The Tuberculin Skin Test Is Unreliable in School Children BCG-vaccinated in Infancy and at Low Risk of Tuberculosis Infection

Background: The tuberculin skin test (TST) is often used to screen for latent tuberculosis infection (LTBI) in school children, many of whom were bacille Calmette-Guerin (BCG)-vaccinated in infancy. The reliability of the TST in such children is unknown. Methods: TSTs performed in low-risk BCG-vaccinated and -nonvaccinated grade 1 and grade 6 First Nations (North American Indian) school children in the province of Alberta, Canada, were evaluated retrospectively. To further assess the specificity of the TST, BCG-vaccinated children with a positive TST (>= 10 mm of induration) and no treatment of LTBI were administered a QuantiFERON-TB Gold In-Tube test (QFT-GIT, Cellestis International). Results: A total of 3996 children, 2063 (51.6%) BCG-vaccinated and 1933 (48.4%) BCG-nonvaccinated, were screened for LTBI. Vaccinated children were more likely than nonvaccinated children to be TST positive (5.7% vs. 0.2%, P < 0.001). Vaccinated children with a positive TST were more likely to have a recent past TST as compared with those with a negative TST (6.8% versus 2.8%, P = 0.01). Among 65 BCG-vaccinated TST-positive children who underwent a QFT-GIT, only 5 (7.7%; 95% CI: 2.5%, 17.0%) were QFT-GIT positive. A TST of >= 15 mm was more likely to be associated with a positive QFT-GIT than a TST of 10 to 14 mm, 16.0% (95% CI: 4.5%, 36.1%) versus 2.5% (95% CI: 0.1%, 13.2%), P = 0.047. Conclusion: The TST is unreliable in school children, BCG-vaccinated in infancy, and who are at low risk of infection. The QFT-GIT is a useful confirmatory test for LTBI in BCG-vaccinated TST-positive school children.