Setting: Lymphadenitis is the most common extra-pulmonary manifestation of tuberculosis (TB). Conventional diagnostic methods such as smear microscopy and culture are not very accurate for TB lymphadenitis. Nucleic acid amplification tests (NAAT) may offer additional diagnostic benefit.
Objective: To assess, in a systematic review, the performance of NAAT for the diagnosis of tuberculous lymphadenitis.
Design: We performed searches of the literature and identified 36 articles containing 49 comparisons between NAAT and a reference standard for TB lymphadenitis. Sensitivity and specificity estimates from each study were displayed in forest plots and summary receiver operating characteristic (SROC) plots.
Results: Overall study quality was fair, but the quality of reporting was poor in many studies. Estimates of sensitivity and specificity of NAAT were highly heterogeneous across studies, possibly due to variations in populations, study quality and test techniques. Estimates of sensitivity varied between 2% and 100%, and specificity estimates varied between 28% and 100%. Commercial NAAT assays, assays that used more than 20 mul of template and reports containing discrepant analysis provided significantly higher diagnostic accuracy. Blinding, template volume and discrepant analysis may account for some of the observed heterogeneity.
Conclusion: Studies on NAAT for TB lymphadenitis produce highly variable and inconsistent results, precluding the determination of clinically meaningful estimates of accuracy. Study reports are not well standardised and often do not contain enough information. Because both false-positive and false-negative results are possible, NAATs will need to be applied in conjunction with conventional methods and interpreted in the context of clinical suspicion.