CEO SUMMARY: Doctors rely on labs for the definitive diagnosis of the bacterium that causes whooping cough. Yet in several recent cases, the use of a molecular test for a definitive diagnosis led to questions about the effectiveness of the test. Molecular pathologists recommend labs may need to do a second test to confirm initial findings and labs may need to be diligent about any equivocal findings to avoid reporting false positive results.
IN RECENT MONTHS, SEVERAL HOSPITALS made headlines as they coped with outbreaks of Bordetella pertussis (whooping cough). In some cases, the news was not the identification of Bordetella pertussis within the institution, but the fact that laboratory testing appeared to be generating a high rate of false positives.
As more cases of Bordetella pertussis surface, health officials are facing difficult questions about how to diagnosis this infection. The issue is important to pathologists and lab directors because they will increasingly be asked to identify pertussis and other conditions using molecular diagnostics. “To do so accurately, definitive protocols need to be developed that may require labs to perform additional testing,” said Michael Mihalov, M.D., Chair of Pathology and Director of the Diagnostic Molecular Pathology Lab at Resurrection Medical Center, in Chicago.
Mihalov is uniquely qualified to comment on this topic because he is an expert in molecular pathology and he trained as a pediatric resident at Chicago’s Cook County Hospital before going into pathology. As a result, he has clinical experience with whooping cough patients.
Last year, infectious disease specialists at the Dartmouth-Hitchcock Medical Center in New Hampshire identified what they thought was an outbreak of pertussis. Over the course of the outbreak, about 1,000 health care workers were tested, 142 were told they had the condition, and more than 4,500 hospital workers got the acellular pertussis vaccine.
A Case of Misplaced Faith?
Last month, a number of newspapers reported that epidemiologists and infectious disease specialists had decided that healthcare experts had placed too much faith in the new, molecular test for pertussis. “In fact, the issue is not so much that experts placed too much faith in molecular testing, but rather health systems have not yet developed a clinically relevant algorithm for interpreting molecular results,” observed Mihalov.
“Recent articles on this outbreak focused only on the molecular test and did- n’t elucidate the difficulties of identifying the specimen with a culture and using a molecular test,” Mihalov said. “Growing a culture is just as difficult, if not more difficult, as doing the molecular test. Yet, these reports said the molecular test failed. They didn’t say that the test it is replacing is also fraught with problems.
“For cases of pertussis and other similar infections, we’re in a transition period where the medical field is still attempting to define what constitutes a clinical infection. Culture is probably not the gold standard any longer because it is too insensitive. If the clinician obtains an adequate specimen and has a lab nearby that is proficient in doing cultures, then you could do it by culture,” Mihalov said. “But in most cases, we don’t have that, which is why molecular technology is replacing the growing of cultures. The Illinois Department of Health does molecular testing almost exclusively for pertussis, for example.
Molecular Testing Caveats
“But there are caveats to molecular testing which we have to understand,” he added. “Every test has a positive range, an equivocal range, and a negative range. If the original tests are in the equivocal range, then pathologists and clinicians should think carefully before assuming that the specimens are clinically diagnostic of a disease. In the Dartmouth-Hitchcock case, clinicians made an assumption that the result was positive, meaning there was an outbreak. Those assumptions may need some reworking, particularly if any of the initial results were in the equivocal range. In fact, the data from Dartmouth-Hitchcock is still being analyzed, I believe, and it remains to be seen if pertussis organisms were present or if there was some cross reactivity with other nucleic acid that was present during the original testing.
“Molecular testing has inherent limitations that must be taken into account,” Mihalov continued. “Many of these tests are developed by individual labs, and not every lab uses the same algorithm to confirm positive results.
“As we move forward, and more of these situations occur in which clinicians use lab data to define outbreaks, the tests and the clinical expertise will both improve,” he said. “First, many labs will employ a second molecular test to confirm the initial result. And second, over time, the FDA is likely to approve methods for testing such specimens. When the FDA approves a specific test, it is likely to have undergone rigorous testing to define the specificity of the test.
“What’s happening today is that many labs are using analyte specific reagents (ASRs), or home brews, for the molecular test they use to identify pertussis,” Mihalov explained. “These home brews are not FDA approved tests but the FDA has allowed labs to develop in-house procedures using these ASRs. When labs use them, they also need to validate the test in their own labs. The problem is that in most labs, we are limited in the number of organisms against which we can test our new reagents. Furthermore, most labs don’t have enough experience to know the exact specificity of the test when using a home brew assay.
“As more labs use molecular tests,” Mihalov said, “they will need to more rigorously determine the specificity of their tests. In addition, every lab will need to ensure that the result is unequivocally positive before it labels something a definite infection, especially with an organism like Bordetella pertussis.
“To confirm a diagnosis, it may be necessary to perform additional testing,” he said. “There are two ways to do that. First, you can ask the clinician for another sample, particularly when the first result is equivocal. Or, second, you could get a confirmation of an initial positive result with a second molecular test. Ideally the second test would target a different region of the organism’s nucleic acid than the initial test. Then, only if both of those tests are positive would we call it a positive result.”