Amount received from the foundation during the reporting period:	$9,043,535
Amount of interest earned on received grant funds in your last fiscal year:	239,461
Balance of grant funds remaining (or overspent) at time of report:	$9,282,996

 

Subgrantee/Subcontractors	Total Amount $U.S.	Duration From/to dates	Grant or contract
London School of Hygiene & Tropical Medicine - ZAMSTAR	$ 3,294,667	09/01/2006-08/31/2007	Contract
London School of Hygiene & Tropical Medicine – Biostatistics Core	$    412,587	09/01/2006-08/31/2007	Contract
Aurum Institute	$ 2,738,793	09/01/2006-08/31/2007	Contract
Rio de Janeiro Secretariat of Health	$    698,475	09/01/2006-08/31/2007	Contract
World Health Organization – PAC	$   262,451	09/01/2006-08/31/2007	Contract

 

 

 

 

 

 

 

 

 

 

 

Public Project Description:

is a global research and implementation project targeting tuberculosis (TB) control in areas with severe HIV/AIDS epidemics. CREATE’s purpose is to evaluate novel strategies to reduce TB morbidity and mortality in populations with high rates of HIV and TB co-infection. TB is the leading killer of people with HIV infection worldwide, and the HIV epidemic has seriously undermined TB control efforts globally. New strategies using currently existing tools will enhance TB control and will facilitate future interventions designed to reduce the incidence of morbidity and mortality from HIV-related TB.

CREATE is conducting three large, community level studies evaluating alternative strategies for control of TB in Brazil, South Africa and Zambia.  The Thibela TB project is assessing the impact of community-wide isoniazid preventive therapy (IPT) in South African miners.  The THRio study is evaluating extensive use of IPT in HIV-infected patients with access to antiretroviral therapy in Rio de Janeiro.  The ZAMSTAR study is a community-randomized trial of enhanced TB case finding and household interventions to reduce risks among contacts of TB patients in Zambia and South Africa.

CREATE has a Biostatistics Core at the Johns Hopkins Bloomberg School of Public Health and the London School of Hygiene and Tropical Medicine.  The CREATE Policy and Advocacy Core is shared by the Administrative Core, the World Health Organization and each of the individual projects.  The Core maintains communication with the TB/HIV global community, and advocates policies to improve TB/HIV control at the local, national and global levels.  An Education and Training Core provides support to study sites for training a new generation of public health researchers who will have an enduring impact on health improvement in the years to come.

 

 

Other Sources of Project Support: Did the project receive any additional unreported monetary support from outside donors during the last reporting period? Y 

Are any new sources of additional monetary support expected in the next funding period?   Y

If you answered yes to either of the above questions, please complete the table below, and provide a brief narrative.  Be certain to list any changes in previously reported sources from your last report.

FIND: 

   Thibela: Donated MGIT machine and funds for supplies for culture of specimens during study

   THRio:   Donated MGIT machine and funds for supplies for culture of specimens during study

   ZAMSTAR:   Donated 3 MGIT machines and funds for supplies for culture of specimens during pilot                              study

                       Donated Quantiferon-gold In-Tube tests and funds for evaluation

 

NIH:       KO1 award for Jonathan Golub to work on THRio study

               R01 for studying MDR/XDR TB diagnosis and natural history (pending)

NIH/Fogarty:    Educating Brazilian Public Health Professionals on TB/HIV Research

 

NIH/Fogarty:    Educating South African Public Health Professionals on TB/HIV Research (pending)

 

DFID:   Tuberculosis Knowledge Program is supporting evaluation of pilot studies in Lusaka

 

Donor	Amount In $U.S.	% of project	Committed or Potential	Expected Receipt date
FIND - Thibela TB FIND – THRio FIND – ZAMSTAR	$  97,731 $  71,140 $203,000	2.1% 6.0% 5.5%	Committed Committed Committed	Received Received Received
NIH – THRio	$45,000	3.9%	Committed	Received
DFID – ZAMSTAR	$100,000	2.8%	Committed	Received
NIH-ICOHRTA, Brazil	$100,000	2%	Committed	Received
Sanofi-Aventis, Thibela TB (donated Isoniazid)	$240,000	4%	Committed	Received
NIH- M/XDR TB	$250,000	<5%	Potential	01/08
NIH – ICOHRTA, South Africa	$100,000	<5%	Potential	9/07

 

 

 

 

 

 

 

 

 

 

 

 

 

Introduction

The Consortium to Respond Effectively to the AIDS/TB Epidemic (CREATE) is a global health consortium with the goal of identifying, evaluating and advocating for novel strategies to reduce tuberculosis incidence in populations with high rates of HIV and tuberculosis co-infection.  CREATE is coordinated by the Johns Hopkins Center for Tuberculosis Research and has multiple international partners.  Phase II is the implementation phase and was initiated in June 2004.  This report provides a summary of activities for year 3, from 1 September 2006 through 31 August 2007.

Year 3 of CREATE was marked by tremendous progress with research operations, extensive reconfiguration of our policy and advocacy activities, and sobering challenges in meeting the timelines and recruitment goals for our unprecedentedly large and complex studies.

All three CREATE studies are fully operational and amassing a wealth of data.  THRio has collected detailed clinical data on >14,000 HIV-infected patients in Rio de Janeiro; Thibela has enrolled >8,000 miners in 4 intervention clusters (with 4 control clusters) into INH preventive therapy; and ZAMSTAR has full-fledged field operations in 24 communities providing extensive case finding and/or household interventions to >1 million people in Zambia and South Africa.  Nonetheless, these ambitious trials have also faced considerable challenges, discussed below.

In close consultation with the BMGF, CREATE has thoroughly revamped its Policy and Advocacy structure, with a major expansion and strengthening of policy/advocacy functions within the central Administrative Core, as well as within each of the CREATE projects.  This enhanced organizational focus on policy and advocacy adds considerable strength to the global effort to transform TB/HIV policies and practices, complementing the leadership, normative and advisory functions provided by the World Health Organization.  This effort culminated in the recent award of $4.4 million of supplemental funding from the BMGF for Policy and Advocacy.

Despite the notable progress with implementation of the CREATE studies, serious challenges have emerged that have required frank and harsh reassessments of our ability to deliver the research results that our portfolio is intended to provide.  Each of these challenges has been met by a thorough consideration of the options for addressing them, bringing in the expertise of our independent Data Safety and Monitoring Committee (DSMB), as well as the judgment and expertise of the CREATE Executive Committee.  A brief summary of the issues is given below, with more detailed discussion of our proposed solutions in the sections that follow. 

 

• THRio:  The THRio study has published baseline data from 11,000 patients that convincingly demonstrate the value of INH preventive therapy in addition to antiretrovirals.  Nonetheless, the power of  the main trial to show a population-level benefit is less than planned (20-50%), because the prevalence of latent TB is lower than expected (~15% of tested patients vs. 25% assumed).  In spite of this reduced power, the DSMB strongly recommends continuing the trial, with the expectation that individual level data will be important and compelling, even if the population level data do not achieve the originally expected result.  The power of the study to achieve its main objective depends on both the coefficient of variability (k) of TB rates within clusters (HIV clinics) and the uptake of the intervention.  The k for the study is not yet known, but is expected to be lower than assumed in our sample size calculations, which would increase power.  The study team is striving to maximize uptake of the intervention in the coming year, which will also increase the likelihood of a positive result.

• Thibela:  The Thibela trial is a massive undertaking on an unprecedented scale.  Based on previous experience, the study team assumed that enrollment would average 150 participants per day.  Actual enrollment, however, has only been ~30 per day.  Coupled with delays in implementing the study, this slower than anticipated accrual will require prolongation of the study timeline, at considerable added expense.  The Thibela team presented an exhaustive review of all possible options for completing the trail, ranging from immediate closure to full implementation, to both the DSMB and the Executive Committee.  Following intensive discussion, the DSMB and Executive Committee each recommended an option that would accelerate recruitment by enrolling in multiple mines simultaneously and a 2-year extension of the study timeline.  This option has the greatest likelihood of providing a definitive answer to the study’s principal question, but also contains two go/no-go decision points where the DSMB could recommend stopping the trial.  The additional cost of this change is appreciable: $9.5 million.  A fuller description is provided below.

 

• ZAMSTAR:  The ZAMSTAR study is likewise an enormously complex undertaking, with 24 sites in two countries and a population coverage of ~1.5 million.  Full implementation of the ZAMSTAR interventions was delayed by human subjects infractions (improper consenting) in the pilot studies, as well as by political events in Zambia.  As a consequence, the study is approximately 6 months behind schedule.  The impact of the intervention was initially assumed to require at least 3 full years to accrue; thus, ZAMSTAR anticipates needing an extension of funding to permit the intervention to be completed.  This has no implications for Year 4, but would affect Years 5-6.  In the coming year, the ZAMSTAR team, the Biostats Core and the Modeling Project will further examine data from the baseline studies and conduct additional modeling to determine the optimal duration of the intervention.  Depending on the outcomes of these analyses, ZAMSTAR anticipates requesting supplemental funding to extend the intervention by 6, 12 or 18 months, beginning at the onset of Year 5 of CREATE.

 

Accomplishments:

Overall goal:  To develop, evaluate, and promote new public health strategies to reduce the incidence of tuberculosis in settings of high HIV prevalence.

 

Objective 1.  Design, implement and evaluate a portfolio of community–level trials of new strategies designed to reduce TB incidence in communities with high HIV prevalence.

 

• The Administrative Core performed on target for its activities for the year, and undertook a number of new activities related to policy and advocacy, as well as communications.  Subcontracts and fiscal oversight for partners were maintained. Two meetings of the CREATE Executive Committee were held, the first in London in April and the second in Annecy, France, in June.  The Consortium will hold its third annual meeting in Rio de Janeiro, Brazil in September 2007.  Administrative and protocol monitoring site visits were carried out by the Administrative Core throughout the year.  The CREATE DSMB held its third meeting in June 2007.  A new electronic newsletter has been launched, with two issues distributed to approximately 400 partners and key allies.  A third edition will be posted in August.  The administrative core led the process of developing a new Policy and Advocacy plan, and has successfully negotiated a supplement for the new plan with the BMGF. 

• The Biostatistics Core had another successful although somewhat stressful year.  It prepared the first full DSMB report, with a large number of analyses of baseline and pilot data.  Standardization of reports across studies was largely successful, enabling the DSMB to navigate to the desired elements. 

 

Thibela TB devised an INH resistance substudy, and spent a great deal of effort tracking recruitment and making projections, which have been critical for study execution and planning.  For ZAMSTAR, electronic capturing of TB treatment and lab register data using web-based applications has been made operational. A data management system for the 3 year cohort study (Secondary Outcome Cohort) of 12,000 individuals over 24 communities has been implemented. ZAMSTAR undertook preparations for full GPS mapping, including buying the requisite hardware, and prepared design papers.  Their mixture analyses of data on thousands of TST results from Zambia and South Africa proved highly intriguing. THRio continued analyses of baseline data, looking at effects of drug regimens and prior TB on subsequent TB incidence, and perfected data merging algorithms that permit incorporation of municipal and state data bases.

• A modeling project led by Dr. Elizabeth Corbett of the London School of Hygiene and Tropical Medicine received a $997,000 award from BMGF to utilize data from the CREATE projects and other sources to model the potential impact of public health interventions for the African TB/HIV epidemic.

 

Thibela TB

In order to maximize the coverage of the intervention the study is now co-enrolling in three regions over a longer period. This change was agreed upon by the CREATE executive committee and the Gates Foundation Project Officer and it was agreed that intensification of enrolment activities was key to the success of the project.

• Year 3 was focused on recruitment and the rollout of the study to new mine shafts. The concurrent enrollment process in underway, with two additional intervention and control shafts opened up in the North West and Free State.  More than 8000 individuals have been placed on INH preventive therapy.
• Community mobilization program was implemented at four additional clusters. Enrolment is occurring concurrently at three sites to maximize study participation.
• Baseline prevalence survey completed enrolment at 6 of 13 clusters.                       
• TB Case notification and TB Case Assessment is occurring on an ongoing basis. Six clusters are enrolled.
• As indicated by the Markinov study (see Activity Table), Thibela TB has successfully shifted attitudes towards TB and TB prevention in the community and created an enabling environment for the successful implementation of the study.
• Thibela has been nominated for two categories in the inaugural awards recognizing excellence in clinical research organized by the International Good Clinical Practice Journal (www.gcpj.com), including “Clinical Trial which Best Promoted Access to Medicine” and “Most Innovative Patient Recruitment Strategy”.

 

THRio

• The intervention will be implemented in 24 of 29 study clinics as of 31 August 2007, as planned. 
• Over 14,000 HIV (+) patients have been entered into the study database since its inception, including >2400 newly diagnosed patients entering treatment since the study start date of September 1, 2005.
• A cost-effectiveness study INH prophylaxis intervention will be implemented in three phases. 
• The MGIT substudy supported by FIND has been implemented and is providing TB culture results at the intervention clinics.  As of May, 370 specimens have been analyzed.  72% of positive cultures are sputum smear negative, underscoring the importance of this technology.
• Baseline data were analyzed, demonstrating a significant impact of the combination of IPT and ART for preventing TB in HIV-infected patients.  These results were presented at the International AIDS Conference in Toronto and are now published in AIDS.
• An additional analysis of the rates of recurrent TB was presented at the American Thoracic Society meeting in May 2007.  This analysis shows that patients with incompletely treated prior TB have an extremely high risk of developing recurrent TB.  These data are being utilized to identify high-risk patients in need of IPT.
• The THRio methods paper was published in Controlled Clinical Trials.

 

ZAMSTAR 

The third year of the ZAMSTAR study has been characterized by consolidation of the interventions in all 24 communities; continued engagement with local and national stakeholders; analysis of the data collected during the baseline studies and planning for the future in order to maximize the chances of measuring the impact of the study.  The study infrastructure has been strengthened.

• Interventions are well established in all 24 sites.
• Secondary outcome cohort recruitment is ongoing in all 24 sites.
• Laboratories and research infrastructure upgraded to receive samples and data from SOCS and substudies.
• Dissemination is ongoing at local, national and international levels.

 

o   Additional Studies

• • Case-control study of poverty and prevalent TB fieldwork completed.
  • MGIT demonstration study phase 1 and 2 completed laboratory work.
  • Capilia demonstration project data analysis ongoing.
  • Quantiferon-Gold In-tube test (QGIT) demonstration project began.
  • Adherence to ARVs in ZAMSTAR cohorts under study.
  • Cost-effectiveness of TB culture manuscript ready for submission.
  • Study on understanding of informed consent for prevalence surveys completed.

Education and Training Core 

• The Education and Training Center has focused on continued on-site needs assessments and evaluations to determine training effectiveness and ongoing training needs; identification and development of education and training resources; and providing information and opportunities for participation in training activities.
• By the end of August, more than 20,000 person hours of training will have been delivered.
• Four CREATE personnel are enrolled in advanced degree programs, one is completing a medium term training course in bioethics and 15 others are engaged in long term training.
• A collaboration has been formed with Clinical Care Options for HIV (CCO-HIV, www.clinicaloptions.com/hiv), a medical education company with a widely used internet portal for HIV education, to establish on-line TB/HIV educational modules, highlighting the objectives of CREATE.  This program will be used on the company’s website, which reaches a large audience of HIV providers and policy makers in the US, Europe, and Africa.  In addition, the modules will be translated into Portuguese and posted on the CCO Brazilian website.

 

Modeling

• The Modeling project has been approved and is launching activities immediately.  A modeling meeting will be held in London in late August, and data needs are already being determined.  An effort has been made to locate as many original data as possible from the USPHS Bethel INH prophylaxis study in the 1950s-1960s, with the cooperation of Dr. George Comstock.

 

Objective 2.     Transform global policies for HIV-related TB through evidence-based advocacy.    

 

In Policy and Advocacy, the year has been tremendously productive in achieving CREATE’s second objective to transform global policies by building infrastructure and capacity to accomplish this goal.

Over the past nine months the Policy and Advocacy Core has been strengthened by restructuring to play to the strengths of the World Health Organization and PAC team members in Baltimore.  Dr. Lois Eldred was recruited from the PEPFAR program at the Health Resources and Services Administration to lead policy and advocacy activities within the CREATE secretariat, and each of the CREATE projects has hired or will soon hire a policy and advocacy coordinator.  The WHO has hired Ms. Colleen Daniels to coordinate activities within the UN agencies, as Dr. Alasdair Reid has moved to UNAIDS (where he has had considerable impact on raising the visibility of TB).

WHO PAC activity has mainly focused in promoting the visibility of TB/HIV and accelerating the implementation of collaborative TB/HIV activities particularly in the 63 priority countries which constitute for 98% of the global burden, in addition to CREATE-specific activities.  The CREATE secretariat and the WHO have worked jointly to organize and promote several high-profile meetings at professional and scientific meetings, in order to increase awareness of and attention paid to TB/HIV among scientists working in both fields.

• The WHO’s work around implementation focused on building partnership with Office of the Global AIDS Coordinator, which resulted in a joint meeting directly supported by the BMGF in March 2007 for five selected countries and an additional pledge of $50 million in PEPFAR funds for TB/HIV in 2007, with additional funds likely in the 2008 budget.
• WHO and CREATE organized a TB/HIV priority research meeting in conjunction with CROI 2007 in Los Angeles, USA, which was co-chaired by the NIH’s director of HIV Therapeutics (Dr. Sandra Lehrman) and the Chair of the STOP-TB TB/HIV Working Group (Dr. Diane Havlir). The meeting raised the awareness of the TB/HIV WG and presentations on key areas by internationally renowned experts highlighted ongoing research efforts and priorities.
• A new working relationship has been established with International AIDS Society to ensure the mainstreaming of HIV/TB into the Secretariat's work, including upcoming HIV conferences.  CREATE, the WHO and the Forum for Collaborative HIV Research will sponsor two meetings on TB/HIV at the IAS meeting in Sydney in July, one a symposium on critical aspects of TB/HIV research and care, and the second a convocation of researchers and funders to discuss priorities for targeted research and collaboration.
• A session during the Stop TB Partnership Coordinating Board meeting that was held in April 2007, which was attended by the Director General of WHO, Executive Director of UNAIDS and the UN Special Envoy on TB that discussed key strategic issues, which includes the inclusion of TB in the next UNGASS in 2008 and also in the upcoming Unified Budget Work plan of UNAIDS.
• Several symposia at the 37^th Annual Meeting of the International Union Against TB and Lung Disease featured CREATE studies and speakers.
• Thibela’s work with the mining industry and, specifically, the Minster of Mineral and Energy has potential to change policy towards INH prophylaxis at a national level, with global implications.
• In response to the M(X)DR-TB outbreak in Tugela Ferry, South Africa, CREATE retained a consultant, Dr. Celine Gounder of the Massachusetts General Hospital,  to conduct field assessments and policy reviews for the two South African projects.  Dr. Gounder visited Thibela and ZAMSTAR to review strengths and vulnerabilities of the programs.  Infection control procedures were adequate in the studies; recommendations were made to refine protocols and practice to lessen the risk of TB transmission.  Several recommendations were made for studies to assess the best means to reduce M(X)DR-TB, including modeling the prevalence of MDR-TB at which rifampicin resistance testing is cost effective and determining the most cost-effective infection control methods in these settings.  One outcome of this activity was the preparation of an NIH grant proposal to study the epidemiology and natural history of M(X)DR in Thibela, and to evaluate rapid diagnostic tests for rifamycin resistance.
• CREATE has worked with the Global Alliance, Medecins sans Frontieres, the Open Society Institute, the Treatment Action Group and other partners to promote TB literacy, increase awareness of TB/HIV coinfections as part of the HIV response, and to advocate for greater research on new tools and  strategies in a number of meetings in the US, Europe, Brazil and South Africa.
• Policy and Advocacy received a $4.4 million award to enhance global policy and advocacy activities.

Additional Specific steps being taken can be seen under VII, Progress in Achieving a Global Strategy. 

 

Changes:

Administration and Policy and Advocacy:

• Dr. Lois Eldred has been hired as the CREATE Project Director and the Director of Policy and Advocacy.
• Ms. Colleen Daniels has been hired for the CREATE policy and advocacy position at the World Health Organization.
• With supplemental funding from the Gates Foundation, Policy and Advocacy activities will be substantially accelerated beginning at the end of Year 3 and continuing into Years 4-6, including but not limited to hiring a PAC coordinator at each of the study sites and coordinating with partners to advance an HIV-related TB policy agenda.

 

Education and Training Core:

• Ms. Linda Watkinson retired on June 30, 2007, ending a 25 year career at Johns Hopkins.  A replacement to lead the ETC is being sought.

 

Modeling:

• Dr. Liz Corbett has joined the CREATE Executive Committee in recognition of her contributions and duties as head of the Modeling project.

 

Thibela:

• The target recruitment rates and assumptions in the initial protocol have proven to be unachievable in practice. In order to maximize the coverage of the intervention the study is now co-enrolling in three regions over a longer period. This change in implementation plan was discussed with the DSMB and the CREATE Executive Committee, and it was agreed that intensification of enrolment activities was key to the success of the project. The effect however is to accelerate the level of expenditure on the project and this will necessitate a drawdown of funding in Year 3 against the future grant at a rate greater than the plan presented in the 2006 Annual Report.  A full discussion is presented below under “Challenges.”

 

THRio:

• THRio will maintain lab and clinic infrastructure necessary to continue MGIT through the end of the project, with FIND continuing to provide the supplies.
• Due to multiple problems with availability of the original members of the THRio CAB (including the death of one member and time commitments of others), the CAB has been completely reformed and has resumed regular monthly meetings.  In addition, the new CAB will include participation of NGOs that work at the community level, not just with affected groups (i.e. those with HIV and TB).

 

Challenges:

THRio:

• Uptake and adherence to IPT remains a challenge for the project team.  The community mobilisation and incentive programme has been intensified to support recruitment and retention activities.
• Strategies used to improve uptake include regional and site launches, community education, providing incentives and encouraging leadership by mine management and labor.
• Strategies used to improve adherence include education adherence road shows, incentive events and similar strategies to promote retention.

• Assuring proper protocol implementation is challenging.  The responsibilities of the project team were reorganized so that the supervisors (or integration agents) no longer have day-to-day responsibility related to data collection.  This task has been assigned to data collectors who have demonstrated excellence in their work.  The supervisors now have more time to promote the intervention, allowing more time to discuss and facilitate TST placement and problem-solve logistics.    Emphasis is being given to our seven largest units that together, contribute nearly 50% of the total number of patients
• TST uptake is not reaching the expected level. Despite training and individuals dedicated to promoting the intervention, administrative and logistical problems at the units have impaired greater advances in TST placement in HIV patients. On the one hand, THRio has the advantage of working in an established infrastructure which provides potential for sustainability. On the other hand, THRio deals with problems intrinsic to Brazil´s health infrastructure, especially decentralization, high turnover of HCWs, insufficient resources, and laboratory deficiencies.  To address the issue of uptake, we are planning a promotional campaign employing pharmaceutical-style marketing, such as detailing, clinical tools (PPD rulers, pocket-sized treatment cards), and educational symposia with luminary guest speakers from the Brazilian HIV establishment to encourage implementation of the TB/HIV guidelines.  In addition, we are considering whether the indicators established initially may need to be revised to take into account the difficulties of calculating the amount of time each patient contributes to the intervention.  This has created statistical challenges which prevent the most accurate picture of our progress.

 

ZAMSTAR:

• As discussed in last year’s report, there are a number of reasons why the interventions have been slower to be implemented than originally proposed:
  • Protocol violation during TST surveys in schools.
  • Delays from the ethics committee.
  • Overall scale of ZAMSTAR.
  • Local elections interfering with field work.
  • Underestimation of time needed to gain full buy in from communities.
  • Underestimation of time needed to train field staff.

• In order to test the effectiveness of the interventions, we need to be able to implement them consistently for long enough to make an impact on transmission of tuberculosis, which in turn will be reflected in a reduction in prevalence of disease.
• The Study rejected alternative primary endpoint measures, and our baseline data confirm that changes in notification rate or changes in prevalence of tuberculin sensitivity in school children are not appropriate endpoints to determine the potential impact of the interventions.
• The study needs to continue the interventions until the end of year 5, rather then the initial proposal which was to continue until the end of year 4.  This will delay the final prevalence surveys by a year to year 6 and the final year of analysis and writing to year 7.
• We are requesting an additional budget to cover the additional year of interventions.  This will be an exact duplicate of year 4 (US$ 3,516,015.72 – allowing for this year’s exchange rate).  In addition we will need to ask for a supplement to the biostatistics core to cover the ZAMSTAR specific costs of the extra year for statistical support

 

Biostatistics Core

• The Biostatistics Core has been updating power and sample size estimates based on accumulating data.  THRio prepared a special report to the DSMB, which showed that while TB incidence rates are slightly lower than anticipated,  theoretically, there would be sufficient power to detect 50% overall reductions in TB incidence.  Further analysis, however, indicated much of the TB would be occurring in those “ineligible” for the intervention due to prior INH use or TB.  Accordingly, the analysis plan will be supplemented with secondary analyses that take into account delays in uptake and focus on individual-specific effectiveness, as opposed to just clinic-level effectiveness.

 

Thibela:

The rate of enrolment into Thibela TB has been slower than expected, such that the enrolment rate has been around 30 per day per site rather than 150 as projected. In order to maintain high coverage of the intervention, the enrolment period has had to be increased; to minimize the effect of slower enrolment on study timelines, enrolment of clusters has been scaled up to enroll in three regions simultaneously. This has resulted in increased study costs and timelines.

In order to critically assess strategies to complete the study while containing costs and meeting timelines to the greatest degree possible, the Thibela study team undertook a formal evaluation of the feasibility, scientific impact, political consequences and financial costs of 11 alternative scenarios for completing the trial, which included changing the outcome measures, the number of clusters included and the study duration. Strategies to minimize cost and time considered included:

• Stopping the study now
• Doing what can be done within the initial budget and timeline
• Scaling up enrolment (with consequent increase in costs)
• Extending the enrolment period to maximize uptake of the intervention
• Doing an early outcome analysis to guide funding decisions
• Dropping elements of the study, such as TB Case Ascertainment Study (TCAS), primary measurement period and final culture prevalence survey, in order to reduce costs but without compromising the ability to evaluate the study.

The Thibela investigators believe that the best option to definitively answer the study question and influence public health policy is to enroll all eight intervention clusters as initially planned and to retain TB incidence as the primary outcome. At the end of Years 4 and 5 there would be data from 4 and 8 clusters, respectively, that will have completed the intervention (enrolment and treatment), at an additional cost of $2.46 million by the end of Year 4 and $6.71 million by the end of Year 5. Study progress can to be reassessed at the end of Years 4 and 5. Unless these assessments suggest that the intervention has not affected TB incidence, we propose that the study be allowed to continue to completion for a total additional cost above the original budget of $9.53 million.

The CREATE independent Data Safety and Monitoring Board (DSMB) reviewed preliminary data using TB case notification rates (CNRs) up to the end of May 2007 at its June 2007 meeting. The DSMB concluded that it is too early to use TB case notification rates (CNRs) to determine whether the intervention is having an effect or not. This may be possible when at least four intervention clusters have completed the intervention (i.e. the last person enrolled has completed treatment). We therefore propose using CNRs as an interim measure of the impact of Thibela TB on TB incidence, recognizing that there are several limitations to these data and there is a risk in using CNRs to guide a decision to stop the study:

• CNRs contain surveillance data, rather than verified case reports, and may overestimate actual TB incidence;
• if CNR data suggest a large effect of the intervention this could lead to stopping the study early. However, we would then be unable to assess the durability of the intervention, which is a critical measure of its public health significance. Both modeling and empirical data suggests that after a mass intervention incidence may rebound or even over shoot;
• a futility analysis based on CNRs could be potentially misleading, both because the data may include many NTM cases that are not affected by the intervention and because the time course of the impact of the intervention may be prolonged. While it could be argued that if a specified difference between intervention and control clusters is not reached the study will not achieve its goals an intervention which achieves a rapid reduction in TB incidence, even if the effect is smaller than our target, could be very important from a public health perspective.

• In spite of these pitfalls, we believe that interim review of CNRs by the DSMB is prudent, as the mining companies have access to these data and will be reassured knowing an expert, independent panel is evaluating the outcomes, and because the results could either bolster or refute the assumptions underlying the trial, leading to important modifications in the study’s conduct.

 

Cost and timelines

The original project timeline was 4.5 years and the budget was $14.1 million. The study encountered a significant delay in starting enrolment, largely due to the protracted negotiations with stakeholders and the time taken to establish the necessary infrastructure for very large community trials. Discounting foreign exchange losses, the project expenditure up to the end of year 3 was in line with the original budget. However due to scaling up enrolment and extending the enrolment period the cost of most alternative scenarios considered will exceed the initial total project budget.

The impact on timelines and costs of three of the alternative scenarios reviewed by the Executive Committee and the DSMB are presented below, and the full analysis is available on request. The financial implications are shown with each of timelines, and include the cost of including or excluding the TB Case Ascertainment Study, the principal means of determining true TB incidence.  The cost drivers include the study timeline, number of shafts and which elements of the study are included. A Year 4 budget based on the recommended scenario of eight clusters is presented at the end of this report.

The investigators believe that the best option to definitively answer the study question and influence public health policy is the scenario with eight intervention clusters, retaining TB incidence as the primary outcome. At the end of years 4 and 5 there would be data from 4 and 8 clusters respectively that have completed the intervention at an additional cost of $2.47 million by the end of year 4 and $6.71 million by the end of year 5. There is an additional $972,000 required to address foreign exchange losses incurred in the first three years of the study. Study progress will be reassessed at the end of years 4 and 5. Unless these assessments suggest that the intervention has not affected TB incidence, we propose that the study be allowed to continue to completion.

*Additional budget with/without TB Case Ascertainment Study

Key

 

E.  Lessons Learned

1.  CREATE is intervening in communities in unprecedented ways, affecting a wide variety of parties.  Stakeholder management is the key to the success of projects of this nature.

• All three studies are learning that ongoing stakeholder involvement through the duration of the projects is crucial and requires considerable energy and commitment. For example, to further engage community groups, THRio staff have been working with local leaders from TB and HIV NGOs to raise awareness of issues related to the study and more broadly for TB and HIV.

 

2.  CREATE’s studies are extremely ambitious and difficult.  The time to fully implement study protocols for extremely large community based trials may have been underestimated at the onset of the trials.

• Thibela TB: They will require an additional 24 months to fully reach their objectives.
• ZAMSTAR:  They will require an additional 12 months to fully reach their objectives.

 

3.  Training local field workers and clinical staff proving to be more time-consuming and laborious than anticipated.

• ZAMSTAR: Communities, especially vulnerable ones, need time and trust for interventions to work. Training of local field workers is an ongoing process and more time-consuming than anticipated.
• THRio: After discussions with HCWs, it became clear that the various steps required to place and read a TST, rule out active TB and start INH prophylaxis involve a lot of collaboration among different players at the health unit and other sectors (e.g. lab and x-ray).  THRio study supervisors are now working with the key players in each unit to try to understand barriers within the steps described above, and develop ways of streamlining or improving the process.

 

F.  Publications and Data Sharing.

Ayles H and Muyoyeta M.  Isoniazid to prevent first and recurrent episodes of Tuberculosis.  Tropical Doctor 2006; 36:83-86

Ayles H, Beyers N, Enarson D, Godfrey-Faussett P.  Surveillance studies and interpretation in The Tuberculosis Handbook (edited by Kaufman and van Helden – in press 2007).

Chaudhary, M.A., Moulton, L.H.  A SAS Macro for Constrained Randomization of Group-randomized Designs.  Computer Methods and Programs in Biomedicine, Vol. 83, 205-210, 2006.

Churchyard GJ. Working paper 1. Latent tuberculosis infection: burden and control. Report on the Scientific Working Group Meeting on Tuberculosis, 2005. TDR/SWG/06.

Churchyard GJ. Preventing TB & opportunistic infections among HIV-infected adults. Combined SA / Royal College of Physicians meeting, Cape Town, Feb 2007 (invited talk).

Churchyard GJ. IPT for silica exposed workers. South African Thoracic Society conference, Cape Town, April 2007 (invited talk).

Churchyard GJ, Scano F, Grant AD, Chaisson RE. TB preventive therapy in the era of HIV infection: Overview and research priorities. JID 2007 (in press).

Clark DA, Coetzee L, Crawford P. Data Management in Public Health. Phase Forward International User Group Conference, Las Vegas, USA, October 2006 (invited talk).

Clark DA, Coetzee L, Crawford P. Data Management in Public Health – Thibela TB Case Study. Phase Forward European User Group Conference, London, UK, May 2007 (invited talk).

Dorman SE, Chaisson RE.  From Magic Bullets back to the Magic Mountain: the rise of extensively drug resistant tuberculosis.  Nat Med 2007; 13:295-8.

Getahun H, Harrington M, O’Brien R, Nunn P. Diagnosis of smear negative pulmonary tuberculosis in people living with HIV/AIDS in resource constrained settings: informing urgent policy changes. Lancet 2007;16;369:2042-9 .

Golub JE, Saraceni V, Cavalcante SC, Pacheco AG, Moulton LH, King BS, Efron A, Moore RD,

Chaisson RE, Durovni B.  The impact of antiretroviral therapy and isoniazid preventive therapy on tuberculosis incidence in HIV-infected patients in Rio de Janeiro, Brazil.  AIDS 2007;21:1441-8.

Hanifa Y, Corbett EL, Abdool Karim S, Fielding K, Grant AD, Churchyard GJ. SIMRAC050801. Prevalence of latent TB infection among South African gold miners. 2007 (final SIMRAC report).

Hanifa Y, Molefe P, Chikwava S, Opperman L, Fielding K, Lewis J, Grant AD, Churchyard GJ. Prevalence of latent TB infection, radiological TB disease and silicosis among gold miners in South  Africa. IUATLD Congress, 2006 (invited talk).

Hanifa Y, Mngadi K, Lewis J, Fielding K, Churchyard GJ, Grant AD. Evaluation of the Arkansas method of urine testing for Isoniazid in a South African population. (submitted).

Hanifa Y, Mngadi K, Sefuthi C, Tautesile G, Mafube E, Mahlatsi N, Mantsoe I, Opperman L, Fielding K,  Lewis J, Churchyard GJ, Grant AED. Evaluation of Arkansas method of urine testing for Isoniazid in a South African population. IUATLD Congress, 2006 (invited talk).

Moulton, L.H: Power and Constraint: A Phased Implementation Cluster Randomized Study.  NIH/NIAID/OCR/Biostatistics Research Branch, Bethesda, 2007 (invited talk).

Moulton, L.H., Golub, J.E., Durovni, B., Cavalcante, S.C., Pacheco, A.G., Saraceni, V., King, B., Chaisson, R.E. Statistical Design of THRio: a Phased Implementation Clinic-randomized Study of a Tuberculosis Preventive Therapy Intervention.  Clinical Trials 2007;4:190-199.

Page K, Godfrey-Faussett P, Chaisson RE.  Tuberculosis and Human Immunodeficiency Virus Infection: in, Tuberculosis, a comprehensive international approach (edited by MC Raviglione, Marcel Dekker 2007).

Schaap, A:  Experiences with large HIV/TB survey implementations in Zambia. Conference on Surveying Health in Complex Situations, Brussels, 2007 (invited talk).

Sismanidis, B., et al. Restricted Randomization of a 2x2 Factorial Cluster Randomised Trial: Lessons Learned from the ZAMSTAR Experience (submitted).

WHO/HTM /TB/2007.376 WHO/HIV/2007.01. Improving the diagnosis and treatment of smear-negative pulmonary and extrapulmonary tuberculosis among adults and adolescents: Recommendations for HIV-prevalent and resource-constrained settings. WHO, Geneva, 2007.

 

Progress in Achieving a Global Access Strategy

The global strategies envisioned in CREATE’s mission are yet to be achieved, and challenges to the success of these approaches remain.  Reluctance to accept INH preventive therapy is widespread, and efforts to promote active or enhanced TB case finding have been weak in most countries.  Nonetheless, CREATE is making important progress in changing attitudes towards these strategies, and is providing an accruing body of evidence that will further transform global practices.  The THRio preliminary data which show that IPT in addition to ART significantly reduces TB risk have been broadly commented on and the just-published paper is being widely distributed.  The baseline prevalence data from ZAMSTAR will be published in the near future and should likewise have a substantial impact on policy discussions.  Overall, we feel that the atmosphere for changing practices and policies on TB/HIV has improved since the start of CREATE, but much work remains.

Specific progress at the project level is summarized below.

• Aurum has established partnerships with South African Business Coalition on HIV/AIDS (SABCOHA) and the Global Business Coalition. Through representation by Aurum on SABCOHA’s Board of Governors a link with Business Unity South Africa (BUSA) has also been established. Contact has also been established with the World Economic Forum and the International Labour Organisation’s HIV department. The use of IPT in the workplace has been actively promoted at all meetings. Thibela’s PI participated in an international workshop to develop guidelines for TB preventive therapy in silica exposed workers.
• Thibela has engaged regularly with the National Department of Health on TB / HIV issues and the use of IPT. Dr Churchyard has contributed to the national TB / HIV review, National TB strategic Plan and the National HIV strategic plan. Results of the study will be presented at relevant meetings of South African research institutions to build awareness within the local research community.
• Results and lessons learnt from the Thibela study have been presented at international conferences such as the World Lung Conference, the EDCTP Conference, and two Phase Forward International Users Conferences.  Principal Investigators play leading roles on a number of WHO committees and working groups where the Thibela experience and IPT strategies are shared.   
• ZAMSTAR has a strategy of engaging local and national institutions throughout the course of project. The strengthened PAC core will providing more time and expertise to prepare and disseminate information for use by local stakeholders.
• Presentation of THRio’s baseline data to Brazil’s National Tuberculosis Program (NTP) has resulted in a series of additional presentations around the country, following the NTP regional workshops agenda. This is contributing to the advocacy of INH prophylaxis in the country.  Upon publication of these results in AIDS, we will translate and disperse the results to key players in HIV/AIDS and TB control in Brazil.
• THRio has supported an international workshop with the participation of local and national NGOs, HCWs and CHWs groups to raise awareness about tuberculosis and research.

• A qualitative study addressing the knowledge and attitudes of IPT among health care workers has started, after receiving IRB approval.

 

IP & Regulatory Assurances Checklists

If nothing has changed from your previous submission, check here:  __X_No change   (Please still check off the appropriate boxes below)

1. Project has/will involve collaboration with for-profit companies	YES     NO
2. Project will involve use, further development or creation of technology that is or will be subject to intellectual property	YES     NO
3. Report contains proprietary information	YES     NO
4. If you checked “yes” for any of the above 3 statements, have any facts and circumstances changed from your last report/proposal submission?	YES     NO
5. Project has/will involve research using vertebrate animals	YES     NO
6. Project has/will involve research using human subjects	YES     NO
7. Project will involve clinical trials	YES     NO
8. Project will involve use of  any of the following substances: recombinant DNA pathogens/toxins identified as “select agents” by U.S. law: (http://www.aphis.usda.gov/vs/ncie/pdf/agent_toxin_list.pdf) biohazards or genetically modified organisms	YES     NO