Screening for latent infection with Mycobacterium tuberculosis: a plea for targeted testing in low endemic regions.

The higher specificity proposed for IFN-γ-release assays (IGRAs) compared with tuberculin skin testing (TST) may indicate superiority in identifying patients at risk for progression towards active tuberculosis. Unfortunately, available assays may frequently show discordant results. To aid in interpreting discordant results, the article under evaluation has comparatively analyzed results of the TST and two commercially available IGRAs in a low-risk population for tuberculosis. TST-positive/IGRA-negative results were strongly associated with bacillus Calmette-Guérin vaccination or nontuberculous mycobacteria sensitization, whereas the etiology of TST-negative/IGRA-positive results or discrepancies among IGRAs remain unresolved. As the three tests identified different individuals in the majority of positive cases, this indicates that most positive results obtained in the absence of additional risk factors are falsely positive. By contrast, increasing concordance of positive results was associated with tuberculosis risk factors. This has direct implications for screening practices in low prevalence regions where a targeted screening approach may reduce both unnecessary testing and treatment.