Intended for healthcare professionals


Absence of dyspeptic symptoms as a test for Helicobacter pylori eradication

BMJ 1996; 312 doi: (Published 10 February 1996) Cite this as: BMJ 1996;312:349
  1. Perminder S Phull, clinical research fellow, gastroenterology unita,
  2. David Halliday, director, section of nutrition researcha,
  3. Ashley B Price, consultant pathologista,
  4. Meron R Jacyna, consultant gastroenterologista
  1. aNorthwick Park Hospital, Harrow, Middlesex HA1 3UJ
  1. Correspondence to: Dr Phull.
  • Accepted 23 November 1995

Eradication of Helicobacter pylori changes the course of duodenal ulcer disease and effectively cures it.1 One issue that limits the more widespread use of eradication treatment, however, is whether patients should undergo a test to confirm successful eradication. Many of the tests necessitate endoscopy and are therefore invasive, time consuming, and expensive. The carbon-13 or carbon-14 labelled urea breath test has been suggested as probably the best method of follow up after eradication treatment.2 This test is non-invasive, well tolerated, and easy to perform, but widespread use is limited by availability, expense (in the case of 13C) or problems with radioisotope handling (14C). We have previously shown that eradication of H pylori results in a highly significant reduction of dyspeptic symptoms in patients with duodenal ulcer disease.3 In this study we assessed the absence of dyspeptic symptoms as a measure of H pylori eradication, using the 13C-urea breath test as the gold standard for measuring the presence of H pylori.

Methods and results

Data were collected prospectively on all patients positive for H pylori with endoscopically proved duodenal ulcer disease who were given eradication treatment at this hospital in 1991-4. Patients with a history of ulcer haemorrhage or perforation and those taking non-steroidal anti-inflammatory drugs were excluded. Whether H pylori was present was determined by histology or the 13C-urea breath test4 before treatment. Treatment was with: colloidal bismuth subcitrate, tetracycline, and metronidazole for seven days; omeprazole, colloidal bismuth subcitrate, tetracycline, and metronidazole for seven days; omeprazole and amoxycillin for 14 days; or lansoprazole and clarithromycin for 14 days. Patients were advised not to take any acid suppressing medication after treatment.

All patients were reviewed one and six months after completion of treatment, when the urea breath test was performed and data collected on dyspeptic symptoms using the gastrointestinal symptom rating scale.5 Patients were asked to grade epigastric discomfort, heartburn, nausea, vomiting, and wind on a severity scale of 0-3 (nil, mild, moderate, severe). Patients were unaware whether H pylori was still present or not until after their six month review.

One hundred and twelve patients (mean age 48 years, range 23-75; 71.4% men; 33% smokers) had undergone eradication treatment, and in 80 (71.4%) H pylori had been eradicated, as defined by a negative urea breath test. Table 1 shows the correlation of dyspeptic symptoms with the breath test result. With absence of all symptoms as a measure of successful eradication, the sensitivity was 87.5% at one month and 97.5% at six months, with specificity being 56.3% and 90.6% respectively. The figures for sensitivity were improved when only absence of epigastric discomfort was used for assessment (88.8% and 100%), with little change in specificity (50% and 90.6%).

Table 1

Correlation of dyspeptic symptoms with results of 13C-urea breath test after H pylori eradication treatment. Results are numbers (and percentages) of patients

View this table:


Our results suggest that in patients with duodenal ulcer a conventional test to assess whether H pylori has been eradicated after treatment may not be necessary. At six months the symptom based method for confirming eradication had a high sensitivity and specificity. At one month the specificity was lower, as about half the patients who remain positive for H pylori experience a temporary improvement in their symptoms after treatment. Although bias cannot be entirely excluded, the patients were unaware of their status until the six month review, so the one month results were in effect double blind as neither the patient nor the interviewing doctor were aware of the patient's H pylori status.

This study excluded patients with a history of haemorrhage or perforation, and we do not recommend using only symptoms to assess treatment in these high risk patients. Similarly, we advise further studies to evaluate symptom based assessment in patients with gastric ulcers. Our results do, however, suggest that patients with uncomplicated duodenal ulceration who are asymptomatic after eradication treatment do not need further investigation or treatment. Patients can simply be advised to return to their doctor if they experience further symptoms.


  • Funding PSP is partly funded by a grant from Glaxo Research and Development, UK.

  • Conflict of interest None.


  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.
View Abstract