Evaluating Tuberculosis Screening Tools: A Decade-Long Network Meta-Analysis of Sensitivity and Specificity with...

Tuberculosis (TB) remains a significant global health challenge despite numerous innovations in diagnostic methods. Accurate early identification is essential for effective TB management and prevention of disease spread.

This study combined bibliometrics and network meta-analysis (NMA) to evaluate publication trends and compare the sensitivity and specificity of various TB screening tools against the gold standard, culture. Literature searches were conducted in seven databases over the period 2014-2024. Data were analyzed using a random-effects model with adjustment for quality of evidence using Confidence in NMA.

Of the 22 studies that met the inclusion criteria, 13 diagnostic tools were compared in 34 direct and 44 indirect comparisons. Culture remained the method with the best accuracy. TB molecular bacterial load assay (TB-MBLA), mNGS, and RT-Mycobacterium tuberculosis (RT-MTB) performed closest to culture in terms of sensitivity and specificity, making them promising candidates for TB screening. No significant incoherence was found, but there was high inconsistency and heterogeneity.

This study highlights that molecular methods provide rapid detection and superior diagnostic performance compared to conventional methods, while also necessitating consideration of specificity factors and the risk of overdiagnosis. These findings highlight the urgent need to prioritize affordable, rapid, and scalable TB screening tools such as TB-MBLA and RT-MTB in low-resource, high-burden settings to improve early detection and reduce transmission.