice daily. The antibiotic regimen was amoxicillin, 500 mg 3 times daily, and metronidazole, 400 mg 3 times daily. Patients who were allergic to penicillin received tetracycline, 500 mg 3 times daily, instead of amoxicillin.

Main outcome measure

Relief of symptoms (score of 0 or 1 on the Glasgow Dyspepsia Severity Score). Secondary outcome measures were quality of life, which was assessed by the Medical Outcomes Study Short Form-36 Health Survey (SF-36), and H. pylori infection status.

Main results

Analysis was by intention to treat. More patients in the antibiotic group than in the placebo group had symptom relief after 1 year (P < 0.001) (Table), but average dyspepsia scores at 1 year were similar (mean score 5.4 vs 6.2, P = 0.07). Treatment groups did not differ for quality-of-life scores. More patients in the antibiotic group than in the placebo group tested negative for H. pylori at 1 year (85% vs 12%, P < 0.001).


In patients with nonulcer dyspepsia, eradication of Helicobacter pylori with omeprazole plus antibiotics relieved symptoms in more patients than did omeprazole alone.

Source of funding: U.K. Medical Research Council.

For correspondence: Dr. K. McColl, Department of Medicine and Therapeutics, Western Infirmary, Glasgow G11 6NT, Scotland, UK. FAX 44-141-339-2800.

Table. Omeprazole and antibiotics vs omeprazole and placebo in nonulcer dyspepsia*

Outcome at 1 y Antibiotics Placebo RBI (95% CI) NNT (CI)
Treatment success 21% 7% 200% (60 to 469) 7 (5 to 15)

*Abbreviations defined in Glossary; RBI, NNT, and CI calculated from data in article.


When considering any study that uses symptom-based diagnoses and outcomes, one must note that symptoms relating to the gastrointestinal tract, even in recognized organic disease, are heterogeneous, poorly defined, and difficult to quantify.

Most studies of functional disorders have used hospital patients, who may have high psychological morbidity and tend to exhibit a high short-term placebo response. Peptic ulceration in the absence of nonsteroidal anti-inflammatory drugs is usually associated with H. pylori infection, and eradication is an effective treatment. The prevalence of H. pylori in NUD has not been convincingly shown to differ from that in asymptomatic controls. Although gastritis is clearly present in infected persons, it does not have a consistent effect on gastric motility (1) or sensitivity (2). If H. pylori has a role in NUD, it is inconsistent and can only be implicated in a few patients. Before these publications, similar studies in NUD had been criticized for being too small, having insufficient follow-up, and using insensitive outcome measures (3). However, the studies by McColl and Blum and their colleagues were large, had 1-year follow-up, and used previously validated scoring systems to assess outcome.

Both groups studied only hospital patients, but McColl and colleagues included most patients who presented with symptoms in the upper gastrointestinal tract. Using the relatively strict criteria of complete resolution of symptoms, they found a statistically significant benefit for eradication therapy. Despite this, treatment did not decrease the average symptom score more than placebo did. This may have been because the study was not powerful enough, despite its size, to show a substantial difference, or because only a subgroup responded to treatment. In particular, the study group may have been enriched with patients with ulcers that had been “missed.” There was a high prevalence of peptic ulceration in symptomatic patients in the placebo group who subsequently had repeated endoscopy (4 of 6 patients); this would tend to exaggerate the benefits of eradication.

Blum and colleagues had more restrictive selection criteria that included only patients with upper abdominal pain and discomfort. Despite this restriction, they showed no statistically significant benefit compared with placebo when either the strict criteria of resolution of symptoms or average symptom scores were used. They were more rigorous than McColl and colleagues in excluding peptic ulceration (which was found in 3% of patients) by performing repeated endoscopy in all patients at 3 and 12 months.

The explanation of the difference in outcome of the 2 studies was the poor placebo response (7%) in McColl and colleagues' study. This was uncharacteristically low for studies of functional disease, but it may be explained by the long follow-up or the more exacting definition of treatment success. Alternatively, blinding may have been compromised by the use of metronidazole, which frequently causes side effects, especially when taken with alcohol.

Assuming that inadequate blinding was not an issue, McColl and colleagues' study supports the hypothesis that some patients with NUD achieve long-term benefit from H. pylori eradication, although these patients may actually be “ulcer” patients with no, minimal, or missed findings at initial endoscopy. Both studies show that most patients do not benefit, and indeed, some may deteriorate. The findings of these well-designed studies do not support routine eradication therapy for NUD but suggest that a subgroup of patients (who are difficult to identify) may benefit in the long-term.

J. I.W. Jones, MD
C. J. Hawkey, MD
University of NottinghamNottingham, England, UK

J. I.W. Jones, MD
University of Nottingham
Nottingham, England, UK

C. J. Hawkey, MD
University of Nottingham
Nottingham, England, UK


1. Goh KL, Paramsothy M, Azian M, et al. Does Helicobacter pylori infection affect gastric emptying in patients with functional dyspepsia? J Gastroenterol Hepatol. 1997; 12:790-4.

2. Mearin F, de Ribot X, Balboa A, et al. Does Helicobacter pylori infection increase gastric sensitivity in functional dyspepsia? Gut. 1995;37:47-51.

3. Veldhuyzen van Zanten SJ, Cleary C, Talley NJ, et al. Drug treatment of functional dyspepsia: a systematic analysis of trial methodology with recommendations for design of future trials. Am J Gastroenterol. 1996;91:660-73.