Why TB Blood Tests Give Mixed Results in Low-Risk Communities

Interpreting blood tests for tuberculosis can be surprisingly complicated, even in places where TB is rare. Tests called interferon-γ release assays (IGRAs) flag people who may have latent tuberculosis infection (LTBI) when they are IGRA positive (IGRA+). Many asymptomatic IGRA+ people are labeled as presumed LTBI and are offered treatment to reduce the small risk of developing active disease. But when these blood tests are repeated or used to confirm an initial result, they sometimes disagree with themselves. Those discordant results — a positive becoming negative or a negative becoming positive — make it hard for doctors to know who truly needs preventive therapy and who does not. To shed light on this problem, Charles Kyriakos Vorkas and colleagues carried out a retrospective review of QuantiFERON-TB Gold Plus (QFTTB) testing at Stony Brook Medicine from October 2020 through March 2024. They aimed to identify demographic and clinical factors tied to varying QFTTB values and to describe how common inconsistent, or discordant, results were in this low-endemicity setting.

The study examined 11,641 QFTTB tests ordered during the study period and analyzed 743 subjects in detail. Of those tests, 436 were QFTTB-positive (3.7%), and 16 people were diagnosed with active TB. For comparison, the team included a random sample of 307 age- and sex-matched QFTTB-negative controls. The investigators looked at clinical and demographic variables including age, sex, race, comorbidities, and medication use, and they analyzed quantitative QFTTB values including NIL, TB1, and TB2. Among 203 people who had serial QFTTB testing, 170 (83.7%) had concordant results on repeat testing, while 33 (16.3%) had discordant results. Of the discordant group, 23 (69.7%) showed reversion (positive to negative) and 10 (30.3%) showed conversion (negative to positive). Conversions occurred in significantly older individuals (mean age 51.1 ± 15.0 versus 37.0 ± 15.6, p = 0.025) and over longer testing intervals (415.1 versus 91.2 days, p = 0.026). The study also found that comorbidities such as cardiovascular disease, infections, and diabetes correlated with changes in NIL, TB1, and TB2 values.

These findings matter because inconsistency in QFTTB readings can directly affect patient care. In a low-incidence region, the pre-test probability that a positive IGRA indicates true latent infection is lower, so a single positive result may be less reliable without supporting clinical risk. The study emphasizes the value of confirmatory testing and of combining QFTTB results with epidemiologic risk factors and a patient’s clinical history before starting LTBI therapy. It also shows that age, the time between tests, and common health conditions can influence the numerical readouts (NIL, TB1, TB2) that underlie the test result. For clinicians, public health workers, and researchers — including those designing TB vaccine and biomarker studies — a clearer grasp of IGRA variability will help avoid unnecessary treatment, improve patient counseling, and strengthen study design where accurate measurement of infection status is critical.