Incidence and risk factors for tuberculosis among people with HIV on antiretroviral therapy in the UK

OBJECTIVE: The UK has a low tuberculosis incidence and earlier combination antiretroviral therapy (cART) is expected to have reduced incidence among people with HIV. Epidemiological patterns and risk factors for active tuberculosis were analysed over a 20 year period among people accessing HIV care at sites participating in the UK CHIC observational study. DESIGN: Cohort analysis. METHODS: Data were included for individuals over 15 years old attending for HIV care between 1996 and 2017 inclusive, with at least three months follow up recorded. Incidence rates of new tuberculosis events were calculated and stratified by ethnicity (white/black/other) as a proxy for tuberculosis exposure. Poisson regression models were used to determine the associations of calendar year, ethnicity and other potential risk factors after cART initiation. RESULTS: 58,776 participants (26.3% female; 54.5% white, 32.0% black, 13.5% other/unknown ethnicity; median (interquartile range) age 34 (29-42) years) were followed for 546,617 person-years. 704 were treated for active tuberculosis (rate 1.3 [95% confidence interval (CI) 1.2-1.4]/1000 person-years). Tuberculosis incidence decreased from 1.3 [1.2-1.5] to 0.6 [0.4-0.9]/1000 person-years from pre-2004 to 2011-2017. The decline among people of black ethnicity was less steep than among those of white/other ethnicities, with incidence remaining high among black participants in the latest period (2.1 [1.4-3.1]/1000 person-years). 283 participants (191 (67%) black African) had tuberculosis with viral load < 50 copies/ml. CONCLUSIONS: Despite the known protective effect of cART against tuberculosis, a continuing disproportionately high incidence is seen among black African people. Results support further interventions to prevent tuberculosis in this group.