More than half of repeat blood tests for tuberculosis infection may have been unnecessary, while nearly half of patients who were not retested may have met criteria for repeat testing, according to a study of more than 65,000 patients at U.S. community health centers.
The findings, published in Open Forum Infectious Diseases, suggest that repeat interferon-gamma release assay testing could be better targeted to patients with weakly positive initial results who are at low risk for tuberculosis infection.
“This project highlights both the need for better diagnostic stewardship surrounding TB infection testing, as well as some of the limitations of current tests for TB infection in general,” first author Jeffrey I. Campbell, MD, MPH, of the Section of Pediatric Infectious Diseases at Boston Medical Center, told IDSA Reporter.
Dr. Campbell and colleagues conducted a retrospective study using electronic health records from OCHIN, a network of about 1,200 health care sites across 40 states. They identified 65,728 patients who had a positive IGRA result between 2012 and 2024. They examined how often patients underwent repeat testing, which factors were associated with a positive result changing to negative (reversion), and whether repeat testing followed published recommendations.
Repeat testing and reversion
Of the 65,728 patients, 8,978 (14%) underwent repeat testing. Among 8,174 second tests with sufficient data, 6,176 (76%) remained positive, 1,904 (23%) changed to negative and 94 (1%) could not be interpreted.
The strength of the initial positive result was the strongest predictor of reversion. The estimated likelihood of a repeat test turning negative was 53% for weakly positive results, 31% for moderately positive results and 7% for strongly positive results.
Overall, 82% of patients whose results changed from positive to negative had initial antigen-minus-nil values below 1.00 IU/mL. When the investigators used a stricter definition that required a change from clearly positive to clearly negative, only 7% of tests reverted.
Children younger than five years and patients with HIV infection had higher odds of a repeat test turning negative. Patients born in countries with high TB rates or who preferred a language other than English had lower odds.
An analysis of consecutive test pairs also found that patients who received corticosteroids between tests were more likely to have a positive result change to negative.
Opportunities to improve testing
Based on published recommendations, the investigators classified 40% of 10,851 repeat-test pairs as likely appropriate, 9% as possibly appropriate and 51% as probably unnecessary.
About 51% of repeat tests had less than a 5% predicted chance of turning negative, suggesting they were unlikely to affect clinical decisions.
Among 53,618 patients who did not undergo repeat testing and had sufficient data for assessment, 25,473 (48%) met criteria suggesting that repeat testing was likely or possibly appropriate.
The investigators noted several limitations, including a lack of information about why tests were ordered, tests performed outside the network, and reliance on electronic health records. Country of birth was often missing, and the analysis focused mainly on QuantiFERON tests.
Dr. Campbell and colleagues concluded that repeat testing should better follow published recommendations and be performed “when it is most likely to change care.”
The study received funding from the Robert Wood Johnson Foundation, Boston University Clinical and Translational Science Institute and National Institute of Allergy and Infectious Diseases, with additional research infrastructure support from the Patient-Centered Outcomes Research Institute. The authors reported no conflicts of interest.
Author perspective
Dr. Campbell offered additional perspective on the findings.
Based on your findings, when should clinicians repeat IGRA after an initial positive result?
Dr. Campbell: The main reason for repeating an IGRA is to “confirm” the initial positive result before recommending treatment for TB infection. Because there is no perfect test for TB infection, a second test is most likely to be useful if the patient has a low baseline likelihood of having TB infection (i.e., that the initial test was most likely to be falsely positive). For instance, patients who have a positive IGRA despite not having any clear risk factors for TB infection could be candidates for having a repeat test; if the second test is negative, and the pre-test probability of TB infection is low, then the patient might be able to avoid taking preventive therapy. What our findings add is that if the first test was “strongly” positive (i.e., that there was a strong specific immune response to TB antigens), then the second test is very likely to also be positive. Obtaining a second test for these patients probably won’t provide additional information to the clinician beyond what the first test result showed, but could add time and expense to the patient’s TB care.
How should clinicians interpret a repeat IGRA that turns negative, particularly in patients with HIV or those receiving corticosteroids?
Dr. Campbell: For patients who are immunocompromised, such as from HIV or long-term steroid use, the negative test should be viewed with caution, for two main reasons. First, the IGRA test is designed to test immune memory to TB. For patients with weakened immune systems, the test may not be able to adequately measure TB infection, which could result in a falsely negative test (and leading to the pattern of a positive test followed by a negative test). Second, these patients experience much higher rates of progression from TB infection to TB disease than patients with fully intact immune systems. Thus, the consequences of missing a true infection are more consequential. In my view, we shouldn’t be seeking to confirm positive tests in these patients with additional IGRAs, but rather use repeat tests primarily for patients at low risk of TB infection and TB progression.
What practical steps can health care organizations take to reduce unnecessary repeat IGRA testing while ensuring that patients who may benefit from retesting are not overlooked?
Dr. Campbell: Probably the main intervention a health system could do to mitigate overuse of repeat testing after a positive result is to minimize testing of low-risk populations in the first place. There are several TB risk factor screeners in widespread use for low TB incidence settings. These aren’t perfect, but developing reminders or prompts to limit testing in patients without these risk factors could avoid scenarios in which a clinician is trying to decide what to do with a positive test that they don’t really believe. Once a positive test has resulted, I think this is a problem in which electronic health record prompts or alerts could be very helpful to avoid low-value repeat testing. In clinical practice, my sense is that repeat tests are often performed because clinicians are not aware of previous positive tests. A simple flag in the health record alerting the clinician to the prior positive result, accompanied by a brief explanation of the likelihood of reversion and risks for TB infection and progression, might avoid many of the serial tests that are performed in busy clinical settings. Right now, electronic health records can be extremely complex to navigate, and finding prior results is often harder than simply ordering a test again.