Xpert MTB/RIF Ultra on nasopharyngeal swabs has lower sensitivity for pulmonary tuberculosis than tongue swabs: a nested case-control study in South African primary care

BACKGROUND: Non-sputum alternatives for tuberculosis (TB) diagnosis are a priority. Nasopharyngeal swabs (NPSs), routinely used for viral testing, may have promise. Comparative performance to tongue swabs (TSs), which are recommended for TB detection but have suboptimal sensitivity, is unknown. METHODS: We evaluated Xpert MTB/RIF Ultra (Ultra; Cepheid, USA) on stored wet or dry NPSs (NPSU) from adults with presumptive TB consecutively presenting to primary care in Cape Town, South Africa. Using a nested case-control design, all TB-positive participants had NPSU (n=148) and random selection of TB-negative participants frequency-matched by HIV-status (n=117). We used a microbiological reference standard (MRS) and extended MRS (eMRS). A consecutive subset participant (n=130) also had TSs collected and of these, 62 TB-positive and 68 TB-negative people had TS Ultra (TSU) done. RESULTS: NPSU sensitivity was 35% (95% CI 27, 43) and specificity 88% (80, 93). Wet swabs had higher sensitivity than dry swabs [45% (34, 57) vs 21% (12, 34)]. Overall sensitivity was lower in HIV-positive participants [24% (14, 36) vs 44% (32, 56)]. TSU had higher sensitivity [67% (53, 79); p<0.001] than NPSU in head-to-head comparisons. Specificities were comparable. CONCLUSION: NPSU sensitivity was lower than TSU, especially among dry NP swabs, and neither NPSU sensitivity nor specificity met WHO benchmarks for non-sputum confirmatory tests. NPSs do not appear useful for TB diagnosis without further optimisation.