Performance of a rapid strip test for the serologic diagnosis of latent tuberculosis in children.

Abstract

The serodiagnostic tests for tuberculosis (TB) present a high variability in terms of sensitivity and specificity. Data on patients with latent TB infection (LTBI) and children in high prevalence settings are still limited. The present study aimed to evaluate an in-house strip test for detection of anti-M. tuberculosis antibodies in TB patients, mostly children aged under 15 y, grouped into four diagnostic categories: active TB, LTBI, healthy TB contacts, and other non-TB diseases. The diagnostic performance of strip test was compared with the tuberculin skin test (TST) and interferon-gamma release assay (IGRA). Sensitivity and specificity were assessed for all three diagnostic tests. The detection accuracy among the tests was calculated by using a receiver operating characteristic analysis. TST and IGRA could diagnose the active TB cases correctly (100%). The sensitivity of strip test for active TB was 58.3% and 37.5% for LTBI, while the sensitivities of TST and IGRA for LTBI were 90.3% and 37.5%, respectively. The overall specificities of strip test and IGRA were 91.5% and 95.7%, respectively, which were superior to that of TST (68.1%). The strip test did not appear to be useful for diagnosis of active TB in comparison with the current diagnostic standard. The assay may be particularly significant in situations where TB is clinically difficult to diagnose like LTBI and could be a meaningful tool in terms of high specificity and simplicity for ruling in pediatric TB in countries with high TB infection rate. Further studies are needed to determine whether strip test can be improved in its sensitivity and should be implemented into routine clinical practice.