Impact of on-site Xpert on TB diagnosis and mortality trends in Uganda

SETTING: Since 2012, Uganda expanded the Xpert® MTB/RIF network for diagnosis of TB.OBJECTIVES: We compared TB care cascades at health facilities with on-site Xpert vs. facilities that accessed the assay through specimen referral.DESIGN: We analysed secondary aggregate data of the National TB and Leprosy Program (NTLP) from 2016 to 2019. We computed the proportions of notified TB cases and mortality ratios in relation to the estimated TB burden.RESULTS: TB case notifications per annum increased from 24,287 in 2016 to 30,739 in 2019, and the proportion of cases diagnosed at facilities with on-site Xpert testing increased from 62% (15,070/24,287) to 81% (24,829/30,739) (P < 0.001). TB mortality at facilities with on-site Xpert decreased from 8.6% (1,302/15,070) to 7.8% (1,938/24,829) (P = 0.41), while it increased at facilities without on-site Xpert from 6.9% (638/9,217) to 8.8% (521/5,910) (P = 0.23). Furthermore, mortality among TB-HIV co-infected patients at facilities with on-site Xpert dropped from 5.0% (760/15,070) in 2016 to 4.8% (1,187/24,826) in 2019 (P = 0.84) compared to 4.4% (407/9,217) in 2016 to 5.3% (315/5,910) in 2019 (P = 0.57).CONCLUSION: Wider installation and decentralisation of Xpert leads to increased case-finding. However, the impact on reduction in mortality remains limited. Interventions to address TB-related mortality in addition to Xpert roll-out are required.