Tuberculin and QuantiFERON-TB-Gold tests for latent tuberculosis: A meta-analysis

Abstract

Background Up to now, there has been no universal consensus on the agreement between the tuberculin skin test (TST) and the QuantiFERON-TB-Gold test (QFT) in the detection of latent tuberculosis infection (LTBI) among high-risk populations. Aims To estimate the agreement between TST and QFT among health care workers (HCWs). Methods A meta-analysis in which all major electronic databases, including Medline, Scopus, Web of Sciences and Ovid, were searched until June 2014. All cross-sectional and cohort studies addressing the agreement between TST and the QFT were included. The extracted data were analysed and the results were reported using random effect models. Results The overall kappa statistic between TST and the QFT was 0.27 [95% confidence interval (CI) 0.22, 0.32] and the adjusted kappa statistic for prevalence and bias was 0.41 (95% CI 0.32, 0.50). The kappa for subjects with and without bacillus Calmette-Guérin (BCG) vaccination was 0.27 (95% CI 0.18, 0.36) and 0.31 (95% CI 0.15, 0.46) respectively. The figures were 0.30 (95% CI 0.16, 0.43) and 0.82 (95% CI 0.74, 0.90) for prevalence-adjusted and bias-adjusted kappa, respectively. Conclusions The overall agreement between TST and QFT in the detection of LTBI among HCWs was poor. After adjusting for the prevalence and bias indices, kappa statistics reached fair agreement. The utility of each of these two tests is dependent on the prevalence and burden of tuberculosis as well as the BCG vaccination status.