University of California, San Francisco Logo

University of California, San Francisco | About UCSF | Search UCSF | UCSF Medical Center

Antiretroviral therapy and TB notification rates in a high HIV prevalence South African community
Global Health Sciences Literature Digest
Published April 11, 2011
Journal Article

Middelkoop K, Bekker LG, Myer L, Johnson LF, Kloos M, Morrow C, Wood R. Antiretroviral therapy and TB notification rates in a high HIV prevalence South African community. J Acquir Immune Defic Syndr. 2011 56:3, 263-269.


To assess the population-level impact of increased access to highly active antiretroviral therapy (HAART) on community tuberculosis (TB) rates in an area with high HIV prevalence.


A peri-urban township in South Africa with a population of 15,000 and an HIV prevalence of 23%. Data were obtained from the township’s sole primary care clinic, which provides care for all TB patients.

Study Design

Modeling and assessment of trends based on census data, survey data, and clinical databases


Adults over 15 of age

Main Outcome Measures

Age-standardized annual TB rates (per 100,000) and trends over time; age-standardized rate ratio (RR) of TB in HIV-infected adults on HAART vs. not on HAART; TB completion rates, new TB diagnoses (retreatment TB), and all-cause fatality in TB patients.


TB notification data were obtained from the local TB clinic for 1997-2008. Data on HIV status, CD4 counts and the number of people on HAART were obtained from clinic files and databases. Annual and age-specific TB rate calculations were based on population denominators obtain from national and community census data. Community HIV prevalence data from 1996-2004 was estimated using the Actuarial Society of South Africa (ASSA) 2003 AIDS and Demographic model,(1) adjusted for the results of two community-based random cross-sectional HIV prevalence surveys in the community. This mathematical model was used to estimate the number of HIV-infected persons in the community. Direct standardization methods were used to calculate annual TB rates among those on and off HAART. 2005 was considered the first year of widespread HAART availability.


Over the 12 year period of analysis, 1,974 TB cases were identified, and 1,741 cases were among adults. HIV testing was done among 77% of the adult cases, among whom 69% were HIV infected. In 2003, 1% of HIV positive persons were receiving HAART. In 2005, this increased to 13%, and in 2008, to 21%. From 1997 to 2005, TB notification rates increased steadily, by approximately 187 cases per year (p<0.001). Starting in 2005, adult cases decreased by an average of 202 cases a year (both p<0.001). From 1997 to 2004, TB rates in HIV-uninfected patients increased slightly (49 cases per year), but not significantly; from 2005, cases decreased by 143 cases per year (p<0.001). Among HIV-infected persons, in contrast, TB rates increased by 432 cases per year from 2002-4, and then decreased by 578 cases per year (both p<0.01). TB completion rates did not differ over time or between HIV-infected and uninfected persons.

TB rates among HIV-infected adults not on HAART increased by 362 cases per year from 2002-4, and declined by 416 cases per year afterwards. TB rates among patients on HAART after 2004 decreased even more, by 1,156 cases per year. In 2008, the RR of TB among patients off HAART compared to those on treatment was 1.98. Overall TB retreatment rates increased significantly from 1997-2004 (52 cases per year), and then stabilized. Retreatment rates among HIV positive persons not on HAART increased initially, and then decreased after 2004 (by 157 cases per year). Retreatment rates among those on HAART decreased by 824 cases per year from 2004-2008. TB case fatality rates remained relatively constant in HIV-negative persons, but dropped significantly in HIV-infected patients (from 13% in 2002, to 4% in 2008, p=0.001). Mean CD4 counts of persons initiating HAART increased slightly over time, but remained <200 cells/µl.


TB notification decreased significantly in both HIV-infected and uninfected persons, after the implementation and high coverage of HAART in the community. Deaths among TB patients who were HIV infected dropped to the same level as those who were HIV-uninfected. The authors estimate that if similar coverage rates of HAART were achieved nationally in South Africa, there could be a reduction of more than 20% in TB rates.

Quality Rating

This was a high quality study. Some information was incomplete, such as HIV testing among TB patients; data were also dependent on the quality of clinical records. However, the authors performed a sensitivity and extreme case scenario analysis that did not alter their findings, and performed a 10% quality assurance evaluation of medical records. The generalizability of these findings to other high prevalence settings needs to be confirmed.

In Context

The STOP TB Partnership promotes the use of the “3-Is” to reduce the burden of TB in HIV-infected patients: intensified case finding, isoniazid (INH) preventive treatment, and infection control.(2) Despite this strategy and Directly Observed Therapy - Short Course (DOTS) programs, TB rates continue to rise, including in South Africa.(3) Studies have shown that new cases of active TB in HIV+ patients on HAART is reduced by 70-90%,(4) but the overall risk of TB is still 5-10 times higher than among HIV-uninfected persons.(5) In treatment cohorts, HAART has been associated with a reduction in TB-associated mortality.(6,7) Community level data have not been presented.

Programmatic Implications

This study has shown that despite increased TB notifications in South Africa as a whole, a high coverage HAART program can reduce new cases of TB and TB-associated deaths within a community. HAART roll-out can also decrease the number of TB cases among those who are HIV-uninfected. Reductions in new cases, recurrence and death are greatest for those who are on HIV treatment, probably due to improved immune function, as well as greater active TB screening among HIV positive patients. Regardless, roll-out of HAART may contribute significantly at the community level to a reduction in the TB epidemic, and could be considered an adjunct to the current World Health Organization (WHO) TB control strategy, 3-Is, and DOTS program.(3) At the programmatic level, this is further evidence for the need for active screening for HIV among TB patients, with initiation of ART. Because of the high risk of unmasking TB among patients initiating HAART at low CD4 counts, bringing people into HIV counseling and testing earlier is also important.


  1. Nattrass N. Actuarial Society of South Africa. Journal of Acquired Immune Deficiency Syndromes. 43(5): 618-623
  2. World Health Organization. WHO three I’s meeting: intensified case finding (ICF), isoniazid preventive therapy (IPT) and TB infection control (IC) for people living with HIV. 2008
  3. Global tuberculosis control - surveillance, planning, financing. WHO Report 2008 WHO/HTM/TB/2008.393
  4. Lawn SD, Myer L, Orrell C, Bekker LG, Wood R. Early mortality among adults accessing a community-based antiretroviral service in South Africa: implications for programme design. AIDS. 2005 Dec 2;19(18):2141-8
  5. Lawn SD, Myer L, Bekker LG, Wood R. Burden of tuberculosis in an antiretroviral treatment programme in sub-Saharan Africa: impact on treatment outcomes and implications for tuberculosis control. AIDS. 2006 Aug 1;20(12):1605-12
  6. Akksilp S, Karnkawinpong O, Wattanaamornkiat W, Viriyakitja D, Monkongdee P, Sitti W, Rienthong D, Siraprapasiri T, Wells CD, Tappero JW, Varma JK. Antiretroviral therapy during tuberculosis treatment and marked reduction in death rate of HIV-infected patients, Thailand. Emerg Infect Dis. 2007 Jul;13(7):1001-7
  7. Varma JK, Nateniyom S, Akksilp S, Mankatittham W, Sirinak C, Sattayawuthipong W, Burapat C, Kittikraisak W, Monkongdee P, Cain KP, Wells CD, Tappero JW. HIV care and treatment factors associated with improved survival during TB treatment in Thailand: an observational study. BMC Infect Dis. 2009 Apr 13;9:42