Abstract
BACKGROUND: Current sputum-based diagnostic approaches fail to identify some individuals with active TB, underscoring the need for tests that do not rely on sputum. This study assessed the diagnostic accuracy of the Xpert MTB Host Response (Xpert HR) blood test for detecting pulmonary TB. METHODS: We conducted a prospective study in adults with pulmonary TB symptoms who underwent Xpert HR testing. Diagnostic accuracy was assessed against a microbiological reference standard of sputum cultures and molecular testing. RESULTS: Among 813 participants, 52% were female, median age was 38 years, and 192 (24%) were HIV-positive. For Xpert HR performed using capillary blood, sensitivity was 91% (95% confidence interval [CI]: 86-94), specificity was 45% (95% CI: 41-49), and area under the receiver operating characteristic curve was 0.85 (95% CI: 0.81-0.89). Country-specific differences in performance were observed. Sensitivity was 93% (95% CI: 76-98) and specificity was 18% (95% CI: 13-27) among people living with HIV. CONCLUSION: The Xpert HR TB assay demonstrated promising sensitivity, but its specificity did not meet minimum target product profiles for a TB triage or case detection test. In its current form, Xpert HR cannot replace more sensitive nucleic acid amplification tests in sputum-productive individuals.