Background and objective: We conducted a systematic review and meta-analysis to compare the accuracy of the interferon-gamma release assays (IGRAs) and the tuberculin skin test (TST) for the diagnosis of Mycobacterium tuberculosis infection.
Methods: We systematically searched PubMed, Embase, Cochrane library, and Web of Science databases for relevant published studies in recent decades and calculated pooled estimated sensitivities, specificities, DOR, and SROC curve of the QFT-IT, T-SPOT and TST. Random-effects models were used to assess estimates from studies with significant heterogeneity. Moreover, area under the curve was used to evaluate the accuracy of the tests.
Results: Overall, 9 studies for QFT-IT, 12 studies for T-SPOT, and 16 studies for TST involving 3586 participants were included in this analysis. We found that sensitivities of the QFT-IT, T-SPOT, and TST were respectively 0.842 (95 % CI 0.811-0.870), 0.840 (95 % CI 0.814-0.864), and 0.665 (CI 0.635-0.693); specificities were respectively 0.745 (95 % CI 0.715-0.775), 0.658 (95 % CI 0.621-0.693), and 0.633 (CI 0.605-0.661); positive likelihood ratios were respectively 3.652 (95 % CI 2.180-6.117), 2.196 (95 % CI 1.727-2.794), and 1.825 (95 % CI 1.351-2.464); negative likelihood ratios were respectively 0.212 (95 % CI 0.109-0.414), 0.246 (95 % CI 0.161-0.377), and 0.556 (95 % CI 0.385-0.804); the SROC curves were 19.205, 10.397, and 3.810.
Conclusions: The two IGRAs showed better performance than TST for the diagnosis of the tuberculosis. However, neither of them showed stability in the diagnosis of TB.
Keywords: Diagnosis; Interferon-gamma release assays; Specificity; Tuberculin skin test; Tuberculosis.