Does antiretroviral therapy reduce hiv-associated tuberculosis incidence to background rates? A national observational cohort study from england, wales, and northern ireland

Rishi K. Gupta*, Brian Rice, Alison Brown, Helen Thomas, Dominik Zenner, Laura Anderson, Debora Pedrazzoli, Anton Pozniak, Ibrahim Abubakar, Valerie Delpech, Marc Lipman

*Corresponding author for this work

Research output: Contribution to journalArticlepeer-review

33 Citations (Scopus)


Background Whether the incidence of tuberculosis in HIV-positive people receiving long-term antiretroviral therapy (ART) remains above background population rates is unclear. We compared tuberculosis incidence in people receiving ART with background rates in England, Wales, and Northern Ireland. Methods We analysed a national cohort of HIV-positive individuals linked to the national tuberculosis register. Tuberculosis incidence in the HIV-positive cohort (2007-11) was stratifi ed by ethnic origin and time on ART and compared with background rates (2009). Ethnic groups were defi ned as follows: the black African group included all individuals of black African origin, including those born in the UK and overseas; the white ethnic group included all white individuals born in the UK and overseas; the south Asian group included those of Indian, Pakistani, and Bangladeshi origin, born in the UK and overseas; and the other ethnic group included all other ethnic origins, including black Afro-Caribbeans. Findings The HIV-positive cohort comprised 79 919 individuals, in whom there were 1550 incident cases in 231 664 person-years of observation (incidence 6·7 cases per 1000 person-years). Incidence of tuberculosis in the HIVpositive cohort was 13·6 per 1000 person-years in black Africans and 1·7 per 1000 person-years in white individuals. Incidence of tuberculosis during long-term ART (≥5 years) in black Africans with HIV was 2·4 per 1000 personyears, similar to background rates of 1·9 per 1000 person-years in this group (rate ratio 1·2, 95% CI 0·96-1·6; p=0·083); but in white individuals with HIV on long-term ART the incidence of 0·5 per 1000 person-years was higher than the background rate of 0·04 per 1000 person-years (rate ratio 14·5, 9·4-21·3; p<0·0001). The increased incidence relative to background in white HIV-positive individuals persisted when analysis was restricted to person-time accrued on ART with CD4 counts of at least 500 cells per μL and when white HIV-positive individuals born abroad were excluded. Interpretation Tuberculosis incidence is unacceptably high irrespective of HIV status in black Africans. In white individuals with HIV, tuberculosis incidence is signifi cantly higher than background rates in white people despite long-term ART. Expanded testing and treatment for latent tuberculosis infection to all HIV-infected adults, irrespective of ART status and CD4 cell count, might be warranted. Funding Public Health England.

Original languageEnglish
Pages (from-to)e243-e251
JournalThe Lancet HIV
Issue number6
Publication statusPublished - 1 Jun 2015

Bibliographical note

Funding Information:
This study was done by the UK TB-HIV Research, Epidemiology and Development (UK-THREAD) group. We would like to thank all members of the HIV & AIDS Reporting System and Enhanced TB Surveillance teams for their excellent data management and for providing the data for this analysis. This work was supported by internal funding from Public Health England. Public Health England has powers under Regulation 3 of Section 251 of the NHS Act (2006) to approve the collection, collation, and analysis of confidential patient information for public health purposes and to assess public health programmes to recognise trends in communicable diseases and other risks to public health, and control and prevent the spread of such diseases and risks.


Dive into the research topics of 'Does antiretroviral therapy reduce hiv-associated tuberculosis incidence to background rates? A national observational cohort study from england, wales, and northern ireland'. Together they form a unique fingerprint.

Cite this