Empiric tuberculosis treatment and 12-month mortality among sputum GeneXpert-negative adults living with HIV in Uganda in the era of widespread antiretroviral therapy: A prospective cohort study

Background: In sub-Saharan Africa where both tuberculosis (TB) and HIV are prevalent, empiric TB treatment in people living with HIV (PLHIV) persists due to limited sensitivity of sputum-based TB tests. We evaluated mortality among molecular test-negative presumptive TB adult PLHIV in a population with widespread antiretroviral therapy (ART) coverage, comparing those empirically treated and not treated for TB. Materials and methods: From November 2017 to December 2020, Xpert-negative adult PLHIV were recruited at Mulago Referral Hospital and Kisenyi Health Centre-IV in Kampala, Uganda. Clinical data including TB symptoms, chest X-ray, and empiric TB treatment decision were collected. Laboratory investigations included CD4 cell count, serum cryptococcal antigen (CrAg), urine TB-lipoarabinomannan (TB-LAM), microbiological blood cultures, and sputum mycobacterial growth indicator tube (MGIT) cultures. Participants were followed monthly for 12 months to ascertain vital status. Results: Overall, 300 participants were enrolled; 61.3% inpatients, 55.7% female, median age 37 (IQR 29–45) years, 82.3% on ART, median CD4 206 cells/mm³ (IQR 36–507). Of the 300, 68 (22.7%) received empiric TB treatment, of which 53 (77.9%) were inpatients. 12-month mortality was 31.0% (93/300); 72% within three months post-enrolment. TB cultures were positive in 5.0% (15/300), 12.3% had positive CrAg, and 3.7% had positive blood culture. Empirically treated participants had higher mortality (42.7%) than non-treated (27.6%) although the difference was not statistically significant after adjusting for factors indicative of severe illness (adjusted Mortality Rate Ratio- aMRR 1.05, p = 0.850). Inpatient (aMRR 3.80, p = 0.008) and poor functional status (aMRR 3.5 95%, p = 0.001) were independently associated with 12-month mortality. Conclusion: We found high 12-month mortality among Xpert-negative PLHIV, predominantly inpatients and within three months post-enrolment. Although empiric TB treatment was associated with higher mortality, this association was lost after adjustment for severe illness indicators. Cryptococcal antigenemia and bacteremia were not uncommon, underscoring need for comprehensive evaluations for non-TB conditions alongside empiric TB treatment.