• This email address is being protected from spambots. You need JavaScript enabled to view it.

The Thibela TB study

The Thibela TB study was to establish whether community-wide isoniazid preventative therapy (IPT), administered to a whole at-risk community, was more effective when compared to standard TB control measures which included amongst others IPT given to high-risk individuals only, particularly those with HIV/AIDS or silicosis.


Background to the study

During the 1990s, it became clear in the South African gold mining industry that despite meeting World Health Organization (WHO) targets for tuberculosis (TB) detection and cure, the incidence of TB among employees had risen sharply and was the principal cause of death in the workforce.

It was noted too that the five-fold increase in the rates of TB coincided with the onset of the HIV epidemic in South Africa. By 2001 TB case notification rates among South African gold mines were extremely high, exceeding 4000 per 100 000 for 1999 and declined to approximately 3000 per 100 000 by 2008, owing to occupational silica dust exposure and a high prevalence of HIV, estimated at approximately 29% in 2001.

In recent years the HIV-fuelled TB epidemic had been worsened by increasing drug resistance, and TB remained the main cause of death among the mining workforce.

The movement of people throughout the Southern African Development Community (SADC) region to South Africa to work in the mines remains a driver of TB.

Standard TB control measures that included DOTS, active TB case finding, targeted IPT, active promotion of HIV counselling and testing, and freely available antiretroviral treatment for HIV failed to control TB in the gold mines. Despite implementing the best TB control strategies, the incidence of TB in the mining industry remained high (three to six times higher than in the general population), and it was most rampant among gold miners, who made up 32% of all miners.

It was imperative that an alternative approach to controlling TB be investigated. The strategy that was chosen was isoniazid preventive therapy that was used in a novel way.

Most often, isoniazid preventive therapy (IPT) is used as a targeted prevention strategy, meaning that people at high risk of TB such as HIV infected individuals or contacts of people with active TB disease are offered IPT and people whose high risk is not identified will not get it.


By contrast, the strategy of community-wide IPT is appropriate where the community as a whole is at high risk of TB. This strategy was successfully used in Alaska in the 1960s where there was an epidemic of TB, prior to the HIV/AIDS epidemic. An entire community was offered IPT and this radically reduced TB prevalence in that community. In community-wide IPT, IPT is offered to everyone, regardless of perceived individual risk and the effect of community-wide IPT is measured across the entire community. With such high TB notification rates, the gold mining community is an entire group-at-risk. We formulated plans to evaluate the effects of community-wide isoniazid preventive therapy on the incidence and prevalence rates of TB in the South African gold mining industry.


The Consortium to Respond Effectively to the AIDS and Tuberculosis Epidemic (CREATE) was established to develop innovative and radical approaches not just to TB but to HIV/AIDS control as well, in order to reverse the alarming trends of these diseases, which were exacting a devastating toll worldwide especially in the developing countries. CREATE aimed to coordinate, organise, implement and evaluate novel strategies to reduce the incidence of TB and related mortality in communities with high HIV infection rates.

CREATE was led by the Johns Hopkins University School of Medicine, Centre for Tuberculosis Research. Research partners were the Aurum Institute, the London School of Hygiene and Tropical Medicine, the Municipal Health Secretariat (Brazil) and the University of Stellenbosch.

At the 15th International AIDS Conference in Bangkok in July 2004, The Bill and Melinda Gates Foundation announced the award of a $45 million grant to CREATE to conduct three, large-scale community studies over seven years in South Africa, Zambia, (including the one conducted by the Aurum Institute in the South African gold mining industry). As part of this grant, Aurum received $14 million over five years.

In announcing the three studies, Richard E. Chaisson, Professor of Medicine at Johns Hopkins University and Principal Investigator of CREATE, stated, “CREATE’s community-level studies will assess bold new approaches for driving down skyrocketing rates of TB in areas with severe HIV epidemics.”

It was envisaged that the findings of the CREATE research portfolio would be used to develop new global policies to combat TB/HIV, a key criterion of projects supported by The Bill and Melinda Gates Foundation. CREATE could make a major contribution to identifying effective strategies to reduce suffering and death from HIV-related TB worldwide.


The rationale supporting this approach was that all individuals at risk of developing TB in a population would be treated rather than just those identified as being at high risk for TB. If successful, such a strategy would have the added advantage of reducing the transmission of TB resulting in fewer new cases of TB, which would, in turn, lead to improved control of the disease.

Aim and Objective

The aim of the Thibela TB study was to establish whether community-wide isoniazid preventative therapy (IPT), administered to a whole at-risk community, was more effective when compared to standard TB control measures which included amongst others IPT given to high-risk individuals only, particularly those with HIV/AIDS or silicosis.

The main Thibela TB study was a cluster randomised intervention trial. For the purpose of implementing the intervention, clusters were defined as comprising all permanent employees and temporary contractors at participating mine shafts and associated hostels. The study was implemented in three provinces of South Africa: Free State, Gauteng and North West province. Fifteen clusters were identified that ranged in size from 1 000 to almost 12 000 miners. The clusters were randomised to 8 intervention and 7 control clusters.

Measurement of outcomes was restricted to permanent employees and excluded temporary contractors. The outcomes were measured among all permanent employees, regardless of whether or not they took IPT.

Thibela TB’s primary objective was to accomplish a 60% reduction in the incidence of TB in the community-wide arm that received IPT, compared to the control arm 13 to 24 months after enrollment, at 90% power.

In addition to the primary objective stated above, various secondary objectives were formulated:

  • To achieve a reduction of at least 40% in TB case notification rates in the community-wide IPT arm, compared to the control arm 0 to 24 months following enrollment.
  • To accomplish a 60% reduction in the sputum culture prevalence of TB among HIV-infected individuals in the community-wide IPT arm compared to the control arm at the end of follow up.
  • To identify and communicate trends in TB case notification following the intervention, the safety of community-wide IPT, all-cause mortality and the prevalence of isoniazid resistance among TB cases.

Major Collaboration

While the Thibela TB study was overseen and managed by the Aurum Institute in South Africa, it was one of three studies designed by CREATE, of which Aurum was a member (see box above on CREATE). The research conducted by Aurum, under the auspices of the Mine Health and Safety Council (MHSC) and CREATE, followed extensive consultation and collaboration with three South African gold mining companies, AngloGold Ashanti, Gold Fields and Harmony, and with the South African government Departments of Health, Labour, and Minerals and Energy.

Extensive discussion and debate also took place with representatives of the National Union of Mineworkers and other labour unions representing mineworkers, to obtain their support for the study.

The pharmaceutical giant Sanofi-Aventis made a truly powerful gesture when the company decided to donate the isoniazid requirements for the study as part of its global access corporate social investment programme. This very welcome undertaking by the pharmaceutical company signalled clear broader support for Thibela TB.

Pilot Studies

Before the main study commenced, four pilot studies were conducted, whose results and findings led the way for the parent study and provided useful data to inform how the main study was conducted.


This cross-sectional qualitative study explored attitudes towards, and perceptions of, TB and associated therapies. The study looked at how communities responded to research being done on the disease in order to plan and coordinate community education programmes for Thibela TB, so communities could be better informed about the studies and their aims. Participants in this invaluable pilot study included mining employees and ‘key informants’, including team leaders, union representatives, hostel leaders, and ex-employees who had been medically boarded owing to TB but who still lived in the surrounding community. Recruitment of participants to the study commenced in October 2005 and was completed in August 2006. The results of pilot study A assisted recruitment for the main Thibela study and ensured that community mobilization and messaging were finetuned to address the real concerns of the communities regarding TB.


Main Thibela Study

The main Thibela study (parent study) consisted of the following components:

  • The baseline survey: In total 13,958 participants were enrolled from both control and intervention clusters, of which the total number of clusters was 15. From all these 15 clusters, 15,609 volunteers were enrolled in the baseline survey.
  • The IPT intervention in intervention clusters: the IPT intervention included participant screening and follow-up over 9 months. More than 23 500 individuals were eventually enrolled in taking IPT.
  • A TB prevalence survey at the end of the follow-up period (month 24): a cross-sectional survey was carried out to determine the prevalence of TB based on the culture of a sputum specimen. A random sample of 750 employees was selected from each cluster.
  • TB episodes were ascertained from routine data collected on TB patients treated by the mine health services as part of the TCAS study described above.


After months of preparation and putting systems in place, trained Thibela study workers welcomed volunteers at each intervention shaft. Participant recruitment was conducted at the crush, hostel areas and access points, on a one-on-one basis. Consenting individuals were screened for active TB using a symptom questionnaire and chest radiograph. Individuals with no evidence of active TB were offered INH. Workers suspected of having active TB had a sputum sample taken for microscopy and culture, and were referred to the mine health service for further management. Participants in the main study were given calendars to record their use of INH. Peer educators actively encouraged adherence to IPT. Participants were also inspired to adhere through incentives of small value handed out to participants in a variety of events and campaigns, such as participants selected at random at a pick-a-box show once a month.

A nine-month duration of IPT was chosen as a compromise between common IPT recommendations for those with HIV infection (six months) or those with silicosis (twelve months). A nine-month IPT course also had the advantage of maximising the probability that a large number of participants would receive an adequate course, defined as 180 doses (six months), even if adherence was suboptimal.


Participants’ adherence to taking isoniazid was carefully measured at monthly follow-up visits as follows:

  • self-reported pill taking in last three days
  • urine test on a random selection of participants
  • return for visits, and
  • pill counts.

At monthly follow-up visits, an impressive 89% of participants reported missing no tablets in the previous three days.

The median age of participants in the study was 40 years, 95% were men and 11% reported they previously had TB. The majority (54%) lived in hostels. Just over half the participants were South African
citizens (59%).

Lessons learned from The Thibela Study

A number of important lessons emerged even before the report for the main study was completed:

Data from the Thibela TB main study showed that adverse effects of INH were very rare. Despite a population with a relatively high median age (40 years) who would traditionally be considered at high risk of hepatotoxicity, only 17 individuals (0.07%) experienced symptoms suggesting symptomatic hepatotoxicity. Only four events fulfilled criteria as being serious adverse events. The data suggest that clinical criteria can safely be used for screening prior to and monitoring during IPT.

Main Study Outcomes

The key results from the Thibela TB study after all the years of work and analysis of the data are that:

  1. IPT radically reduces TB incidence in individuals who take it (by 66%), but the effect is not lasting.
  2. IPT used as a community-wide strategy has no population-level effect on either incidence or prevalence of TB.


During the IPT treatment period, TB incidence was reduced by nearly two thirds among those who started IPT. This is strikingly shown in the graph below, where the orange and yellow lines on the right diverge so clearly


IPT was effective and of strong benefit to those at risk who took the medication, but the effects quickly wore off when IPT was stopped.

So IPT used on a community-wide basis did not lead to a population-level reduction in TB incidence or prevalence, as the investigators had hoped.

It did, however, suggest that IPT used a longer term or even on a life-long basis to prevent TB in individuals at risk may be an appropriate future strategy.


Policy and Advocacy

During the course of the Thibela study and beyond, Thibela’s investigators contributed to the transformation of global policy in three particularly notable areas:


A major obstacle to wider implementation of IPT was the lack of an evidence-based policy on the most appropriate screening tool to exclude active TB prior to starting IPT. Aurum and its collaborators contributed two datasets to a meta-analysis led by WHO using individual patient data which addressed this question. As a result, a four-symptom screening tool was recommended with high sensitivity and good negative predictive value in most populations. A paper describing this meta-analysis was published by PLoS Medicine.


Publications arising from
The Thibela Study

Most notably, a supplement of the journal AIDS which showcased the experience with IPT of all three of the CREATE projects was published in November 2010.

The work undertaken as part of the Thibela TB programme and published in this supplement included: a cross-sectional analysis, a cross-sectional analysis describing the prevalence and risk factors for tuberculosis at screening prior to offering IPT, a cross-sectional study describing the adverse events occurring in the Thibela TB study, and a description of a programme of community education and mobilisation to promote uptake in a cluster-randomised trial of tuberculosis preventive therapy offered to all members of the intervention clusters.

Three other articles published in this supplement were based on work within Aurum’s HIV care programme. These comprised a qualitative study of clinician and patient barriers to implementation of IPT, an observational study showing an association between IPT started around the time of ART start and reduced mortality, and a viewpoint discussing why randomised controlled trials may have underestimated the potential effect of IPT programmes on mortality among people with HIV.

The impressive list of publications arising from Thibela to date is given below:

  1. Van Halsema CL, Chihota VN, Gey van Pittius NC, Fielding KL, Lewis JJ, van Helden PD, Churchyard GJ & Grant AD. Clinical relevance of non-tuberculous mycobacteria isolated from individuals investigated for tuberculosis in a gold mining workforce in South Africa. Emerging Infectious Diseases (submitted).
  2. Chihota VN, van Halsema CL, Grant AD, Fielding KL, van Helden PD, Churchyard GJ & Gey van Pittius NC. Spectrum of non-tuberculous mycobacteria identified using standard biochemical testing versus 16S sequencing. Int J Tuberc Lung Dis 2013 Feb;17(2):267- 9. doi: 10.5588/ijtld.12.0425. Epub 2012 Dec 8.
  3. Lewis JJ, Fielding KL, Grant AD, Chihota VN, Popane F, Luttig M, Muller D, Coetzee L & Churchyard GJ. Eligibility for isoniazid preventive therapy in South African gold mines. PLoS One (submitted)
  4. Lewis JJ, Chihota VN, van der Meulen M, Fourie B, Fielding KL, Grant AD, Dorman SE & Churchyard GJ. Proof-of-concept evaluation of an automated sputum smear microscopy system for tuberculosis diagnosis. PLoS One (accepted)
  5. Dorman SE, Chihota VN, Lewis JJ, Maunank S, Clark D, Grant AD, Churchyard GJ & Fielding KL. Performance characteristics of the Cepheid Xpert MTB/RIF test in a tuberculosis prevalence survey. PLoS One. 2012;7(8):e43307.
  6. Dorman SE, Chihota VN, Lewis JJ, van der Meulen M, Mathema B, Beylis N, Fielding KL, Grant AD, Churchyard GJ. Genotype MTBDRplus for Direct Detection of Mycobacterium tuberculosis and Drug Resistance in Gold Miners in South Africa. J. Clin. Microbiol. JCM.05723-11; published ahead of print 11 January 2012
  7. Van Halsema CL, Fielding KL, Chihota VN, Lewis JJ, Churchyard GJ & Grant AD. Trends in drug-resistant tuberculosis in a gold mining workforce in South Africa, 2002-2008. Paper submitted to IJTLD February 2011.
  8. Churchyard GJ, Fielding KL, Lewis JJ, Chihota V N, Hanifa Y, & Grant A. Symptom and chest radiographic screening for infectious tuberculosis prior to starting isoniazid preventive therapy: yield and proportion missed at screening AIDS 2010: 24 (suppl 5); S19-S27.
  9. Chihota VN, Grant AD, Fielding K, Ndibongo B, van Zyl A, Muirhead D & Churchyard G. Mycobacterial growth indicator tube compared with Löwenstein-Jensen medium in the investigation of pulmonary tuberculosis: performance and cost in a resource-constrained setting IJTLD 2010:14; 1024-1031.
  10. Van Halsema CL, Fielding KL, Chihota VN, Russell EC, Lewis JJC, Churchyard GJ & Grant AD. Tuberculosis outcomes and drug susceptibility in individuals exposed to isoniazid preventive therapy in a high prevalence HIV setting. AIDS 2010, 24:1051-1055.
  11. Fielding KL, Grant AD, Hayes RJ, Chaisson RE, Corbett EL & Churchyard GJ. Thibela TB: design and methods of a cluster randomised trial of the effect of community-wide isoniazid preventive therapy on tuberculosis amongst gold miners in South Africa. Contemp Clin Trials. 2011 May;32(3): 382-92.
  12. Hanifa Y, Mngadi K, Lewis J, Fielding K, Churchyard GJ & Grant AD. Evaluation of the Arkansas method of urine testing for isoniazid in South Africa. Int J Tuberc Lung Dis. 2007 Nov;11(11): 1232-6.
  13. Hanifa Y, Grant AD, Lewis J, Corbett EL, Fielding K & Churchyard GJ. Prevalence of latent tuberculosis infection among gold miners in South Africa. Int J Tuberc Lung Dis. 2009 Jan;13(1): 39-46.
  14. Grant AD, Coetzee L, Fielding KL, Lewis JJ, Ntshele S, Luttig MM, Mngadi KT, Muller D, Popane F, Mdluli J, Mngadi N, Sefuthi C, Clark DA & Churchyard GJ. ‘Team up against TB’: promoting involvement in Thibela TB, a trial of community-wide tuberculosis preventive therapy. AIDS. 2010, 24 (suppl 5):S37-44.
  15. Grant AD, Mngadi KT, van Halsema CL, Luttig MM, Fielding KL & Churchyard GJ. Adverse events with isoniazid preventive therapy: experience from a large trial. AIDS. 2010, 24 (suppl 5) :S29-36.
  16. Grant AD, Fielding KL, Charalambous S, Chaisson R & Churchyard GJ. Why have trials of isoniazid preventive therapy among people with HIV infection not demonstrated an effect on mortality? Did a close examination of the trees obscure our view of the wood? AIDS 2010, 24 (suppl 5):S15-S18.
  17. Lewis JJ, Charalambous S, Day JH, Fielding KL, Grant AD, Hayes RJ, et al. HIV infection does not affect active case finding of tuberculosis in South African gold miners. Am J Respir Crit Care Med 2009;180(12):127112788.

The Thibela Team

The achievements of the Thibela study were only possible because of the richness of skills and personalities represented in the Thibela research team, which made this groundbreaking work possible:


Gavin Churchyard, Katherine Fielding, Alison Grant, Leonie Coetzee, James Lewis, Violet Chihota, Clare van Halsema, Evans Muchiri, Keith Tomlin, Susan Dorman, Dave Clark, Barun Mathema, Richard White, Emilia Vynnycky, Andy Cox, Tom Sumner, Liz Corbett, Peter Godfrey-Faussett, Dick Chaisson, Richard Hayes.

© 2024 The Aurum Institute. All Rights Reserved.