Intended for healthcare professionals

Editorials

Test and treat for dyspepsia—but which test?

BMJ 2005; 330 doi: https://doi.org/10.1136/bmj.330.7483.105 (Published 13 January 2005) Cite this as: BMJ 2005;330:105
  1. Cliodna McNulty (jill.whiting{at}hpa.org.uk), consultant medical microbiologist,
  2. Louise Teare, consultant medical microbiologist,
  3. Robert Owen, head,
  4. David Tompkins, laboratory director,
  5. Peter Hawtin, clinical scientist,
  6. Kenneth McColl, professor of gastroenterology
  1. Health Protection Agency Primary Care Unit, Microbiology Department Gloucestershire Royal Hospital, Great Western Road, Gloucester GL1 3NN
  2. Mid-Essex, Department of Microbiology, Chelmsford CM2 OYX
  3. Campylobacter and Helicobacter Reference Unit, Health Protection Agency, Specialist and Reference Microbiology Division, London NW9 5HT
  4. Health Protection Agency, Yorkshire and the Humber, Leeds Laboratory, Leeds LS15 7TR
  5. Health Protection Agency, Southampton Laboratory, Southampton General Hospital, Southampton SO16 6YD
  6. University of Glasgow, Western Infirmary, Glasgow G11 6NT

    Urea breath test and stool antigen test are better than serological tests

    Managing dyspepsia costs the NHS over £500m annually.1 European dyspepsia guidelines and those from the National Institute for Clinical Excellence (NICE) say that patients with persistent or recurrent uncomplicated dyspepsia should have a non-invasive Helicobacter pylori test and, if the test is positive, receive triple therapy.24 With a policy requiring non-invasive testing and treatment we need to use an accurate test so that the patients receive the correct treatment. The urea breath test and serology were the first non-invasive tests available; the urea breath test is the more accurate. This test detects products of the enzyme urease produced by live H pylori in the stomach and is 95% sensitive and specific.5 The breath test has not been used much in primary care in the United Kingdom, probably because it is time consuming as it requires two breath samples, taken 20 minutes apart.

    Serology is the main non-invasive test used in the United Kingdom and is notably less accurate than the urea breath test.5 6 A positive serology result can mean one of three things: that the patient is infected at the time of the test; that the patient was once infected, but by the time of the test, infection has resolved, either by specific therapy or naturally; or that the test is detecting non-specific cross reacting antibodies.

    Another accurate non-invasive test is now available. The stool antigen test detects H pylori antigens passed in the faeces. The first commercially available test, which used polyclonal antibody raised in rabbits, has been used in thousands of patients across Europe and is almost as specific (91.9%) and sensitive (92.4%) as the urea breath test.7 Some centres have, however, found appreciable variation between batches, and a monoclonal antibody kit is now available commercially, which avoids this.8 The monoclonal test is reported to be as accurate as the urea breath test (specificity 97.5%, sensitivity 94.7%)8 It uses similar laboratory methods to the serology test and can be introduced with ease into routine laboratory practice.9 w1

    Antibody concentrations to H pylori fall slowly after eradication of the infection.10 In contrast to serology, stool antigen testing is useful for confirming eradication of the infection following treatment.7 8 w1Although equivalent to the urea breath test in performance (see table on bmj.com), the stool test is considerably less expensive and less time consuming, and investigators have found it acceptable to patients.11 A disadvantage of breath and stool antigen tests is that patients must stop taking proton pump inhibitors for at least two weeks before the test and H2 receptor antagonists for one day.7 w2 w3Any antibiotics must be stopped four weeks before.

    The accuracy of H pylori tests has been determined mainly in patients at endoscopy in whom the prevalence of H pylori is high and the positive predictive value of all tests therefore high. However, as the prevalence of H pylori falls, the positive predictive value of all tests falls.12 The lower the specificity of a test, the greater the fall in positive predictive value with falling prevalence. When using the urea breath test or monoclonal stool antigen test in developed countries, where typically 25% of dyspeptic patients are H pylori positive, only 3% (62 for stool, 65 for urea breath test of 2000) of patients will receive unnecessary antibiotics.3 5 In contrast, using a serology based test 255 of the 2000 patients tested are likely to receive an incorrect diagnosis of active H pylori infection and receive inappropriate treatment.2 3 5

    Serology leads to at least four times as many false positive results as the urea breath test or second generation monoclonal stool antigen test, with associated unnecessary treatment and increasing risks of antibiotic resistance in other bacterial flora. If the dyspepsia “test and treat” guidance is implemented widely across Europe the number of patients receiving treatment to eradicate H pylori could easily double. We need to have an easy, accurate diagnostic test and the stool antigen test is just that. The European Helicobacter Study Group4 and NICE dyspepsia guidance3 now endorse the use of urea breath tests or stool antigen tests over serology. Any small additional cost to the healthcare provider will be far offset by improved diagnostic accuracy and reduced use of antibiotics. Furthermore, as these tests replace serology and market forces come into play, the price of the breath and stool tests is likely to come down. Clinicians are therefore best advised to inform patients that the minor inconvenience of providing a stool or breath sample is far outweighed by the increased accuracy of the tests. Clinicians should request healthcare providers to fund office based tests or local laboratories to include these tests in their repertoire.

    Embedded Image Additional references w1-w5 and table are on bmj.com

    References

    1. 1.
    2. 2.
    3. 3.
    4. 4.
    5. 5.
    6. 6.
    7. 7.
    8. 8.
    9. 9.
    10. 10.
    11. 11.
    12. 12.
    View Abstract