Jennifer Ho1, Phuong Thi Bich Nguyen2, Thu Anh Nguyen2, Khoa Hien Tran3, Son Van Nguyen3, Nhung Viet Nguyen4, Hoa Binh Nguyen5, Khanh Boi Luu2, Greg J Fox6, Guy B Marks7. 1. Woolcock Institute of Medical Research, University of Sydney, Sydney, NSW, Australia; South Western Sydney Clinical School, University of New South Wales, Sydney, NSW, Australia. Electronic address: jennifer.ho@sydney.edu.au. 2. Woolcock Institute of Medical Research, Hanoi, Vietnam. 3. Ca Mau Centre for Social Disease Prevention, Ca Mau, Vietnam. 4. National Tuberculosis Program, Hanoi, Vietnam. 5. National Tuberculosis Program, Hanoi, Vietnam; Centre for Operational Research, International Union Against Tuberculosis and Lung Disease, Paris, France. 6. Woolcock Institute of Medical Research, University of Sydney, Sydney, NSW, Australia; Central Clinical School, Sydney Medical School, University of Sydney, Sydney, NSW, Australia. 7. Woolcock Institute of Medical Research, University of Sydney, Sydney, NSW, Australia; South Western Sydney Clinical School, University of New South Wales, Sydney, NSW, Australia.
Abstract
BACKGROUND: Community-wide screening for tuberculosis with Xpert MTB/RIF as a primary screening tool overcomes some of the limitations of conventional screening. However, concerns exist about the low positive predictive value of this test in screening settings. We did a cross-sectional assessment of this diagnostic test to directly estimate the actual positive predictive value of Xpert MTB/RIF when used in the setting of community-wide screening for tuberculosis, and to draw an inference about the specificity of the test for tuberculosis detection. METHODS:Field staff visited households in 60 randomly selected villages in Ca Mau province, Vietnam. We included people aged 15 years or older who provided written informed consent and were able to produce 0·5 mL or more of sputum, irrespective of reported symptoms. Participants were tested with Xpert MTB/RIF, then those with positive results had two further sputum samples tested for smear microscopy and culture, and underwent chest radiography at the provincial TB Health Center. The positive predictive value of Xpert MTB/RIF was compared against two reference standards for tuberculosis diagnosis-a positive sputum culture for Mycobacterium tuberculosis, and a positive sputum culture or a chest radiograph consistent with active pulmonary tuberculosis. We then calculated the specificity of Xpert MTB/RIF for tuberculosis detection on the basis of these positive predictive values and disease prevalence in this setting. FINDINGS:43 435 adults consented to screening with Xpert MTB/RIF. Sputum samples of 0·5 mL or greater were collected from 23 202 participants, producing 22 673 valid results. 169 participants hadpositive Xpert MTB/RIF results (0·39% of those screened and 0·75% of those with valid sputum results). The positive predictive value of Xpert MTB/RIF was 61·0% (95% CI 52·8-68·7) when compared against a positive sputum culture and 83·9% (76·8-89·2) when compared against a positive sputum culture or chest radiograph consistent with active tuberculosis. On the basis of these positive predictive values, the specificity of Xpert MTB/RIF was determined to be between 99·78% (95% CI 99·71-99·84) and 99·93% (99·88-99·96). INTERPRETATION: The positive predictive value and specificity of Xpert MTB/RIF in the context of community-wide screening for tuberculosis is substantially higher than that predicted in previous studies. Our findings support the potential role of Xpert MTB/RIF as a primary screening tool to detect prevalent cases of tuberculosis in the community. FUNDING: Australian National Health and Medical Research Council.
RCT Entities:
BACKGROUND: Community-wide screening for tuberculosis with Xpert MTB/RIF as a primary screening tool overcomes some of the limitations of conventional screening. However, concerns exist about the low positive predictive value of this test in screening settings. We did a cross-sectional assessment of this diagnostic test to directly estimate the actual positive predictive value of Xpert MTB/RIF when used in the setting of community-wide screening for tuberculosis, and to draw an inference about the specificity of the test for tuberculosis detection. METHODS: Field staff visited households in 60 randomly selected villages in Ca Mau province, Vietnam. We included people aged 15 years or older who provided written informed consent and were able to produce 0·5 mL or more of sputum, irrespective of reported symptoms. Participants were tested with Xpert MTB/RIF, then those with positive results had two further sputum samples tested for smear microscopy and culture, and underwent chest radiography at the provincial TB Health Center. The positive predictive value of Xpert MTB/RIF was compared against two reference standards for tuberculosis diagnosis-a positive sputum culture for Mycobacterium tuberculosis, and a positive sputum culture or a chest radiograph consistent with active pulmonary tuberculosis. We then calculated the specificity of Xpert MTB/RIF for tuberculosis detection on the basis of these positive predictive values and disease prevalence in this setting. FINDINGS: 43 435 adults consented to screening with Xpert MTB/RIF. Sputum samples of 0·5 mL or greater were collected from 23 202 participants, producing 22 673 valid results. 169 participants had positive Xpert MTB/RIF results (0·39% of those screened and 0·75% of those with valid sputum results). The positive predictive value of Xpert MTB/RIF was 61·0% (95% CI 52·8-68·7) when compared against a positive sputum culture and 83·9% (76·8-89·2) when compared against a positive sputum culture or chest radiograph consistent with active tuberculosis. On the basis of these positive predictive values, the specificity of Xpert MTB/RIF was determined to be between 99·78% (95% CI 99·71-99·84) and 99·93% (99·88-99·96). INTERPRETATION: The positive predictive value and specificity of Xpert MTB/RIF in the context of community-wide screening for tuberculosis is substantially higher than that predicted in previous studies. Our findings support the potential role of Xpert MTB/RIF as a primary screening tool to detect prevalent cases of tuberculosis in the community. FUNDING: Australian National Health and Medical Research Council.
Authors: David J Horne; Mikashmi Kohli; Jerry S Zifodya; Ian Schiller; Nandini Dendukuri; Deanna Tollefson; Samuel G Schumacher; Eleanor A Ochodo; Madhukar Pai; Karen R Steingart Journal: Cochrane Database Syst Rev Date: 2019-06-07
Authors: Adrienne E Shapiro; Jennifer M Ross; Mandy Yao; Ian Schiller; Mikashmi Kohli; Nandini Dendukuri; Karen R Steingart; David J Horne Journal: Cochrane Database Syst Rev Date: 2021-03-23