Diagnostic accuracy of the updated FUJIFILM SILVAMP TB LAM II assay for pulmonary TB among HIV-negative outpatients at a primary-level clinic in rural China

Objectives Non-sputum-based diagnostic tests are a priority for tuberculosis (TB) control. However, studies regarding urine lipoarabinomannan (LAM) tests in HIV-negative population remain limited. This study aimed to evaluate the accuracy of FujiLAM II for pulmonary TB among HIV-negative outpatients in primary-level healthcare settings. Methods A prospective cohort study was conducted among outpatients with presumptive pulmonary TB and healthy controls in rural China. Urine samples from all participants were tested by FujiLAM II. Diagnostic accuracy was assessed against liquid culture, molecular test, microbiological reference standard (MRS) and composite reference standard (CRS). Results 408 presumptive pulmonary TB patients and 144 healthy controls were included, which were entirely HIV-negative. Among presumptive patients, 125 (30.6%) were MRS-positive and 149 (36.5%) were CRS-positive. All healthy controls were LAM-negative. Using liquid culture as the reference standard, FujiLAM II had a sensitivity of 35.8% (95% CI 26.4-46.3) and a specificity of 94.2% (95% CI 90.9-96.5). Against molecular assay, the sensitivity and specificity were 30.3% (95% CI 22.3–39.5) and 94.5% (95% CI 91.0–96.7). The sensitivities were 29.6% (95% CI 21.9-38.5) and 29.5% (95% CI 22.5-37.6) against MRS and CRS, respectively, with the corresponding specificities of 94.7% (95% CI 91.2-96.9) and 96.9% (95% CI 93.8-98.6). LAM-positive cases had a significantly shorter median liquid-culture time to positivity (9.5 vs 14.8 days, P = 0.002) and were more frequently smear-positive (35.3% vs 13.3%, P = 0.018) than LAM-negative cases, suggesting that bacillary burden may influenced FujiLAM II results. Conclusions FujiLAM II could serve as a complementary tool with high specificity but limited sensitivity for pulmonary TB diagnosis among HIV-negative outpatients in primary healthcare settings.