University of California, San Francisco Logo

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

Successful integration of tuberculosis and HIV treatment in rural South Africa: The Sizonq'oba Study
Global Health Sciences Literature Digest
Published June 30, 2009
Journal Article

Gandhi N, Moll A, Lalloo U, et al. Successful integration of tuberculosis and HIV treatment in rural South Africa: The Sizonq'oba Study. JAIDS 2009 Jan;50(1):37-43.


To measure the feasibility, effectiveness, and safety of a tuberculosis (TB) and HIV treatment strategy using directly observed therapy for TB and HIV medications.


The Msinga sub-district of KwaZulu-Natal, South Africa. This impoverished rural area has an estimated HIV prevalence of 25%, incidence of active TB greater than 1,000 per 100,000 population, and more than 80% of TB patients co-infected with HIV.

Study Design

Prospective operational research study


Patients aged 18 years and older who were co-infected with TB and HIV, had a CD4 count of <350 cells/mm3, had no prior TB treatment default, were ART naïve, had completed 1 month of TB treatment by home-based directly observed therapy (DOT), did not have severe hepatic or renal dysfunction, agreed to disclose HIV status to at least one family member, and willingness to take contraception while using efavirenz (women of child-bearing age).


Change in CD4 counts, viral load, weight, TB outcome (i.e., cure, treatment completed, died on therapy, defaulted, or treatment failure), and survival in the first 12 months after initiating ART. Outcomes of TB followed WHO definitions, and adverse events were characterized according to NIH standard protocols.


A government-sponsored TB treatment program was provided with additional resources to integrate services with an HIV treatment program located at the same site. Health care providers, staff, patients, and patients´┐Ż family members all received relevant training. Reliable sources of TB and HIV medications were identified and laboratory support for CD4 and viral load testing was provided. Satisfactory laboratory testing for TB was already in place.

Eligible patients were from either the TB DOT or the HIV program. Patients took their TB and ART doses at home in the presence of a DOT supporter or family member. Treatment for TB was provided Monday through Friday and ART was provided once daily. The DOT supporter documented adherence by inspecting the monthly treatment calendar and by administrating a standardized 7-day adherence measurement instrument monthly.(1) The DOT supporter also assessed patients for adverse reactions to medications. Once TB treatment was completed, the patients were moved to self-administration of their ART. Patients were seen monthly for 12 months at the HIV treatment center and CD4 and viral load testing was performed quarterly.


A total of 174 patients were evaluated for inclusion and 119 enrolled from October 2003 to January 2006. The mean age was 34 years and 56% were women. At the time ART was started, the mean weight was 55.3 kg (SD, 8.0 kg), median CD4 count was 78.5 cells/mm3 (interquartile range [IQR] 42-152), and the median viral load was 250,000 copies/mL (IQR 120,000-590,000).

Patients initiated ART at a median of 67 days after the start of TB therapy. At 6 months the median increase in CD4 cells was 151/mm3 (IQR 81-227; P<0.001) and at 12 months was 211 (IQR 131-283; P<0.001). Weight increased by 6.5 kg (SD 6.5 kg) at six months and by 10.5 (SD 7.9 kg; P<0.001) at 12 months. At 6 months, 83% of the patients had an undetectable viral load (defined as <400 copies/mL) and at 12 months 88% were undetectable. There were 6 patients whose viral load was not suppressed and 5 of these patients were found to have genotypic resistance: 4 to efavirenz, 3 to lamivudine, and 2 to didanosine. Seven patients were lost to follow-up at 12 months.

Eighty-four percent of patients were cured or successfully completed TB therapy. Eleven patients died while on therapy: 5 who defaulted, and 3 who failed treatment. Multidrug resistance was found in the 3 living patients who failed therapy.

Patients attended 93% of the follow-up visits within 1 day of the scheduled appointment and reported taking 99% of their ART doses. There were 34 severe adverse events and 1 case of pancreatitis. Thirteen patients died in the 12 months after ART initiation. Of these, 10 died of confirmed multi-drug resistant TB (MDR TB), confirmed extensively drug-resistant TB (XDR TB), or suspected drug-resistant TB; 3 patients died of unknown or other causes (respiratory failure, hematemesis, and acute diarrhea).


This integrated TB-HIV treatment program resulted in excellent adherence as reported by patients. Clinical improvements consistent with therapy were observed.

Quality Rating

This operations research used a cohort design, and as a cohort study was of fair quality. No comparison group was included; all patients enrolled in the integrated TB-HIV program were followed. The patients were referred from a variety of settings and information regarding any potential or demonstrated biases was not included. The presence of TB and HIV was objectively documented and baseline measurement of outcomes was appropriate and complete. Outcome measures were objective and appropriate. Follow-up time and proportion completing the study was adequate. This study was limited by its operations research design; a randomized intervention trial would have been much stronger, although perhaps not feasible. Comparison of TB outcomes in this setting to published outcomes was provided and did show an improvement. As well, the study demonstrated the extent of deaths associated with multidrug resistant TB - an important, if unexpected, finding. Comparison to ART outcomes, however, was not mentioned and thus we cannot determine whether the integrated approach is an improvement for providing HIV care.

In Context

More than 65% of persons infected with TB are co-infected with HIV, and TB is the leading cause of death among HIV-infected persons.(2) Although TB treatment programs have been well established in developing countries, treatment completion rates are only approximately 60%, due in large part to the increases in TB cases since the late 90s. Although ART is now available in sub-Saharan Africa, infrastructure for ART programs is insufficient. Currently, TB and HIV treatment programs are separate and do not share health care providers, staff, or funds. Integrating ART and TB treatment programs may be one method to increased TB treatment and to provide infrastructure to ART programs.(3,4)

Programmatic Implications

It is particularly important in resource constrained areas to identify the most cost-effective methods to provide quality care. This study demonstrated that treatment for TB and HIV can be given concurrently and can result in clinical improvement. Both of these findings are important. Using the infrastructure provided with TB programs to facilitate HIV ART is an excellent idea and merits additional evaluation. The current distinct funding streams for HIV and TB may not be ideally suited to a region in which large proportions of persons are infected with both TB and HIV. Additional studies to examine the effectiveness of integrated care programs are needed.


  1. Mannheimer S, Friedland G, Matts J, et al. The consistency of adherence to antiretroviral therapy predicts biologic outcomes for human immunodeficiency virus-infected persons in clinical trials. Clin Infect Dis 2002;34:1115-21.
  2. Mukadi YD, Maher D, Harries A. Tuberculosis case fatality rates in high HIV prevalence populations in sub-Saharan Africa. AIDS 2001;15:143-52.
  3. Nunn P, Reid A, De Cock KM. Tuberculosis and HIV infection: the global setting. J Infect Dis 2007;196 (Suppl 1):S5-14.
  4. Friedland G, Harries A, Coetzee D. Implementation issues in tuberculosis/HIV program collaboration and integration: 3 case studies. J Infect Dis 2007;196 (Suppl 1):S114-23.