Mind the Clinic-Community Gap: Re-evaluation of Test Performance and False Positive Results in Community-Wide Tuberculosis Screening

Abstract

Community-wide screening for bacteriologically confirmed pulmonary tuberculosis may reduce the tuberculosis burden, although concerns of overtreatment remain because of false positive diagnoses due to subpar specificity of current bacteriological tests for screening. Our review and data analysis show that clinic-based test specificity estimates of Xpert against culture underestimate performance in communities, for both Xpert MTB/RIF(99.8% for community vs 98.4% for clinic) and Xpert Ultra (99.4% vs 95.6%, respectively), reducing the presumed false positivity of sputum Xpert, using culture as the reference, by 86.8% and 85.4%, respectively, compared with clinic-based specificity estimates. These findings support large-scale evaluation of community-wide screening for tuberculosis.