Article Text
PDF
Statistics from Altmetric.com
QUESTION: In patients with dyspepsia and a positive test result for Helicobacter pylori, is an H pylori eradication strategy more effective than placebo for improving dyspepsia symptoms?
Design
Randomised {allocation concealed*}†, blinded (clinicians, patients, data collectors, outcome assessors, {data analysts, and manuscript writers}†),* placebo controlled trial with 1 year of follow up.
Setting
36 family practices in Canada.
Patients
294 patients (mean age 49 y, 50% men) who were ≥18 years of age and had uninvestigated symptoms of dyspepsia for ≥3 months. Dyspepsia was defined as a complex of epigastric pain including heartburn, acid regurgitation, excessive burping or belching, increased abdominal bloating, nausea, abnormal or slow digestion, or early satiety. All patients had to have positive test results for H pylori on the Helisal rapid blood test and the 13C-urea breath test. Exclusion criteria included gastrooesophageal reflux disease, upper gastrointestinal investigation in the previous 6 months or ≥2 times in the past 10 years, eradication therapy for H pylori in the past 6 months, previous gastric surgery, ulcer disease or endoscopic oesophagitis, and irritable bowel syndrome. Follow up was 87%.
Intervention
Patients were allocated to omeprazole, 20 mg; metronidazole, 500 mg; and clarithromycin, 250 mg (eradication) (n=145) or omeprazole, 20 mg, and placebo metronidazole and placebo clarithromycin (placebo) (n=149) twice daily for 7 days.
Main outcome measures
Global overall severity (GOS) of dyspepsia symptoms assessed with a 7 point scale (1=no problem, 7=very severe problems). Treatment success was a score of 1 or 2. Secondary outcomes were proportion of asymptomatic patients and treatment success according to H pylori status.
Main results
Analysis was by intention to treat, and an analysis with all evaluable patients was also done (n=267). Patients in the eradication group had greater treatment response than did those in the placebo group (table). More patients in the eradication group were completely asymptomatic (table). Treatment was more successful in patients in whom H pylori was eradicated than in those it was not (54% v 39%, p=0.02). Eradication treatment reduced societal costs, but the difference was not statistically significant (Cdn $53, 95% CI −86 to 180).
H pylori eradication v placebo for uninvestigated dyspepsia at 1 year‡
Conclusion
In patients with dyspepsia and a positive test result for Helicobacter pylori, an H pylori eradication strategy was more effective than placebo for improving dyspepsia symptoms.
Commentary
Patients with dyspeptic symptoms should be managed with 2 goals in mind: early detection of malignant disease and cost-effective relief of symptoms. Patients <55 years of age without “alarm features” (eg, weight loss, dysphagia, and anaemia) are at very low risk of malignancy and do not require endoscopic investigation. The question of management then turns on the relative costs and effectiveness of endoscopy, noninvasive tests, and eradication of H pylori and empiric acid suppression.1
The trials by McColl and Chiba and their colleagues provide important information for physicians managing patients presenting with uninvestigated dyspepsia. The study by McColl et al adds to 2 previous studies2–3 that confirm the cost effectiveness of a secondary care based H pylori test and treat service compared with endoscopy based management. All 3 trials of test and treat compared with endoscopy based management showed equivalent effectiveness, but costs were reduced because fewer patients had endoscopy. The trial by McColl et al showed a rate of endoscopy in the year of follow up in the test and treat group of only 8%, whereas Heaney2 and Lassen3 showed rates nearer 30%. Patients positive for H pylori in both groups of the trial by McColl et al received eradication treatment. Thus, any differences caused by the eradication treatment itself were abolished. The trial can therefore be considered to address the question, “Is the cost of endoscopy warranted by the effect on symptoms, quality of life, and patient satisfaction of having the investigation?” The answer is “no”.
The trial by Chiba et al (CADET-Hp) takes the McColl et al trial 2 steps further. Firstly, patient recruitment and the intervention took place in a primary care setting. Secondly, test and treat was compared with acid suppression alone. In contrast to the trial by McColl et al, CADET-Hp was designed to examine the effect of eradication treatment on dyspeptic symptoms, and found a substantial improvement in the proportion of patients with dyspeptic symptoms at the end of the trial. However, the difference in costs was small and not statistically significant.
The reason for the difference in effects and costs between CADET-Hp and McColl et al lies in the use of eradication treatment for H pylori. In CADET-Hp the control group patients did not receive eradication treatment and were therefore at risk of recurrent peptic ulcers that had healed initially with omeprazole
Furthermore, patients with nonulcer dyspepsia may also benefit from H pylori eradication. A Cochrane review of 9 placebo controlled trials of H pylori eradication treatment in patients without peptic ulcers or oesophagitis at endoscopy found a number needed to treat of 15.4
The CADET-Hp trial does not show conclusively that H pylori test and treat is more cost effective in primary care than omeprazole alone, because it was only done in H pylori–positive patients. The cost effectiveness of this strategy needs to be tested by randomising patients with dyspepsia, both positive and negative for H pylori, before noninvasive testing for H pylori, to determine the effect of the management strategy on the whole group. It does, however, lend more support to the eradication of H pylori in all patients known to be infected.
Footnotes
-
Source of funding: Astra-Zeneca Canada Inc.
-
For correspondence: Dr N Chiba, Surrey GI Clinic/Research, Guelph, Ontario, Canada. chiban{at}on.aibn.com
-
Abstract and commentary also published in ACP Journal Club
↵† Information provided by author.