The impact of sputum quality on Xpert positivity in active case-finding for TB.

BACKGROUND: Studies evaluating sputum quality and Xpert® MTB/RIF positivity in the context of active case finding are scarce. We aimed to determine whether sputum quality is associated with Xpert positivity and whether the association differed according to demographic and clinical characteristics. METHODS: A cross-sectional analysis using data from a mass screening programme in Brazilian prisons was conducted from 2017 to 2021. We administered a standardised questionnaire, obtained a chest X-ray and collected a spot sputum sample for Xpert testing. Sputum quality was classified as 'salivary', 'mucoid/mucopurulent' or 'blood-stained'. We used log binomial regressions to estimate the relationship between sputum quality and Xpert positivity, assessing interactions with participant characteristics. RESULTS: Among 4,368 participants for whom sputum quality was assessed, 957 (21.9%) produced salivary specimens, 3,379 (77.4%) had mucoid/mucopurulent sputum and 32 (0.7%) had blood-stained sputum. Xpert positivity was higher among those with mucoid/mucopurulent sputum than among those with salivary samples (12.0% vs. 3.7%). Mucopurulent sputum independently predicted Xpert positivity among individuals with and without symptoms, current smoking and abnormal chest radiographs on CAD4TB. CONCLUSIONS: In our study, sputum appearance independently predicted Xpert positivity, and could be used together with chest X-ray and symptom screening to inform use of Xpert in individual or pooled testing.