Use of Xpert MTB/RIF Ultra assay on stool and gastric aspirate samples to diagnose pulmonary tuberculosis in children in a high-tuberculosis-burden but resource-limited area of China

Lin Sun,Yang Liu,Min Fang,Yu Chen,Yu Zhu,Cheng Xia,Jihang Jia,Shuting Quan,Yacui Wang,Xue Tian,Yan Shi,Li Duan,Xiaomei Shi,Qiong Liao,Chaomin Wan,Adong Shen
DOI: https://doi.org/10.1016/j.ijid.2021.11.012
IF: 12.073
2022-01-01
International Journal of Infectious Diseases
Abstract:OBJECTIVES: Our study analyzed the performance of Xpert MTB/RIF Ultra (Ultra) on stool and gastric aspirate (GA) samples for the diagnosis of pediatric pulmonary tuberculosis (TB) caused by Mycobacterium tuberculosis in a high-burden area of China.METHODS: Children with presumptive TB were enrolled in two hospitals in Sichuan Province (July 2019-Oct 2020). Because of the unavailability of sputum for etiological testing, gastric samples were aspirated and tested by bacterial culture, acid-fast bacillus microscopy, and Ultra. Stool samples were tested simultaneously using Ultra and Xpert.RESULTS: In total, 141 children with active TB and 34 with non-TB respiratory tract infections were enrolled. Ultra-stool (60.3%, 85/141) and Ultra-GA (52.5%, 74/141) tests were similarly sensitive (p = 0.187). Among the subset of 48 children with confirmed TB, Ultra testing was equally sensitive on stool and GA samples (85.4%, 41/48). The agreement between Ultra-stool and Ultra-GA was moderate in children with active TB (kappa value = 0.527). After integrating Ultra-GA and Ultra-stool outcomes, 70.9% (100/141) of the children were considered to have confirmed TB. The specificities of Ultra-stool and Ultra-GA were 97.1% (33/34) and 100% (34/34), respectively (p = 0.314).CONCLUSIONS: In children, stools can be used as alternative samples to GAs for Ultra tests. Stool- and GA-based Ultra tests are appropriate for bacteriological TB confirmation.
infectious diseases
What problem does this paper attempt to address?