The way docs decide who to test for celiac disease might help explain low diagnosis rates
By Amy Ratner, Medical and Science News Analyst
Despite growing awareness, most people who have celiac disease still have not been diagnosed.
Although celiac disease affects about 1 percent of the population— 3 million people in the United States — about 80 percent don’t know they have the genetic autoimmune condition.
Researchers from the Mayo Clinic in Minnesota may have uncovered some of the reasons why. They suggest in a study published in the journal, Alimentary Pharmacology and Therapeutics, that methods commonly used to determine who to test for celiac disease are not effective in detecting undiagnosed cases.
Patients who don’t have typical symptoms or don’t have symptoms at all may account for a significant portion of undetected cases of celiac disease, said Joseph A. Murray, M.D., director of the Mayo Clinic Celiac Disease Program, and a study author.
Who to screen
Celiac disease in clinical practice is driven by testing patients who have symptoms, called case finding, according to the study. Patients are also “tested in the community” when their complaints of symptoms get to a level that prompts their physician to test for celiac disease without the formal structure of case finding.
the practice of testing individuals who are at increased risk for a disease primarily due to the presence of symptoms or conditions associated with the disease.
Despite reports that case finding has increased diagnosis in the primary care setting, “there is increasing suspicion that [it] may not be effective and that indications to test, such as gastrointestinal symptoms, have poor diagnostic accuracy in identifying celiac disease,” the authors wrote, referencing earlier research that looked at the usefulness of symptoms in diagnosis.
“Gastroenterology symptoms are certainly common signs of celiac disease when it becomes obviously symptomatic.So, patients present with diarrhea, abdominal pain, bloating, and anemia with fatigue,” Murray said. “What this study tells us is that many patients are out there in the community who likely have some measure of celiac disease and who have not developed any of these symptoms.This also indicates that, perhaps, celiac disease is being detected much more without those symptoms being present.”
The study was based on screening of nearly 50,000 stored blood samples taken from patients from Olmsted County, Minnesota, treated at the Mayo Clinic from 1995 to 2009. Researchers tested the samples for a biomarker of celiac disease, tissue transglutaminase (tTg) IgA. Samples that were positive were further tested for endomysial antibodies, which are very specific for celiac disease. Ultimately, 400 cases of undiagnosed celiac disease were identified and matched with 400 controls without celiac disease.
Researchers then reviewed the medical records of both those who were identified as having undiagnosed celiac disease and the controls. They looked at whether all 800 patients had conditions that indicated the need for celiac disease testing or had celiac disease blood testing or a biopsy regardless of the reason these tests were done.
Of the 400 cases of undiagnosed celiac disease identified in the study, 159 had indications to screen, including classic symptoms such as diarrhea, and non-classic symptoms, such as other auto immune diseases. Ultimately, 92 cases would have met clinical criteria for diagnosis of celiac disease over time, usually taking more than five years. The remaining 241 cases had no indication to screen and would be likely to remain undiagnosed.
Among the controls, 147 cases had an indication to screen, nearly as many as were found in the cases of undiagnosed celiac disease. None of these were ever found to have celiac disease.
The similarity of these results suggests that case finding based on the presence of accepted indications would be unable to discriminate between those with and without celiac disease, the study said.
In most cases, the conditions that might prompt clinical testing for celiac disease occurred with similar frequency for both those who had undiagnosed celiac disease and controls. However, hypothyroidism was significantly more likely in those with celiac disease. Meanwhile, undiagnosed patients were less likely than controls to have dyspepsia or chronic diarrhea.
For more than a decade, studies have been finding that gastrointestinal symptoms are not the only or predominant signs of celiac disease. More than half of adults with celiac disease have symptoms not related to the gastrointestinal system, including anemia, osteoporosis and joint pain, according to the Mayo Clinic.
Less than 5 percent of all 800 cases included in the study were tested for celiac disease by a provider in the community prior to the blood tests run as part of the study. And testing was less common among those who did have undiagnosed celiac disease – 9 – than in those who did not have the disease – 26. Blood tests were done about equally in the two groups, but controls were significantly more likely to have biopsies.
“The rate of testing for celiac disease was low in both groups, but was more likely in controls,” the study said.
“The fact that controls were more likely tested than those with undiagnosed is interesting and probably tells us that our current strategies for identifying people at risk for celiac disease are not especially good and was part of the point of this study,” Murray said.
When researchers looked at the total number of blood samples, they found that among the 47, 056 who did not have celiac disease, 17,410 had an indication to screen. That means if case finding was used to diagnose celiac disease, 17, 569 people would have to be tested to identify 159 cases, but about 60 percent of those would still remain undetected.
“A more effective method of detection of celiac disease is needed,” the study concluded.
Those with undiagnosed celiac disease were more likely over time to develop osteoporosis, dermatitis herpetiformis, chronic fatigue, thyroiditis and autoimmune diseases. The development of conditions increased the chances that someone would eventually be found to have celiac disease, the study said, calling it “a later opportunity for diagnosis.”
Natural history of disease:
progression of a disease in a person over time, in the absence of treatment
This long-term look at what happens to undiagnosed patients helps fill in the natural history of celiac disease. “The most important thing about natural history of undiagnosed of celiac disease is to determine what is the outcome if it is left undetected and untreated compared to the early detection and intervention,” Murray said, noting more study is needed. But that’s a challenge because it would require participation of patients who were diagnosed with celiac disease that was then left untreated.
Undiagnosed patients were also more likely to incidentally be found to have villous atrophy and to have a family member diagnosed with celiac disease.
Lack of detailed family history about celiac disease was one of the limitations of the study, and the authors noted they were likely unable to accurately assess how often it occurred. Family history has been found to be highly indicative of being at risk for celiac disease. Lack of biopsy confirmation of celiac disease diagnosis was another study limitation.
The study raises the question of how best to detect celiac disease if case finding is not effective. The authors note that some experts support mass screening of the general population, but they point out that the US. Preventive Services Task Force recently concluded there is insufficient evidence to assess the balance of benefits and harms of screening for celiac disease in asymptomatic individuals. The task force is a panel of experts in disease prevention that works to improve the health of Americans by making evidence-based recommendations about clinical preventive services.
The challenge is to increase detection of undiagnosed symptomatic celiac disease but also decrease the number of people without celiac disease put at risk by screening, the study said.The risks include an incorrect diagnosis of celiac disease based on a borderline blood test result and unnecessary biopsies. “
Natural language software: a branch of computer science that employs linguistics to analyze regular speech or text. The software can be used to translate doctors’ and nurses’ clinical notes into standardized data that can be used to track patterns of disease.
Solutions proposed in the study are: use of natural language software to process electronic medical records; systematic acquisition of symptoms and family history; testing of those with combinations of symptoms or indications to test; and identification of new indications to test based on additional research.
Currently, patients are not consistently asked about family history of celiac disease, according to Murray. “I believe should be routine but it is not,” he said. “The question, in regard to what symptoms would be good indicators, is very important because this is not a question that is obvious at all and is one of the major challenges of celiac disease detection.”