Comparison of Xpert MTB/RIF to Microscopy and Culture for the Diagnosis of Tuberculosis in a Referral Laboratory in Nepal

抄録

Sputum microscopy and Xpert MTB/RIF are the primary rapid diagnostic methods for tuberculosis (TB) in Nepal. However, disagreements among Xpert, microscopy, and culture, for example, cases that are Xpert positive and microscopy negative, are frequently observed in Nepal, including in our reference laboratory. The objective of this study was to compare the effectiveness of Xpert with that of culture and microscopy for the diagnosis of TB in Nepal. A total of 125 TB suspected sputum samples were processed for Xpert microscopy and culture. Comparison of the Xpert results to the culture results showed 100% sensitivity and 97.4% specificity, with excellent agreement (kappa coefficient = 0.96), whereas comparison of microscopy to culture showed 43.2% sensitivity and 98.7% specificity, with moderate agreement (kappa coefficient = 0.4). The sensitivity and specificity of microscopy, when compared with Xpert, were 43.5% and 100%, respectively. Importantly, the majority of the Xpert-positive samples with medium MTB detection and all samples with low and very low MTB detection were missed by microscopy. Our study showed that Xpert MTB/RIF is a reliable tool for the diagnosis and management of TB in Nepal. However, because of its high cost and lack of sustainability, alternative simple, rapid diagnostic methods with similar high efficiency would be helpful for controlling TB in Nepal.