False positives plague TB screening of health care workers
Retest and use clinical judgment, experts advise
Most of the positive results in routine tuberculosis screening of health care workers are false positives. That statistical artifact is creating headaches for employee health professionals as they try to find the best TB testing method and struggle with unexpected results.
A recent study involving 2,122 health care workers found that more than three in four positive conversions on TB blood tests (interferon gamma release assays, or IGRAs) reverted to negative when workers were retested six months later.1
Discrepancies in test results led the American College of Occupational and Environmental Medicine (ACOEM) to suggest that a higher threshold for positive results might be appropriate for the IGRAs. (See related story, p. 39)
The bottom line advice from experts if a health care worker has a positive TB test and no known exposure or risk factors: Use clinical judgment and consider retesting.
That strategy "is consistent with current CDC guidelines," says Thomas R. Navin, MD, chief of the Surveillance, Epidemiology & Outbreak Investigations Branch in CDC's Division of Tuberculosis Elimination.
Yet don't simply discount unexpected positive results, cautions Susan Dorman, MD, associate professor of medicine at Johns Hopkins University in Baltimore and lead author of the paper, which was supported by the CDC's Tuberculosis Epidemiologic Studies Consortium.
"One should not automatically assume that they're false," she told HEH. "One probably wants to go back and ask the question, 'Is there something that went on in our institution that could have led to TB transmission?'"
IGRAs better for some HCWs
When the first IGRA was introduced in 2001, it held great promise as a modern method of screening for latent tuberculosis. Despite the concerns about false positives, the IGRAs still have clear advantages over the century-old tuberculin skin test (TST), experts say.
The IGRAs are more specific and do not react to prior BCG vaccination, which is common in countries where TB is endemic. Many hospitals use IGRAs to screen health care workers who have previously received a BCG vaccine.
"We confirmed that the IGRA tests appeared to be more specific and did not cross react with BCG vaccine. So the tests are clearly helpful in that way," says Dorman.
Dorman and colleagues from medical centers in four states administered TSTs and IGRAs to the health care workers every six months for 18 months. About one in four (27%) of the health care workers who tested positive with a skin test at baseline were negative on both IGRAs. In fact, when TST-positive health care workers received a second skin test, more than half of them (54%) reverted to negative.
But the IGRAs — QuantiFERON-TB Gold In-Tube, produced by Cellestis, a Qiagen Company based in Valencia, CA; and T-SPOT. TB, produced by Oxford Immunotec, based in Abingdon, UK — showed high rates of reversion, as well. Some 57% of health care workers who tested positive with QuantiFERON at baseline and 64% of T-SPOT baseline positives reverted to negative.
Variable results continued throughout the study. Some 76% of those converting from negative to positive with QuantiFERON reverted to negative with further testing, as did 77% with a positive T-SPOT.
"We looked very hard to find relationships and understand why these conversions and reversions happened," says Dorman. "Despite intensive investigation, we really were not able to come up with any biological or medical or epidemiologic reason why these conversions and reversions happened."
Many of those conversions/reversions occurred when the positive result was near the cut point, she says.
"We concluded that most of these conversions were very likely to be false conversions," she says. Even retesting the same specimen or immediate retesting of an individual with a new specimen often led to a reversion to negative, Dorman says.
Is the cut point too low?
The Dorman study mirrors what many employee health professionals are finding in their daily practice and raises renewed questions about how to define a positive TB test result among low-risk health care workers.
Some of the problem stems from the potential variability in the way the tests are handled, processed or interpreted, she says. "What distinguishes these tests from other tests we do frequently in medical clinical practice is that these tests rely on a biological response," she says.
Technique also matters. For example, how vigorously the QuantiFERON tube is shaken can affect the results, she says. (The T-SPOT is currently processed in a central laboratory by Oxford, the manufacturer.)
Furthermore, the greatest number of false positives occur near the cut point — which calls into question the definition of a positive result, says Wendy Thanassi, MA, MD, National Project Lead for Tuberculosis with the Office of Public Health, Occupational Health at the U.S. Department of Veterans Affairs and chief of occupational health with the Palo Alto (CA) VA Health Care System.
The U.S. Food and Drug Administration set the QuantiFERON cut point at .35 IU/mL to achieve specificity of 99%, says Thanassi.
"What we see is a few percent of the total population testing right near the cutoff point," she says. "When we retest them, they're negative. Because they were always negative. There's just always maximum variability right around the cutoff."
Thanassi suggests setting a cut point of 1.1 IU/mL for low-risk health care worker screening. The results between .35 and 1.1 would be a "retest zone" in which employee health professionals should use professional judgment, she advises. "Retesting causes no harm. But chest x-rays, medications, consultations and fear cause a lot of harm," she says.
Limit screening programs
One other important way to reduce the likelihood of false positives, says Thanassi, is to follow CDC recommendations on risk assessment and conduct only baseline testing in low-risk settings. Inpatient settings with more than 200 beds are low risk if they have treated five or fewer TB patients in the past year. Those with less than 200 beds are low risk if they treated two or fewer TB patients.2
"The classification of low risk should be applied to settings in which persons with TB disease are not expected to be encountered, and, therefore, exposure to M. tuberculosis is unlikely," the guidelines state. "This classification should also be applied to HCWs who will never be exposed to persons with TB disease or to clinical specimens that might contain M. tuberculosis."
Over-testing of health care workers has led to the problem of false positives, says Thanassi. "This is an absolute creation of the fact that we have such a low prevalence for the disease that the false positive rate will be higher than the true positive rate. It's a mathematical law," she says.
At the same time, employee health professionals should be aware that health care workers may travel to countries where TB is endemic, says Amy J. Behrman, MD, FACP, FACOEM, medical director for Occupational Medicine at the University of Pennsylvania Health System. "At my institution, we have an increased number of global health programs, with doctors and nurses choosing to volunteer or take paid rotations in areas where TB is endemic," she says. "We're seeing more and more people who are choosing to spend months at a time in Haiti or sub-Saharan Africa."
Fewer tests, fewer dilemmas?
Hospitals that are able to restrict or even eliminate annual TB screening because of their low-risk status should still be prepared for the possibility that could change. Marshfield (WI) Clinic has typically been low-risk, rarely seeing even one TB patient in the entire system. But last year, three clinic facilities saw three TB patients, suddenly putting them in the medium risk category.
As Marshfield began testing 2,500 employees with an IGRA, an unexpected number of positives occurred. Bruce Cunha, RN, MS, COHN-S, manager of Employee Health and Safety, has been carefully evaluating the cases to determine if the employees could have had any potential exposure at work or elsewhere.
"If they don't have any risk factors when do we send somebody to infectious disease to have them evaluated for latent TB?" says Cunha. "I think the doctors are as frustrated as we are as to what to do with these patients."
Children's Hospital of The King's Daughters in Norfolk, VA, is a low-risk facility that once screened all employees with TB skin tests. The hospital switched to baseline testing only, initially using QuantiFERON. When a number of tests came out as indeterminate, the hospital switched again, to T-SPOT, says Occupational Health Director Patricia Higazi, RN, COHN.
Overall, Higazi has been happy with the change. By limiting the testing program, the hospital saves time and money, she says. And with the IGRAs, no one has to track down employees who forgot to come back to have their skin tests read, she says.
- Dorman SE, Belknap R, Graviss EA, et al. Interferon-gamma release assays and tuberculin skin testing for diagnosis of latent tuberculosis infection in healthcare workers in the United States. Am J Respir Crit Care Med 2014; 189:7787.
- Centers for Disease Control and Prevention. Guidelines for preventing the transmission of Mycobacterium tuberculosis in health-care settings, 2005. MMWR 2005; (RR)54:1-141.