Abstract
BACKGROUND
Over half of childhood tuberculosis (TB) remains undiagnosed yearly. WHO recommends Xpert-Ultra as a first paediatric diagnosis test, but microbiological confirmation remains low. We aimed to determine the diagnostic performance of Xpert-Ultra on stool and urine in presumptive paediatric TB cases in two high-TB burden settings.
METHODS
This Médecins sans Frontières cross-sectional multicentric study took place at Simão Mendes hospital, Guinea-Bissau (July 2019 to April 2020) and in Malakal hospital, South Sudan (April 2021 to June 2023). Children 6 months to 15 years with presumptive TB underwent clinical and laboratory assessment, with one respiratory and/or extrapulmonary sample (gold standard (GS)), one stool and one urine specimen analysed with Xpert-Ultra.
RESULTS
A total of 563 children were enrolled in the study, 133 from Bissau, 400 from Malakal; 30 were excluded. Confirmation of TB was achieved in 75 (14.1%) while 248 (46.5%) had unconfirmed TB. Of 553 with GS specimen, the overall diagnostic yield was 12.4% (66/533). A total of 493 and 524 samples were used to evaluate Xpert-Ultra on stool and on urine, respectively. Compared to GS, sensitivity and specificity of Xpert-Ultra on stool were 62.5%(95%CI:49.4-74) and 98.3%(95%CI:96.7-99.2), whereas on urine were 13.9%(95%CI:7.5-24.3) and 99.4%(95%CI:98.1-99.8), respectively. Nine patients were positive on stool and/or urine but negative on GS.
CONCLUSIONS
Xpert-Ultra on stool showed moderate to high sensitivity and high specificity when compared to GS and an added diagnostic yield when GS was negative. Xpert-Ultra on stool was useful in extrapulmonary cases. Xpert-Ultra in urine showed low test performance.
