A multicountry evaluation of the Xpert MTB/RIF assay for the diagnosis of intrathoracic tuberculosis in children using alternative specimens (nasopharyngeal aspirate and stool): A prospective cohort study conducted in Madagascar, Ivory Coast and Cameroon (TB KIDS project)

Objectives: Tuberculosis (TB) diagnosis in children remains challenging due to the paucibacillary nature of specimens and the difficulty in obtaining suitable samples. The use of alternative samples like nasopharyngeal aspirate (NPA) and stools, alongside Xpert MTB/RIF testing, offers promising improvements. This study aimed to assess the diagnostic performance of the Xpert MTB/RIF test on NPA and stool samples for detecting intrathoracic TB in children from Madagascar, Cameroon, and Ivory Coast. Methods: Children under 15 years with suspected intrathoracic TB were enrolled in hospitals in these countries’ capitals. Samples for analysis included standard specimens (gastric aspirate or sputum), NPA, stools, with additional HIV serology, tuberculin skin test tests, and chest X-rays. We used a composite reference standard to estimate the accuracy of the Xpert MTB/RIF test with alternative samples. Results: Of 1146 children analyzed, the sensitivity of Xpert MTB/RIF was 58.3% for NPA and 45.5% for stool samples, with a high specificity of more than 95%. The diagnostic performance of Xpert MTB/RIF with alternative samples did not differ according to age group or HIV status. Conclusions: The findings support the World Health Organization's recommendation for using Xpert MTB/RIF with alternative samples in childhood TB diagnosis, underscoring its utility across different settings and HIV statuses.