Evidence for Tuberculosis in Individuals With Xpert Ultra “Trace” Sputum During Screening of High-Burden Communities

Abstract

Background. “Trace” results on Xpert MTB/RIF Ultra (“Ultra”; Cepheid) —a molecular diagnostic test for tuberculosis (TB)—are often interpreted as an indication for TB treatment, but may also represent detection of nonviable bacilli or analytical error. In community-screening settings where individual TB risk is low, there is limited guidance on how to interpret Ultra-trace results. Methods. We conducted systematic Ultra TB screening of adults and adolescents (≥15 years) in Kampala, Uganda, through door-to-door and event-based sputum collection. We enrolled individuals with trace-positive sputum for detailed clinical, radiographic, and microbiological (including 2 sputum cultures, repeat Ultra, and for people with HIV, urine lipoarabinomannan) evaluation, and compared those findings with similar evaluations in controls with Ultra-negative and Ultra-positive (non-trace) sputum. Results. Of 21 957 people screened with Ultra, 211 (1.0%) tested positive, including 96 (46% of positives) with trace results. Of 92 people enrolled with trace-positive sputum; 12% (11/92) were HIV-positive and 14% (13/92) had prior TB. The prevalence of TB among participants with trace-positive sputum results was 14% (13/92) by culture, 24% (22/92) using broader microbiological criteria, and 26% (24/92) after accounting for clinical diagnosis. The prevalence of cough and of abnormal chest computed tomography (CT) findings were 32% and 26%, respectively, if Ultra-negative; 34% and 54% if trace-positive/non–microbiologically confirmed; 72% and 95% if trace-positive/microbiologically confirmed; and 71% and 93% if Ultra-positive (more than trace). Conclusions. Most individuals with trace-positive sputum in Ugandan communities did not have microbiologically confirmed TB but had more symptoms and chest CT abnormalities than people with Ultra-negative sputum.