PERFORMANCE OF FUJILAM VERSION 1 FOR THE DIAGNOSIS OF TUBERCULOSIS IN CHILDREN UNDER 5 YEARS IN UGANDA - A BRIEF REPORT.

Background: Diagnosing tuberculosis in children remains challenging due to the paucibacillary nature of the disease, non-specific symptoms, and difficulty collecting samples. New tests are urgently needed. We evaluated the diagnostic accuracy of FujiLAM version 1 urine test in children under 5 years with presumptive tuberculosis in Uganda. Methods: This was a cross-sectional evaluation nested in the NOD-pedFEND diagnostic study. Children under 5 years with signs or symptoms of tuberculosis were recruited from three hospitals in Uganda. All participants underwent comprehensive baseline investigations, including culture and Xpert MTB/RIF Ultra on reference samples and chest X-ray. Culture and Xpert MTB/RIF Ultra results were used to define a microbiological reference standard. FujiLAM version 1 testing was performed on fresh urine at the baseline visit. Results: Seventy-nine children were included in the study. Fifteen participants (19%) were classified as having microbiologically confirmed TB, 44 (56%) as unconfirmed TB, thirteen (16%) as unlikely TB, and seven participants (9%) were unclassifiable. Only one participant (with unlikely TB) had a positive FujiLAM test. Culture and Xpert MTB/RIF Ultra on reference samples were negative in this child, indicating poor diagnostic accuracy with a sensitivity of 0% (95% CI: 0 - 21.8) of FujiLAM version 1 against the microbiological reference standard. Conclusion: The poor diagnostic accuracy of FujiLAM version 1 in children under five years of age makes it unsuitable as a diagnostic test in this age group. Further evaluation with the optimized version 2 of this test is needed.