Therapy for Helicobacter pylori in Patients with Nonulcer Dyspepsia. A Meta-Analysis of Randomized, Controlled Trials. Laine L, Schoenfeld P, Fennerty MB.
Overall Study Question
Helicobacter pylori (H. pylori) infection is a common manifestation with a Canadian prevalence of 20-40% in the general population and ~50% in patients with nonulcer dyspepsia. While the eradication of H. pylori is the standard of care in patients with peptic ulcer disease, the role of this organism in nonulcer dyspepsia is controversial. The purpose of this meta-analysis was to determine whether treatment of H. pylori infection in patients with nonulcer dyspepsia significantly improves symptoms. A secondary outcome was to determine whether patients whose H. pylori infection was cured had significantly less dyspepsia compared to those with persistent infection.
Are the Results of the Study Valid?
1. Did the overview address a focused clinical question?
2. Were the criteria used to select articles for inclusion appropriate?
Yes. Criteria for study selection were defined a priori and included studies that:
1) involved patients with non-ulcer dyspepsia and documented H. pylori infections;
2) compared dual, triple or quadruple therapy for H. pylori with a placebo control;
3) were randomized and controlled trials;
4) followed patients for at least one month after conclusion of therapy; and
5) assessed symptoms of nonulcer dyspepsia.
Ten randomized, controlled trials from 1995-1999 that used combination therapy with proven effectiveness against H. pylori were included in the meta-analysis. Control therapy consisted of placebo or antisecretory therapy alone.
3. Is it unlikely that important, relevant studies were missed?
Yes. Two independent investigators reviewed Medline and HealthStar as well as abstracts from the American Gastroenterological Association, the American College of Gastroenterology, and the European H. pylori study Group. Reference lists were checked from all articles that met the selection criteria. Efforts were also made to identify unpublished trials from manufacturers and experts in the field, although no relevant trials were identified in this manner.
4. Was the validity of the included studies appraised?
Yes. Each article was given a score (1-5) by two independent investigators based upon the description of randomization, blinding, and patient withdrawals. There was 97% agreement between the two reviewers. Seven of the ten trials scored a rating between 3-5 and only one trial had a rating score of 1.
5. Were the assessments of studies reproducible?
Yes. A pooled random-effects estimates of odds ratios (OR) and 95% confidence intervals (CI) was calculated as described by Fleiss. Tests for heterogeneity across studies were performed. A p value of less than 0.5 indicated statistical heterogeneity, meaning that the trials were too different to combine in a meta-analysis.
What were the Results?
1. What are the overall results of the review?
One hundred and fifty-four articles were identified, of which ten met the inclusion criteria. Seven trials including 1544 patients were used to assess the primary outcome (proportion of patients with treatment success at least 1 month after completion of eradication therapy compared to control therapy); three of these trials were in abstract form only. One of the abstracts utilized a 5 day triple therapy eradication regimen and most H. pylori guidelines recommend a minimum 7 day treatment period. (1) A second abstract used a double therapy regimen (amoxicillin plus omeprazole) which has an eradication success rate of 50-80% compared to 85-98% for triple and quadruple regimens.
Overall, no differences in dyspepsia symptoms were found between treatment and control groups. The OR was 1.29 (95% CI, 0.89 to 1.89, p = 0.18) for the 7 trials. Significant heterogeneity was found which resolved upon exclusion of one trial. The OR was 1.07 (95% CI, 0.83 to 1.37, p > 0.2) for treatment success with H. pylori eradication for the 6 remaining trials. The trial that was excluded for heterogeneity was the only trial showing a significant result in favour of H. pylori eradication.
For the secondary outcome (treatment success in H. pylori cured patients versus those with persistent infection), the OR was 1.17 (95% CI, 0.87 to 1.59, p>0.2) for treatment that resulted in cure rather than persistent infection. No evidence of heterogeneity was found in the five trials used to assess this outcome.
2. How precise were the results?
Several different analyses were performed, all with similar results. Four studies that provided a specific definition of dyspepsia showed an OR for treatment success of 1.04 (CI, 0.80 to 1.35, p>0.2) with no evidence of heterogeneity. Four studies with the highest methodological quality (excluded the abstracts) demonstrated statistical heterogeneity with an OR of 1.41 (CI, 0.85 to 2.33, p=0.18). Analysis of six studies that had a follow-up period of at least a 6 months also revealed statistical heterogeneity with an OR of 1.24 (CI, 0.84 to 1.84, p>0.2).
Seven studies provided a scoring system to assess dyspepsia. To allow comparison of the mean scores between study groups, the mean numerical change in symptom score was divided by the maximum possible symptom score for each study. In 5 of these studies, both the raw and adjusted mean dyspepsia scores varied by less than 5 percentage points between the treatment and control groups. The smallest trial (n=41) showed the greatest benefit in favour of H. pylori therapy (decrease in score of 35% vs. 16% for control group), however this trial had the lowest methodological rating by the investigators with a score of 1.
3. How much does allowance for uncertainty change the results?
The results do not change. See number 2 above.
Will the Results Help Me in Caring for My Patients
1. Can the results be applied to my patient care?
Yes. Nonulcer dyspepsia represent a relatively common disorder with an prevalence ranging from 25 to 41%.2 This meta-analysis suggests that alternate therapy to H. pylori eradication is advisable in patients with diagnosed nonulcer dyspepsia who do no have an underlying organic disorder, such as duodenal or gastric ulcer. Other therapy to be explored in these patients include prokinetic agents, acid suppression, psychotropic therapy, or antinociceptive therapy.
2. Were all clinically important outcomes considered?
3. Are the likely treatment benefits worth the potential harms and costs?
No. The results of this trial do not support treatment of H. pylori infection in diagnosed nonulcer dyspepsia. While H. pylori is easily detectable by urea breath test or serology, overtreatment of H. pylori with double or triple antibiotic regimens can lead to antibiotic resistance and increase in health care costs.
This meta-analysis provides further support for the recommendations made by National Institute of Health in 19944 that H. pylori should not be eradicated in diagnosed nonulcer dyspepsia. What this meta-analysis does not answer, is whether H. pylori should be eradicated in uninvestigated dyspepsia. Complete investigation for dyspepsia is recommended for all patients greater than 50 years of age or those who present with alarm symptoms (e.g. bleeding, weight loss, vomiting). For younger patients with no alarm features and who are not receiving NSAID or present with reflux symptoms, the 2000 Canadian Dyspepsia Working Group recommend investigating for H. pylori and treating all positive patients; this decision is based upon the assumption that between 5-15% of these patients will have underlying peptic ulcer disease and thus benefit from eradication of the organism. Further study is required to elucidate the appropriate management of uninvestigated dyspepsia.