John: Hello, I'm John Hilton, editor in the Cochrane Editorial and Methods department. Cochrane Diagnostic Test Accuracy reviews bring together the evidence that should help practitioners to choose the most appropriate way to diagnose an illness or condition. This is now the case for the stomach bug, Helicobacter pylori, and we asked Lawrence Best from the UCL Medical School in London England to tell us about the findings of this March 2018 review.

Lawrence: Helicobacter pylori is a type of bacteria that is present in the stomach of some people. It is believed to cause a number of cancers, including stomach, pancreatic, and throat cancer; and is also linked with problems such as stomach ulcers, heart burn, and a bloated feeling.
Finding out if someone has Helicobacter pylori allows treatment to be started. Testing can be done in different ways. The definitive way is by an endoscopic biopsy which uses a thin flexible tube called an endoscope to take tissue from the stomach that is then examined under the microscope. However, although this is currently the most accurate available test, it causes physical discomfort to the patient and is associated with a very small risk of patient harm. There are other non-invasive tests which are much less uncomfortable and have minimal or no risk of harm, which would make them desirable alternatives if they can be shown to be as accurate as endoscopic biopsy.
That’s why we did this review, in which we compared the accuracy of three common non-invasive tests for diagnosing Helicobacter pylori. Namely, urea breath tests, blood tests, and stool tests.
We found 101 studies which included about eleven thousand people who were tested for Helicobacter pylori. Of these, approximately half were shown to have the infection. All the studies used one of the three tests we were interested in and compared the results with the diagnosis given by endoscopic biopsy. Most of the studies included participants with heart burn or similar problems in the stomach and excluded people who had previously undergone partial removal of the stomach and those being treated for Helicobacter pylori.
Studies varied in the limits they used before saying a test was positive for Helicobacter pylori infection and the type of stains used to examine the tissue from endoscopic biopsy, and were generally of poor methodological quality. When we looked at all the data, we found that urea breath tests were more accurate than blood and stool tests. The results mean that based on the included studies, on average, if 1000 people are tested non-invasively and 537 have Helicobacter pylori, 46 people without Helicobacter pylori will be misdiagnosed as having it, and between 30 and 89 of those who have Helicobacter pylori infection will be missed depending on the test.
In summary, urea breath tests, blood tests, and stool tests may be suitable for finding out if Helicobacter pylori is present in children or adults, who have not previously undergone stomach operations or recently received treatment for Helicobacter pylori infection. However, the threshold at which these tests should be considered positive remains unclear. We also still cannot definitively tell which test is more accurate and need new high-quality studies to resolve this uncertainty.

John: If you would like to read more about the different tests, and watch for future updates of the review should those new studies get done, go to Cochrane Library dot com and search 'diagnostic tests for H pylori' to hit the review.