Diagnostic accuracy of screening and diagnostic tests used in a state-wide tuberculosis prevalence survey in India

Prevalence surveys and active case findings for tuberculosis (TB) typically use symptoms and Chest X-ray for screening and a diagnostic test for microbiological confirmation. We report the performance of screening and diagnostic tests from a state-wide prevalence survey in India. We screened 130,932 individuals using chest X-ray (CXR) and symptom checklist. Sputum samples were obtained from eligible participants and subjected to Xpert/MTB RIF, smear microscopy, and culture. Cough over two weeks had higher sensitivity (41.6%, 95% CI: 31.6-52.1) and lower specificity (72.8%, 95% CI: 72.1-73.5) among all the symptoms. The sensitivity and specificity of abnormal CXR were 86.4% (95% CI:77.9-92.5) and 42.12% (95% CI, 41.3-42.8), respectively. Xpert/MTB RIF in the reference laboratory had the highest sensitivity (96.55% 95% CI: 88.0-99.5), and smear microscopy had the highest specificity (99.7% 95% CI: 99.6-99.8). The combined sensitivity and specificity of Xpert MTB/RIF along symptom screening and CXR were 97.9% (95% CI:92.6-99.7) and 4.9% (95% CI:4.67-5.33), respectively. The mobile van Xpert MTB/RIF was found to be highly specific (99.3%). We recommend a diagnostic algorithm consisting of symptom screening and CXR followed by WHO-recommended rapid molecular assays (mWRD) for future prevalence surveys and active case finding.