Department of Infection, University College London Medical School, Royal Free Hospital, Rowland Hill St., London, UK.
Clin Infect Dis. 2012 Nov;55(9):1171-8. doi: 10.1093/cid/cis631. Epub 2012 Jul 17.
BACKGROUND. There were 1.45 million deaths from tuberculosis in 2011. A substantial proportion of active pulmonary tuberculosis cases in countries where tuberculosis, human immunodeficiency virus (HIV) infection, and AIDS are highly endemic remain undiagnosed because of the reliance on sputum-smear microscopy. This study evaluated the performance of the Xpert MTB/RIF assay at a tertiary care referral center in Zambia, a country where the burden of tuberculosis and HIV infection is high. METHODS. A total of 881 adult inpatients admitted to University Teaching Hospital in Lusaka who were able to produce sputum were enrolled and analyzed in the study, irrespective of admission diagnosis. Sputum specimens were analyzed by fluorescence smear microscopy, the Xpert MTB/RIF assay, mycobacterial growth indicator tube (MGIT) culture,and MGIT drug-susceptibility testing. The sensitivity and specificity of the Xpert MTB/RIF assay were evaluated using culture as the gold standard. RESULTS. Culture-confirmed tuberculosis was found in 201 of 881 patients (22.8%). The specificity of the Xpert MTB/RIF assay was 95.0% (95% confidence interval [CI], 92.4%–96.8%),and the sensitivity was 86.1% (95% CI, 80.3%–90.4%). In sputum smear–negative, culture-positive cases, the assay was 74.7% sensitive (95% CI, 64.6%–82.8%), identifying 71 additional tuberculosis cases that were not detected by smear microscopy.A total of 18 of 111 patients with tuberculosis who were tested (16.2%) had multidrug-resistant (MDR) tuberculosis.The sensitivity and specificity of the Xpert MTB/RIF assay for detecting culture-confirmed, rifampicin-resistant tuberculosis was 81.3% (95% CI, 53.7%–95.0%) and 97.5% (95% CI,90.4%–99.6%), respectively. CONCLUSIONS. The Xpert MTB/RIF assay performs better than smear microscopy in an inpatient setting in a country where tuberculosis and HIV infection are highly endemic. Assessment of its usefulness and cost-effectiveness for increased detection of tuberculosis cases missed by sputum smear and for concomitant screening for MDR tuberculosis among adult inpatients attending tertiary care referral centers in other countries with a high burden of tuberculosis and HIV infection is warranted [corrected].
2011 年有 145 万人死于结核病。在结核病、人类免疫缺陷病毒（HIV）感染和艾滋病高度流行的国家，由于依赖痰涂片显微镜检查，大量活动性肺结核病例仍未得到诊断。本研究评估了 Xpert MTB/RIF 检测在赞比亚一家三级转诊中心的性能，该国结核病和 HIV 感染负担沉重。
共纳入并分析了 881 名入住卢萨卡大学教学医院的成年住院患者，无论入院诊断如何，只要能够产生痰液即可参加研究。痰标本通过荧光涂片显微镜检查、Xpert MTB/RIF 检测、分枝杆菌生长指示管（MGIT）培养和 MGIT 药敏试验进行分析。以培养为金标准评估 Xpert MTB/RIF 检测的灵敏度和特异性。
在 881 名患者中，有 201 名（22.8%）经培养证实患有结核病。Xpert MTB/RIF 检测的特异性为 95.0%（95%置信区间[CI]，92.4%–96.8%），灵敏度为 86.1%（95% CI，80.3%–90.4%）。在痰涂片阴性、培养阳性的病例中，该检测的灵敏度为 74.7%（95% CI，64.6%–82.8%），额外检出了 71 例未通过涂片显微镜检查检出的结核病病例。111 例结核病患者中有 18 例（16.2%）患有耐多药（MDR）结核病。Xpert MTB/RIF 检测对培养确诊、利福平耐药结核病的灵敏度和特异性分别为 81.3%（95% CI，53.7%–95.0%）和 97.5%（95% CI，90.4%–99.6%）。
在结核病和 HIV 感染高度流行的国家，Xpert MTB/RIF 检测在住院患者环境中的表现优于痰涂片显微镜检查。评估其在增加检测因痰涂片漏检的结核病病例以及在结核病和 HIV 感染负担沉重的其他国家的三级转诊中心对成年住院患者进行耐多药结核病同时筛查方面的有效性和成本效益是合理的。
