When a patient suffers from a lingering sore throat, many healthcare practitioners turn to rapid antigen streptococcus tests (RAST), commonly known as rapid strep tests, to help determine the presence of group A Streptococcus bacteria. Given the 13 million office visits annually in the United States for these sore throats – called pharyngitis – clinicians’ reliance on these tests is significant, but despite their popularity, researchers found the accuracy is inconsistent.
To get a clearer picture of the tests’ accuracy, researchers including Benjamin Littenberg, M.D., University of Vermont Henry and Carleen Tufo Professor of Medicine and professor of nursing, and colleagues analyzed and summarized the variances in 59 studies of rapid strep tests encompassing nearly 56,000 patients, mostly in North America and Europe. This systematic analysis was published in the November 4, 2014 PLoS ONE.
Current guidelines from the Infectious Diseases Society of America (IDSA) for treating strep are not based on a meta-analysis that could shed light on why test performance varies, note researchers. The IDSA’s non-systematic review shows that tests produce excellent specificity, correctly diagnosing 95 percent or more of sick people, but have variable sensitivity, accurately returning negative results for 70 to 95 percent of healthy people. The IDSA recommends prescribing antibiotics for patients who test positive.
Littenberg and colleagues’ meta-analysis reveals where studies vary, but not why. They examined possible sources of heterogeneity, including funding sponsors, settings – such as emergency rooms, outpatient clinics or student health centers – as well as number of swabs, methodology (immunochromatographic, enzyme immunoassay or optical immune-assay), age groups, study design and more.
“What we discovered was there weren’t a lot of really good studies,” Littenberg says. “In some cases, there is publication bias. But there are other cases where there is something else going on that we don’t understand. It’s not clear why the test is more accurate in one situation versus another. That inconsistency is a real problem."
The analysis, he says, “will not resolve the issue of whether or not to get rapid strep tests.” But it does make clear that “the sources of heterogeneity need to be addressed in future tests, and we need to get better data.”
Among the team’s finding regarding the most widely used and most studied tests (those that employ lateral flow immunochromatographic methods) are:
· The tests can more accurately identify strep in adults versus children
· Inconsistencies were higher for children than for adults.
· Test results for children are more consistently accurate when used in outpatient clinics versus emergency rooms.
· The tests are shown to be more accurate in evaluations sponsored by manufacturers.
The group’s analysis employed a statistical technique, which uses a graphical curve to compare data points, that Littenberg published in 1993 to compare diagnostic tests with the “gold standard” test. Since then, it has been used hundreds of times to compare positive-negative tests for a wide variety of medical conditions.
Two professors of medicine at the University of Alabama at Birmingham – Robert M. Centor, M.D., regional dean for the Huntsville Regional Medical Campus and the foremost expert on management of adult strep throat, and Carlos A. Estrada, M.D. – called on Littenberg’s expertise for the PLoS ONE study.
Lead authors on the article are Emily Stewart, M.D., of Walter Reed National Military Medical Center, Bethesda, Md., and Brian Davis, M.D., of University of Texas Southwestern Medical Center, Dallas. B. Lee Clemans-Taylor, M.L.S., director of the medical library at University of Alabama at Birmingham-Huntsville, is a co-author.