Contribution of xpert MTB/RIF assay and Urine LF-LAM for the diagnosis of tuberculosis in children aged 5 – 14 years, at selected health facilities in Ethiopia, 2016 – 2019

Abstract

Background Childhood tuberculosis (TB) remains under-reported and undiagnosed. A full complement of diagnostic tests is oftentimes unavailable in resource limited country like Ethiopia. This study assesses the contribution of Xpert MTB/RIF assay and urine LF-LAM for childhood TB diagnosis using sputum and urine samples. Method A facility based cross-sectional study was conducted in children between 5 and 14 years of age. Sputum and urine samples were collected from children with presumptive TB. The samples were tested for TB using LF-LAM, Xpert MTB/RIF assay, concentrated smear microscopy, and culture. Diagnostic performance of Xpert MTB/RIF assay was analyzed and compared against culture, which was used as the gold standard. Urine LF-LAM test result was compared to a composite reference standard. Result Of 576 participants with presumptive TB enrolled in the study, 519 (90.1%) had complete clinical data and bacteriological laboratory test results. Active TB was diagnosed in 14.1% (73/519), and bacteriological confirmation was made in 10.1% (52/515) of children with presumptive TB. The odds of being diagnosed with a bacteriologically confirmed TB are significantly higher in children who have household contact history with TB patient (aOR 2.27, P = 0.03) and age above 10 years (aOR 3.67, P < 0.001). Xpert MTB/RIF test had sensitivity of 79% using culture as the gold standard. Compared to smear microscopy, the sensitivity of the Xpert MTB/RIF assay increased by 50% for children aged 5–9 years and by 40% for children and adolescents living with HIV (C/ALHIV). All bacteriologically confirmed (n = 2) and clinically diagnosed TB children (n = 2) who live with HIV were tested positive for urine LF-LAM. The overall sensitivity of urine LF-LAM was 27.6% when using the composite reference standard, compared to 17.9% when the bacteriological reference standard was applied. Conclusions Pulmonary TB diagnosis was greatly improved with the use of Xpert MTB/RIF assay, particularly in children aged 5–9 years and C/ALHIV who typically have difficulty producing good quality sputum. Urine LF-LAM performed well in children/adolescents who tested positive for HIV, but it performed poorly in the other variables, which suggests that urine LF-LAM testing did not play a critical role in TB diagnosis in children with negative HIV status.