Helicobacter pylori (H. pylori) is a common bacterial infection that as many as 50 percent of all people in the world have been exposed to. Fortunately, not everyone who has the infection will experience symptoms, but some may suffer a wide range of effects, and are at risk to develop peptic ulcers, gastritis or some stomach cancers.
Both the American Gastroenterology Association (since 2005) and American College of Gastroenterology (since 2007) recommend use of either the H. pylori stool antigen test (SAT) or the urea breath test (UBT) as a first-line diagnostic test for suspected H. pylori infection, and recommend that clinicians avoid serologic (blood) testing due to poor clinical performance. However, blood tests remain the most commonly ordered diagnostic test.
The reason for this disconnect is unknown, but anecdotally could be ascribed to the desire to cause less physical or emotional discomfort for patients, as well as a perceived cost benefit. Patients must stop taking acid-suppressing drugs and antibiotics two weeks before either of the tests, and in the case of the SAT, must collect a stool sample. Further, the blood test is less expensive to administer than either of the other two tests.
Researchers from the Mayo Clinic Robert D. and Patricia E. Kern Center for the Science of Health Care Delivery published a study earlier this year that used the OptumLabs Data Warehouse to shed light on this issue. After looking at OLDW medical claims data for patients from 2010-2012, we found more than 500,000 H. pylori first-line diagnostic tests were conducted across the approximately 150 million patients in the data set, with the vast majority receiving blood tests (366,846).
Of those tested using the three methods, a positive test result indicative of active H. pylori infection was found in only 4.2% of blood tests, 18% of UBT and 13% of SAT. This was concerning on several fronts.
First, a blood test cannot distinguish between an active or past H. pylori infection, so there is a high potential for misdiagnosis resulting in inappropriate treatment. Therapy includes an antibiotic, and overuse of antibiotics can result in antibiotic resistance. Conversely, untreated H. pylori infections can result in significant health concerns. Finally, inappropriate testing results in extra costs to patients, providers and insurers.
While certain patient scenarios may necessitate the use of a blood test rather than an SAT or UBT for diagnosis, there is very little likelihood that those reasons could explain the apparent vast overuse of the blood test in this cohort.
This case illustrates the difficulty that medical researchers have in translating findings from discovery to common adoption across clinical practice. Even when the leaders in the field promulgate their well-researched recommendations, achieving widespread practice adoption is problematic. Clearly we need to continue to our efforts to disseminate research findings and educate health care providers. But this study also showed an opportunity for medical providers and payers to partner together in order to influence practice patterns. While some private insurance companies already have stopped reimbursements for H. pylori diagnostic blood tests, more could (and perhaps should) do so.
We are excited by the opportunities that OLDW presents to better understand any number of patient concerns. Our goal in the center is to promote evidence-based health care that enhances patient experience, improves population health, and manages the total costs of care.
Learn more about the science of health care delivery, best practices and current research at Delivery Science Summit 2015, September 16-18, 2015, in Rochester, Minnesota.