It has to be said at the outset that there is no single gold-standard test available to establish if someone has SIBO. In fact, the breath-tests which are the most popular tests in use, may not only be inaccurate but the interpretation of them largely depends on the doctors involved. Some doctors may find a particular reading indicative of SIBO and others may find the same reading is not.
Hydrogen Breath Tests
These tests are used because the human body does not produce hydrogen, but the bacteria within the body do, and they do this when they metabolize carbohydrates. Different sugars are introduced orally to the patient and the level of hydrogen, and sometimes methane, are then measured.
The two types of sugars which are normally used in the tests are lactulose and glucose. Lactulose is not a natural sugar but a synthetic composite which contains lactose. This test is thought to be best at diagnosing problems associated with the lower portion of the small intestine but not the upper. For establishing if the bacterial overgrowth is in the lower part of the small intestine you are looking at the glucose test. Glucose is a sugar which is directly absorbed by the human body and it is done in the upper part of the small intestine. So this test is more suited to those suspected of having an overgrowth of bacteria in the upper part of the small intestine.
However there are clear indications that these tests can be unreliable and the reasons for this are extensive. They can range from something as basic as patients not adhering to dietary guidelines prior to taking the test or to measurements being affected by non-gas producing bacteria. One detailed research paper gives substantial information on the many problems involved and can be found here.
Endoscopy is also sometimes suggested as an alternative to the breath tests. However this procedure is not only invasive and more expensive but it can only examine the bacteria which extend as far as the duodenum. The duodenum is in the first section of the small intestine and so for patients suspected of SIBO, where the bacterial overgrowth is thought to reside in the lower end, it cannot produce any accurate findings.
Other tests which are available relate to examining bacteria found in stool samples. Yet again here the results will only show bacterial levels from the large intestine (large colon or bowel). Again they are not measuring the number of bacteria from the lower end of the small intestine which would be indicative of SIBO.
So the best option currently available for diagnosing SIBO is the lactulose test. However as mentioned earlier it has its limitations and can produce both false negative and false positive results.