LAUSR

Abstract The diagnostic accuracy of oral specimen nucleic acid amplification tests (NAATs) for pulmonary tuberculosis (PTB) remains controversial. We performed a systematic review according to Preferred Reporting Items for Systematic Reviews and Meta-Analyses guidelines, including studies that reported the diagnostic yield of NAATs in oral samples for PTB diagnosis. The pooled estimates, including those of sensitivity and specificity, were calculated, and a meta-regression was performed to investigate heterogeneity, which was determined using χ2 and I² tests. A total of 23 articles were included, and the pooled sensitivity, specificity, and area under the curve of NAATs in oral samples for PTB diagnosis were 50% (95% CI, 37%–63%), 97% (95% CI, 93%–99%), and 0.89 (95% CI, 86%–92%; I2 = 99%; chi-square, 169.61; P < .001), respectively. Our data demonstrated that NAATs using oral samples have a less satisfactory sensitivity and high specificity for PTB diagnosis. However, due to significant heterogeneity, such as heterogeneity in age, the results should be interpreted with caution.