Clinton Health Access Initiative (CHAI) — Community-Based Tuberculosis Household Contact Management (June 2024)

Note: This page summarizes the rationale behind a GiveWell grant to CHAI. It reflects our understanding at the time we recommended the grant in June 2024. CHAI staff reviewed this page prior to publication.

Published: October 2024

Summary

What we think this grant will do

Tuberculosis (TB) is a leading cause of infectious-disease related deaths worldwide, with the highest burden in India.1 Uttar Pradesh and Bihar, the two states in India where CHAI will implement this program, together account for roughly 30% of notified TB cases (i.e., diagnosed and reported TB cases) in India. We think this grant will directly avert about 2,900 cases of TB disease and 975 deaths among children under the age of 5 over the course of the grant period by:

• increasing uptake of TPT for asymptomatic children who are close contacts of people diagnosed with TB disease, and
• increasing detection and treatment of child contacts who already have TB disease (more).

We think the grant will also avert around 800 more deaths among children who would have survived TB but suffered from post-TB sequelae (i.e. complications from TB that can occur after TB treatment), and among older household contacts screened and referred as part of this program. We expect it to also improve later-life incomes for children and reduce illness-related costs for households (more).

Our grant will fund CHAI to provide technical assistance (TA) to Bihar and Uttar Pradesh's state TB programs and financial support to private sector patient support agencies to ensure that health workers visit households of index TB patients to screen household members for TB, provide preventive treatment, and make referrals for suspected active TB cases (more). Currently, our understanding is that the household contact management system in these states is largely facility-based (i.e., requires the index patient to report their household contacts and bring them into a facility for screening and treatment), which we believe imposes barriers to TPT uptake.

CHAI’s activities will include:

• Providing training and tools to build capacity among existing health workers to ensure household-level screening, TPT provision, and referrals can be carried out effectively (more)
• Providing financial incentives as part of contracts with private sector organizations to encourage health workers employed by these organizations to conduct contact tracing and household-level screening, TPT provision and referrals for families of private-sector TB patients (more)
• Refining operating procedures and improving data systems to enable government health workers to fit program activities into their daily routines, and to enable more effective monitoring of contact tracing and TPT provision.
• Hiring program officers to provide ongoing supportive supervision and monitoring of program activities;
• Procuring sufficient supplies of pediatric 3HP, which is a new child-friendly formulation of a short-course pediatric TPT drug regimen that CHAI plans to use in the treatment districts in the cRCT, after in-country drug registration and its inclusion in India’s TPT guidelines (more).

As part of this grant, CHAI will also commission IDinsight to evaluate the program through a large-scale cluster randomized control trial (cRCT) and process evaluation. The cRCT will randomize districts within states to receive the community-based program or to remain on the largely facility-based standard of care until the end of the 2-year evaluation period. It will answer several of our questions on TB HCM; the primary outcome it will estimate is the effect of the community-based program on the TPT initiation rate among all eligible under-5 children. (more)

Additionally, a small portion of this grant will support CHAI in market-shaping activities such as working with drug manufacturers to identify opportunities to further lower the price of the pediatric 3HP drug regimen and make it easier to procure the appropriate quantities of the two drugs in the regimen. We think this could further improve access and uptake of TPT in the future. (more)

Lastly, we think that with advocacy and support from CHAI, the government may take over the program after the grant period ends, leading to additional benefits from sustained higher TPT uptake (more). CHAI staff believe that at least some degree of transition is highly likely because the program largely works within the existing TB health system by leveraging the capacity of existing cadres of health workers (more). In the case that transition to government is only partial, GiveWell may consider funding CHAI to continue lighter touch technical assistance to maintain TPT coverage improvements. This could be a cost-effective use of funds over the longer term that would not be possible without this initial grant.

Why we made this grant

• The intervention has a strong intuitive case for its cost-effectiveness. The program is targeted at part of the population at high risk of dying from TB (unlike mass screen and treat programs): very young children who live in a household where an individual has TB. It also brings the health workers directly to people's homes, which substantially reduces the effort required for the target population to access a preventive treatment that halves their risk of getting TB. (more)
• Cost-effectiveness of the program is likely around our bar. Our CEA suggests this grant is ~8x as cost-effective as unconditional cash transfers, without including any benefits from transitioning the program to government ownership or from the option value of opening up room for more cost-effective funding to support this program in the future. If CHAI is successful in fully transitioning the program to government ownership, cost-effectiveness could be around 12x. (more)
• Market-shaping activities could generate additional benefits that we haven’t accounted for in our CEA. CHAI staff reported that there are opportunities to leverage this model of increasing TPT uptake to work with manufacturers to further reduce the price of the pediatric drugs in the 3HP TPT drug regimen (which were recently negotiated by the IMPAACT4TB Consortium). There may also be opportunities to develop a version of the drug regimen which would reduce supply chain disruptions. While we don’t expect any changes to products or prices to happen in time to affect the delivery of TPT through this grant, we think there’s a chance it will reduce costs and increase demand for these drugs in the longer term. (more)
• We expect to learn from this grant across several questions which could help us identify further cost-effective funding opportunities. These include:
  • The cost-effectiveness of community-based TB HCM. We will learn about this primarily through the evaluation. We suspect community-based TB HCM has a gap of roughly $100 million in cost-effective funding over the longer term ("room for more funding"), to which GiveWell could direct further investments. (more)
  • How market-shaping initiatives work, including the risks and challenges involved in executing these initiatives and predicting their impacts (more)
  • Other TB control efforts and the broader TB ecosystem. Since TB is one of the biggest causes of death in low-resource settings, this would better prepare us to decide whether to direct additional funding to TB programs in the future (more)
• CHAI seems well-positioned to implement this program. We believe that CHAI, through its India affiliate, William J Clinton Foundation (WJCF), has a good understanding of TB programming in Bihar and UP and well-established relationships with local stakeholders. WJCF has recent experience implementing TB programs across several states in India, in collaboration with local public, private and academic stakeholders, as well as experience implementing other health programs across India. (more)
• This program could influence the decisions of other stakeholders to fund similar programs. The results from the cRCT will significantly increase the amount of evidence available to the global TB community on community-based household case management, and shed light on whether promising results from an earlier RCT are generalizable. (more)

A summary of our cost-effectiveness model is shown in the table below. We expect CHAI’s program to increase TPT uptake among under-5 child contacts by 22 percentage points, at a cost of $35 per under-5 child reached. For these children, we estimate that TPT reduces the risk of getting TB by more than half, and that they would otherwise have a ~25% chance of dying from TB if they got it. We estimate the program generates further benefits by reducing deaths among children who already have TB, averting long-term health consequences of childhood TB, increasing the incomes of children later in life, reducing illness-related costs, and increasing TB diagnosis and treatment among older household contacts. Our cost-effectiveness model does not account for sustained impacts from transitioning the program to government. (More)

	Best guess	25th-75th percentile range on key parameters	Modeled cost-effectiveness over that range
Number of under-5 household contacts in target population over 2-year implementation period	~420,700
Share of contacts reached	60%	45-75%	6x - 9x
Cost per under-5 child reached	$35	$25-45	6x - 11x
Total grant amount	~$8,800,000
Among under-5 contacts reached, increase in TPT initiation rate	36pp	15-55pp	4x - 10x
Population-level increase in TPT initiation for under-5 child contacts	22pp
Prevalence of TB infection	30%
Risk of progression from infection to disease in absence of TPT	19%
Effectiveness of TPT initiation in preventing progression from infection to disease	56%	35-75%	5x - 10x
Cases of TB disease averted	~2,900
Case fatality ratio for TB disease in absence of treatment	35%	25-50%	6x - 11x
Overall case fatality ratio, accounting for treatment rates	26%
Deaths averted via prevention of disease progression	~750
Adjustment for deaths averted via increased treatment of TB disease	+30%
Total number of under-5 deaths directly averted	975
Moral weight for averted deaths	114
Initial cost-effectiveness estimate	4x
Benefit streams (% of total)
Direct mortality benefits	48%
Morbidity benefits from averted TB disease	1%
Benefits from averted post-TB sequelae	17%	3-30%	6x - 9x
Long-term increases in income	15%	4-23%	7x - 9x
Cost of illness averted	12%
Benefits for over-5 household contacts	7%
Additional adjustments
Adjustment for downstream costs to others and risk of replacing funding from other funders	-4%	-2 - -22%	6x - 8x
Cost effectiveness (x cash)	~8x

You can see the simple cost-effectiveness analysis for this grant here and the full version here.

Main reservations

Our primary reservations are:

• There's some chance that the program is below our cost-effectiveness bar due to limited evidence about key parameters that we won't learn about from this grant. In particular, we are uncertain about:
  • What proportion of children under-5 with TB die in the absence of treatment, and hence to what extent averting pediatric TB cases averts mortality.
  • How large the long term benefits to health and income are from averting childhood TB disease.
  • How effective 3HP TPT is at preventing TB disease in infants.

  If we are systematically over-optimistic about these uncertain parameters, we might be funding a program that in reality is meaningfully below our bar. (more)

• CHAI’s program requires buy-in from the government, and CHAI has not obtained that yet. While we believe it’s very (>90%) likely that CHAI will get government approval to implement the program, there’s some risk that this process will either take longer than expected or require CHAI to make changes that affect cost-effectiveness or the ability to run a cRCT. (more)
• CHAI will be leveraging the capacity of existing health workers, which could limit their ability to ensure CHWs conduct planned household visits. This program largely leverages existing cadres of health workers (whose time is already paid for). This seems to be a more sustainable program model than hiring a parallel workforce, and may also reduce costs. However, it increases the risk that the effect of the program on TPT uptake will be smaller than we expect because CHAI has less direct influence over CHW activities. We expect to learn more about this risk through the randomized controlled trial and qualitative data collection from health workers about their experience of this program. (more)
• These health workers may be more time-constrained than we expect. This would decrease the cost-effectiveness of this program if:
  • Health workers are not able to reach as many households as we expect; or
  • Health workers accomplish the volume of household visits we anticipate, but at the expense of other important health-promoting activities. (more)
• There’s a small risk that CHAI’s market-shaping activities could adversely affect the availability of TPT drugs, if some manufacturers agree to offer lower prices but this causes other manufacturers to drop out of the market. (more)

The organization

The Clinton Health Access Initiative (CHAI) is an international public health non-profit founded in 2002, which operates in 35 countries.2 GiveWell is the sole funder of the CHAI Incubator, a program within CHAI that exists to find, assess, and implement programs that have high cost-effectiveness and potential to scale. CHAI proposed this TB household contact management program through the Incubator.

CHAI has experience implementing similar programs to the one it will implement under this grant including TB programs to increase uptake of TPT within India and other countries, as well as programs that involve coordinating initiatives across different cadres of health workers. (more)

The intervention

Uttar Pradesh and Bihar, the two states in India where CHAI will implement this program, together account for roughly 30% of notified TB cases (i.e., diagnosed TB cases) in India.3 Our understanding is that currently the household contact management system in these states is largely facility-based (i.e., requires the index patient to report their household contacts and bring them into a facility for screening and treatment), which we believe imposes barriers to TPT uptake.4

Community-based household contact management (HCM) programs, on the other hand, involve health workers visiting households of people diagnosed with TB and screening their close contacts for possible TB. Those with symptoms of TB disease are referred for diagnosis and treatment at health facilities, while those without symptoms are given preventative treatment, TPT, which reduces the risk that they will develop TB disease. CHAI’s program will target household contacts under 5 years of age, since children under 5 are at relatively high risk of progressing to TB disease after becoming infected and less likely to be treated for it. We expect this intervention to substantially lower barriers to take-up of TPT by child contacts. In addition, it may identify additional cases of children who already have TB disease, increasing the likelihood that these children get treated. See our intervention report for more details.

CHAI's program will utilize a pediatric formulation of the 3HP TPT regimen, which involves one weekly dose of a combination of two drugs (isoniazid and rifapentine) for 12 weeks.5 This is a significantly shorter and less frequent regimen than the common alternative, which requires a daily dose of isoniazid for 6 months or more.6 The pediatric formulation, which is taste-masked and dispersible, should also make the drugs easier to administer to young children.7 We expect the short-course regimen and the pediatric formulation to have a positive effect on the TPT uptake and completion rates.

The grant

What we think this grant will do

We expect the grant to accomplish the following:

• Avert about 2,850 cases of TB disease and about 971 deaths among under-5 children over the grant period. It would do this by:
  • Increasing uptake of TPT for asymptomatic child contacts of people diagnosed with TB disease; this would prevent some children from getting TB disease and dying as a result (accounts for about three-quarters of short-term deaths averted).8
  • Increasing detection and treatment of child contacts who already have TB disease; this would prevent some children from dying as a result (accounts for about one-quarter of short-term deaths averted).9
• Generate additional benefits by preventing TB disease in children. These benefits are reduced illness-related costs for families, increased later-life health (as a result of averted post-TB sequelae) and increased income later in life.10
• Generate additional benefits for older household contacts who are screened for TB during household visits. By screening and providing referrals for contacts age 5 years and older, the program could also reduce TB disease and its consequences among this population, though we expect these benefits to be much smaller than for under-5 contacts.11
• At least partly transition the program to government ownership. If the program operates successfully in the first two years, CHAI plans to advocate for and support the government to take over the program. CHAI staff believe that at least some degree of transition is highly likely because the program largely leverages the capacity of the existing TB health system and workforce.12 (more)
• Reduce the price and improve the supply of pediatric 3HP (the short-course TPT regimen that CHAI proposes using for this program), both within and outside India, via market-shaping activities. (more)
• Generate rigorous evidence about the effect of community-based HCM programs, via the large-scale cRCT and other monitoring and evaluation (M&E) activities. (more)

Main grant activities

Program activities

The grant will fund CHAI to provide technical assistance, procure TPT and directly finance some elements of the community-based HCM program.13 These activities will support public-sector health workers and staff of private sector agencies to visit households with children under 5 to screen household contacts of TB patients, and to provide preventative TPT treatment for under-5 contacts without TB symptoms.14 Any household members (regardless of age) who screen positive for TB symptoms will be referred to health facilities for further testing and treatment.15

To enable health workers to effectively conduct these household screening, CHAI staff will undertake the following activities:

• At the district level:
  • Build capacity among existing community health workers and their supervisors to deliver household-based activities. Specifically, CHAI staff will train three cadres of community health workers (CHWs): TB health visitors (known as TB-HVs, who are employed by state TB programs to support public-sector patients in urban areas), ASHAs (who are contracted by the government to cover patients in rural areas) and Patient Provider Support Agency (PPSA) health workers (who are employed by private agencies to cover private-sector patients in urban areas).16
  • Support adoption of centrally-developed tools and processes to optimize CHW time use to ensure CHWs can incorporate household visits into routine activities17
  • Call index TB patients (TB patients known to the health system) to schedule and monitor home visits.18
  • Routinely visit health centers to provide ongoing supportive supervision and monitoring of program activities19
  • Support periodic review meetings with district-level TB Officers and Managers by compiling performance data, setting agendas and suggesting opportunities for improvement.
  • Engage district medical officers and other officials to ensure household contact management activities get prioritized.
  • Engage with district officials to enable evaluation activities.20
  • Refining operating procedures and improve data systems to enable government health workers to fit program activities into their daily routines, and to enable more effective monitoring of contact tracing and TPT provision.21
• At the state level22 :
  • Monitor activities of district-level staff and provide coaching to ensure performance standards.
  • Support periodic review meetings with state-level officials by compiling performance data and setting agendas.
  • Engage state-level leadership to ensure HCM activities are prioritized in the selected districts.
  • Strengthen the quantification and supply chain processes for TPT drugs to reduce the chances of TPT shortages once CHAI stops procuring TPT drugs.
  • Work with state-level staff to streamline CHW incentive payments for TPT initiation, which are already funded by the government.
  • Advocate for the states to oversee and finance CHW activities across all districts, and facilitate transition of project activities from CHAI staff to government and PPSA staff. This would occur first in the treatment districts (while CHAI is setting up the program in control districts after the evaluation period), and CHAI will use this experience to inform the level of support needed to facilitate transition of the program in control districts at the end of the grant.23
• At the national level:
  • Undertake market-shaping activities, such as working with drug manufacturers in India to find opportunities to further reduce the price of pediatric 3HP and reduce supply chain complexity for buyers of the drug in the global market following the grant period.24

CHAI will finance and support procurement of a dispersible pediatric formulation of 3HP, after in-country drug registration and inclusion in India’s TPT guidelines, to cover household-based TPT provision in districts that will receive the program in the first half of the grant (the treatment group in the randomized controlled trial – see evaluation activities below). CHAI expects that the dispersible pediatric formulation of the medication will be easier to administer to young children.25 In the second half of the grant, CHAI expects that the Indian government will procure the dispersible pediatric formulation without support from CHAI. CHAI has informed us that the non-dispersible adult-formulation of the drug (which can be crushed to provide to children) has already been procured by the Indian government and will be available in control group districts prior to large-scale procurement of the dispersible pediatric formulation.26

CHAI will finance additional incentive payments for PPSAs to conduct HCM and TPT provision for households of private-sector index TB patients.27 CHAI has told us that private-sector TB patients account for 50% of patients in Bihar and 28% in UP, suggesting that these incentives cover a substantial amount of the overall staff cost for household-based activities.28 CHAI has told us that additional incentive payments will not be necessary for the other cadres of health workers, either because the government already funds TPT-related financial incentives (in the case of ASHAs) or because (facility-based) HCM is already part of their job description (in the case of TB-HVs).29

Evaluation activities

In addition to programmatic activities, CHAI will contract IDinsight to evaluate the program via a cluster-RCT (cRCT) and process evaluation. The cRCT will randomize districts within states to receive the community-based program or to remain on the facility-based standard of care until the end of the 2-year evaluation period and will estimate the effect of the community-based program on the TPT initiation rate among all eligible under-5 children as the primary outcome.30 The evaluation will also answer several other questions on our learning agenda for TB HCM.31 More details about the cRCT can be found below.

The cRCT will not capture any effect of CHAI’s state-level technical assistance activities because state-level activities would affect both implementation and control districts. However, we perceive the state-level technical assistance to be a more minor part of the program. To address the possibility that these state-level activities affect TPT uptake in both control and treatment districts, the evaluation will also measure the change in state-level TPT uptake over time (via baseline and endline surveys) and collect qualitative data to understand the role of CHAI’s activities and other factors in contributing to the change in TPT uptake over time.

At the end of the evaluation period, CHAI will spend another 2 years setting up and rolling out the program to control-group districts.32 CHAI will ultimately roll out the program in 26 districts in Bihar and 47 districts in Uttar Pradesh, prioritizing districts with higher TB notifications.33 This accounts for roughly two-thirds of the districts in each state. While rolling out the program to control-group districts, CHAI will also work with the states to transition the program in treatment-group districts to government ownership.34

Budget for grant activities

The total budget for the grant is $15.1 million. It is broken down as follows:35

Programmatic costs
Personnel	$5,234,618
Trainings, Review meetings and Supervisory visits	$1,584,490
Rewards and Incentives	$819,108
Other direct - Office, telecom and team travel	$945,920
M&E (tool development and formative research)	$230,000
Subtotal	$8,814,136
Evaluation costs
Impact and process evaluation (IDinsight)	$ 2,651,762
CHAI staff to coordinate and support evaluation	$ 561,926
Supplemental M&E conducted by CHAI on top of main evaluation	$ 200,000
Fees for local co-investigator	$ 240,377
Subtotal	$3,654,064
Other direct costs - Market shaping and procurement
Market shaping	$213,000
TPT procurement	$643,123
Subtotal	$856,123
Total Direct Cost	$ 13,324,322
Indirect Costs (15%)	$ 1,504,416
Indirect (8% on pass through on contractor costs)	$ 263,591
TOTAL COST	$ 15,092,329

The case for the grant

The intervention has a strong intuitive case for its cost-effectiveness

We think this program has a strong intuitive case for being more cost-effective than other TB programs we've considered:

• it targets a segment of the population (i.e. under-5 household contacts) at high risk of dying from TB, which explains why it’s likely to be more cost-effective than other TB programs we’ve considered, which are less targeted;36
• we have a fairly strong prior that taking preventive health services and drugs to households, rather than asking individuals to come to facilities, will substantially increase timely uptake of preventive treatment;37
• it is aimed at increasing uptake of a drug (TPT) which seems to substantially reduce the risk of recently exposed child contacts developing TB disease38 , and is also fairly cheap given recent market-shaping efforts for short-course regimens (~$5 per course)39 ;
• it can use the newly available 3HP TPT regimen40 , which should increase initiation and completion rates since it is shorter in duration (12 weeks vs. 6+ months) and frequency (once a week, vs. daily) than older regimens, and the dispersible pediatric formulation is easier to administer than earlier formulations of 3HP41 ;
• by preventing TB disease in young children, the program probably also has other benefits aside from directly averting deaths from TB: averting post-TB sequelae which can cause disease and early death. Averting this morbidity also reduces household illness-related costs and likely leads to higher income later in life.42
• by having health workers visit households of TB patients and screen contacts for symptoms of TB, this program may also increase (or at least bring forward) detection and treatment of children and adults who have already progressed to TB disease because the health workers would refer household contacts with signs of TB to facilities for further testing and treatment.43

Cost-effectiveness is likely around our bar

Our cost-effectiveness analysis implies CHAI’s program in Bihar and Uttar Pradesh is 8 times as cost-effective as cash transfers. In the first two years of implementation we expect CHAI’s program to increase TPT uptake among under-5 children by 22 percentage points over what would have happened otherwise. We estimate this will cost $35 per under-5 child reached. For these children, we estimate that TPT reduces the risk of getting TB by more than half, and that they would otherwise have approximately a 25% chance of dying from TB if they got it. Our analysis also accounts for the program generating additional benefits by reducing deaths among children who already have TB, averting long-term health consequences of childhood TB, increasing the incomes of children later in life, reducing illness-related costs, and increasing TB diagnosis and treatment among older household contacts.44

We think our analysis probably underestimates the cost-effectiveness of CHAI’s implementation model since it does not account for CHAI’s intention to transition the program to the government by the end of the grant. If successful, we expect this would maintain TPT uptake at a relatively high level after CHAI concludes their grant activities. We think that some degree of transition to government is likely because CHAI’s implementation model largely leverages existing resources to implement program activities (more), but are highly uncertain about the extent of this transition and how long it will take. If CHAI is able to transition all elements of the program to government ownership using this grant, their program could be around 12 times as cost-effective as cash transfers (see details of this estimate in the intervention report). If the program is only partially transitioned to government ownership, we may consider funding CHAI to provide lighter-touch technical assistance to maintain the benefits of the program over a longer period. We expect that this use of funds could be even more cost-effective than this initial grant since it could generate similar benefits at a lower ongoing cost.

Market shaping activities could generate additional benefits

CHAI plans to work with manufacturers to find opportunities to further lower the price of the pediatric formulations of the drugs in the 3HP TPT regimen.45 CHAI expects there may be an opportunity to support development of a version of these drugs which is easier to procure in the right quantities, thereby reducing supply chain complexities.46 Together, they expect these changes in the drugs available on the market to increase the volume of these drugs procured for HCM programs,47 ultimately increasing uptake of TPT for child contacts. While we don’t expect these market outcomes to materialize in time for the program funded by this grant (which will procure 3HP at current prices) the benefits would accrue in the longer-term to government- and NGO-run programs procuring pediatric 3HP.

We haven’t attempted to model the size of the benefits from potential reductions in price and increased demand for 3HP, nor the likelihood of these outcomes materializing, but given the potential size of the pediatric TPT market it seems plausible that the benefits would easily justify the relatively small cost of funding these market-shaping efforts.48

We also expect this part of the grant to have significant learning value for GiveWell. By making our first grant with an explicit market-shaping component that we plan to follow throughout the grant period, we will be able to learn about this potentially cost-effective type of intervention. Our impression is that CHAI is a global leader in shaping health commodities markets, and so are a particularly good partner for us to learn from.

We expect to learn from this grant across several important questions

The cost-effectiveness of community-based HCM

As part of this grant, CHAI will commission IDinsight, an external evaluator, to design and carry out a cluster-randomized controlled trial (cRCT), process evaluation and study of community health worker time-use to understand the effects of the program.49

Through this evaluation we expect to learn about the effect of the program on the rate of TPT uptake among under-5 household contacts (the primary outcome) as well as several other important outcomes including50 :

• TPT completion rates
• The percent of target households reached by health workers
• Health workers’ time use
• The rates of referrals to health facilities for active TB cases
• Program cost per child reached

The two-arm cRCT will compare the community-based HCM set up by CHAI against the existing facility-based standard of care. Randomization will occur at the district-level, with 26 districts randomized in Bihar and 44 in Uttar Pradesh, and will be stratified by district-level characteristics.51

The primary outcome for the cRCT is the percentage of under-5 child contacts of pulmonary TB patients who initiated a course of TPT.52 This will be measured through surveys of caregivers in households of recently diagnosed TB patients, and will also be validated through alternative measures.53 In addition to TPT initiation, the cRCT will measure TPT completion rates, the percentage of households visited by a CHW for TB screening, and referrals for active TB investigation, among other things.54 Outcomes will be measured at baseline and at a ~1.5-year endline (in repeated district-level cross-sections).55

The planned sample size for measuring the effect on TPT initiation is close to 1,600 children aged under-5 who are household contacts of recently diagnosed TB patients.56 We expect that this will allow us to detect an effect on TPT initiation of approximately 15 percentage points at the state-level and approximately 10 percentage points across states (the minimum detectable effect sizes (MDES)).57 This gives us confidence that the cRCT is sufficiently powered to inform us about the cost-effectiveness of this program both because:

• These MDES are smaller than what was observed in a recent study of community-based HCM.58
• If the average effect size was less than 10 percentage points, we would be unlikely to fund this program in the future, even if we lowered our cost-effectiveness bar.59

The process evaluation will complement the cRCT by helping us understand what happened during program implementation. It will be based on semi-structured interviews with a range of stakeholders, including TB patient households, community health workers, government officials, and CHAI staff.60 In addition to the main cRCT and process evaluation, the study will also examine the effect of the program on community health workers via time-use surveys and a shadowing exercise.61

By generating new evidence on these uncertainties, we expect to update our view about the promisingness of this program, which we suspect has a gap of roughly $100 million in cost-effective funding over the longer term ("room for more funding"), to which GiveWell could direct further investments. The "room for more funding" estimate includes the possibility of providing ongoing cost-effective funding to CHAI to support the program in Bihar and Uttar Pradesh if it is not fully transitioned to government ownership during this initial grant.62

This evidence may also inform our view of community-based approaches to preventive health treatment more generally, which could have even broader implications for how GiveWell directs future funding.

How market-shaping initiatives work and CHAI’s approach to market-shaping

Market shaping is a broad category of intervention that aims to improve access to health commodities through market coordination strategies.63 CHAI reports being involved in 135 global market shaping projects that have reduced the price of health commodities and improved access in low-income countries.64 This grant provides the opportunity for GiveWell to learn from CHAI about the risks, benefits, and principles of market shaping initiatives. This will occur through structured workshops delivered by CHAI on market shaping and commodity quality assurance methods, and by GiveWell staying updated on CHAI's progress improving the 3HP market as part of this grant.65 We expect this to have high learning value because market-shaping is a potentially highly cost-effective type of intervention that GiveWell has not previously funded.

TB control efforts and the broader TB ecosystem

TB is one of the leading causes of death in low-resource settings, with potentially large room for more funding.66 Because this will be GiveWell's first large grant towards a TB program, we expect that it will enable us to learn more effectively about TB, which will in turn help us decide whether to direct additional funding to TB programs in the future.

We plan to learn more about the TB ecosystem through this grant by:

• Sharing our intervention report with experts, implementers and other grantmakers to get exposure and feedback on our views about this program model;67
• Continuing conversations with other TB funders and experts and potentially participating in TB-related stakeholder fora;
• Seeing how the public TB program in India coordinates with CHAI's philanthropically funded program;
• Monitoring the grant progress and outcomes to potentially learn about unanticipated factors that would contribute to or hinder the successful implementation of a TB program.

CHAI seems well-positioned to implement this program

We believe that CHAI, through its India affiliate, William J Clinton Foundation (WJCF), has a good understanding of TB programming in Bihar and UP and well-established relationships with local stakeholders. WJCF has recent experience implementing TB programs across several states in India, in collaboration with local public, private and academic stakeholders, as well as experience implementing other health programs across India.

Most recently, WJCF led a 6-year TB program across several Indian states, including Bihar and Uttar Pradesh, as part of Project JEET and JEET 2.0.68 We believe that implementing these projects has given WJCF both a good understanding of the current state of HCM and TPT provision in these states and an opportunity to solidify their relationships with both private sector and public sector TB partners.69 CHAI and WJCF noted that the challenges associated with facility-based HCM, observed in Project JEET, informed their proposal for this program.70

Beyond Project JEET, CHAI reports its affiliate WJCF has long-standing experience implementing public health programs in India, including in Bihar and Uttar Pradesh.71 We think that well-established relationships with government and other health system stakeholders through this work positions CHAI and WJCF well to successfully implement this program.

More broadly, we continue to have a strong qualitative impression of CHAI as an implementer. CHAI stands out as a strong partner due to its successful track record scaling very large, evidence-based programs, its extensive experience working in low- and middle-income countries, and its global footprint.72 Our regular collaboration through the CHAI Incubator has strengthened our confidence in our view of the organization. We have found that the CHAI team answers our questions with exceptional rigor and provides high-quality feedback, causing us to update and correct our work.

This program could lead other stakeholders to fund similar programs in other geographies.

We are currently aware of only one other randomized control trial of the effect of community-based TB household contact management on TPT uptake in children (Bonnet et al 2023). We therefore expect the cRCT funded by this grant to significantly increase the amount of evidence available to the global TB community on this approach to increasing TPT uptake. The results of the cRCT, whether positive or negative, could therefore influence:

• whether other Indian states adopt a community-based approach to TB HCM;
• whether community-based TB HCM is implemented elsewhere beyond India. This is because the cRCT funded by this grant would improve the external validity of the evidence base, since it will be implemented in a different geography to the Bonnet et al 2023 study (northern India rather sub-Saharan Africa73 ), using a somewhat different implementation model (with community health workers initiating TPT rather than nurses74 ) and different drug regimen (the weekly 3HP regimen rather than the daily 3HR regimen75 ).

Risks and reservations

There’s some chance the program is meaningfully below our bar and we won’t find out about this through the grant

We’re particularly uncertain about the following parameters, and the grant will not enable us to learn about them. If we’re systematically off in a particular direction across some/all of them, cost-effectiveness of CHAI’s program could be more like 5x rather than around 10x.

• Pediatric TB case fatality ratio (CFR) in the absence of treatment (best guess: 35%). Our uncertainty is mainly due to the quality of studies from the pretreatment era (which is the only source of evidence about case fatality in the absence of treatment, given ethical considerations)76 as well as some residual uncertainty about the extent to which near-universal BCG vaccine coverage reduces under-5 CFR.77 We may revise this parameter in the future if new research becomes available.
• Benefits from averted post-TB sequelae (best guess: 17% of total benefits). We rely largely on a single modeling study of the impact of post-TB sequelae on health, and the evidence on sequelae from pediatric TB is limited. Quantifying these benefits also relies on assumptions about when in life these sequelae tend to manifest. We've developed what we think is a conservative best guess, but there’s still a chance that we’ve put too much weight on a limited evidence base.78
• Long-term increases in income from averted TB disease (best guess: 15% of benefits in CEA). We benchmark these increases to those observed in studies on averting childhood malaria, using the method described here for our analysis of Vitamin A supplementation. We think it’s reasonable to assume, as a first approximation, that the income benefits from averting TB are similar to those from averting malaria, since our impression is that TB disease tends to last longer than malaria and produce comparably severe symptoms.79 However, we haven’t searched for evidence on the effects of TB on what we think are the mechanisms for long-term income effects (e.g. growth deficits, cognitive impairments and disabilities), so may be overestimating these effects.
• The effectiveness of 3HP TPT among young children in preventing TB disease (best guess: 56% risk reduction). Our uncertainty is driven by the limited evidence about the effectiveness of TPT for infants,80 and lack of dosing guidelines for children under 2 for 3HP.81 To arrive at our best guess, we considered two meta-analyses of the effect of TPT on incidence of TB in children, and adjusted the effect estimate from one of these studies (Ayieko et al. 2014) to account for several external validity considerations relating to expected differences in effectiveness between different TPT regimens and across different age groups.82 However, we are quite uncertain about each of these adjustments, so may be overestimating the effect of 3HP on TB incidence for child contacts under the age of 5 with a TB infection.
• The risk of reinfection and death following completion of TPT. Due to reinfection risk in high-burden settings, we think there’s a small chance that short-course TPT merely delays TB child deaths by a few years rather than averting them.83 This risk should be mitigated in CHAI’s program to the extent that the re-infection risk is due to exposure to new notified cases of TB in the child’s household, since these children should receive another course of TPT in this event. However, there may be some residual risk due to other channels of re-infection. While we have already attempted to account for this risk in our estimate of the effectiveness of 3HP, we may be underestimating it.84

Government buy-in isn’t fully secured yet

CHAI’s program requires buy-in and collaboration with government officials, including permissions from government officials to access Ni-kshay data as well as some revisions in TPT guidelines to introduce pediatric 3HP. Our understanding is that at the time of making this grant CHAI has not fully discussed and agreed on program plans with these officials.85 CHAI reports that confirmation of project funding is usually required to advance the process of obtaining buy-in from senior government officials.86

Because of this, there is a small risk that CHAI will fail to obtain government buy-in and moderate risk that the process of obtaining buy-in will delay grant activities or require CHAI to modify their proposed activities to be more amenable to government officials in a way that reduces the cost-effectiveness of the program.

We think it is very unlikely that CHAI will fail to obtain government buy-in for the program, given their recent experience working with state governments in Bihar and UP87 , their long history working with the Indian government,88 and the alignment of this program with India's current goals and guidelines for TB prevention.89 In the event that CHAI cannot obtain buy-in to implement the program and evaluation with core features intact, we will have the option of requesting a return of funds.

CHAI will be leveraging the time of existing health workers, which could limit their ability to ensure CHWs conduct planned household visits

We think that CHAI’s decision to leverage existing CHWs for household visits is preferable to hiring a parallel set of health workers. This is because it increases the chances of the program being sustained by the government after CHAI leaves and because this seems like a more scalable and cost-effective implementation model than one which requires additional workers to be hired. This view was also expressed to us by multiple external stakeholders, including government officials.90

However, given CHWs’ other responsibilities and priorities, it may be more difficult to maximize the number of households reached and the quality of household visits via this approach than it would be if CHAI hired a parallel set of health workers.91 This could mean that the effect of the program on TPT initiation will be lower than we expect.

We have tried to account for this risk by moderating our best guesses for the percent of households that CHWs will reach (60%)92 and the TPT initiation rate for households who are visited (80%). We expect to learn more about this risk through the randomized controlled trial and qualitative data collection from health workers about their experience of this program.

CHAI staff at the state and distinct levels hope to enable CHWs to reach as many households as possible with high quality care by: by i) ensuring that CHWs have the time, skills, tools, and resources necessary to undertake program activities and ii) working with CHW’s managers to encourage course corrections for the CHWs who do not meet expectations.93 These strategies seem plausibly effective to us and similar to strategies used by other NGOs who successfully implement programs using government workers (such as Malaria Consortium’s SMC programs, one of GiveWell's top charities).

We think that CHAI’s substantial previous experience of partnering with public and private sector organizations, including for TB projects in India94 , suggests they will be able to successfully execute these strategies.

Health workers might be more time-constrained than we expect

We may be overly optimistic about:

• How many households CHWs can reach given the number of CHWs and their existing workload.95
• The opportunity costs of CHW time. For instance, CHWs may prioritize household screening over TB patient treatment or other important health-promoting activities. Although we apply a rough leverage adjustment in our cost-effectiveness model to account for these opportunity costs, we may still be underestimating their magnitude.

We think we can learn about this risk through the grant, and have incorporated it into our learning agenda96 and evaluation plans.

Risks associated with CHAI’s market-shaping plans

We think there’s a slight risk that if existing manufacturers agree to lower the price of the drugs in the 3HP regimen, it could have a longer-term effect of reducing the number of manufacturers in the market by making it unprofitable for manufacturers with higher costs to supply 3HP. This potential increase in monopoly power for existing manufacturers could result in the price of 3HP being higher in future than it would have been in the absence of the short-term price reductions.

We think this risk is small because we put significant weight on CHAI’s judgment that the price reductions they are targeting are sustainable (based on their in-depth understanding of the costs of production from previous market shaping work).

We plan to learn more about this risk by engaging with CHAI about how they set price targets through this grant's follow up and our shared market shaping learning agenda.97

Plans for follow up

We will have regular meetings with the CHAI team to discuss progress of the program and cRCT over the course of the grant period. We expect to receive preliminary results of the cRCT in late 2028, and will consider this program for renewal or expansion once we receive those results.

Internal forecasts

For this grant, we are recording the following forecasts:

Prediction	By time	Likelihood
State governments provide written approval to CHAI and its affiliates to implement the program and evaluation with core features intact	Jan 2025 / Dec 2025	70% / 90%
CHAI and IDinsight receive all necessary ethical and regulatory approval to run the evaluation of the program	Jan 2025 / Dec 2025	70% / 90%
CHAI contracts PPSAs (and agree to financial incentives) for TPT-related HCM of private sector patients	Jul 2025 / Jul 2026	70% / 90%
CHAI has procured at least 125,000 courses of pediatric 3HP TPT (the rough quantity we expect will be required for treatment districts in first two years, assuming 60% of households are reached)	Jul 2025 / Jul 2026	60% / 80%
IDinsight completes baseline data collection in all treatment and control districts	Dec 2025 / Jul 2026	75% / 85%
CHAI and its affiliates complete initial round of capacity building with CHWs (including PPSA staff), supervisors and medical officers in treatment districts	Dec 2025 / Jul 2026	60% / 75%
CHAI and its affiliate staff begin conducting monitoring and supervision visits to check CHW activities in treatment districts	Dec 2025 / Jul 2026	60% / 75%
IDinsight completes endline data collection in all treatment and control districts	Dec 2027 / Jan 2028	50% / 70%
At least 55% of households of PTB patients notified in the past year with children under-5 in treatment districts have been visited for TB screening (our best guess of program reach in year 1)	1 year after implementation begins	50%
At least 65% of households of PTB patients notified in the past year with children under-5 in treatment districts have been visited for TB screening (our best guess of program reach in year 2)	2 years after implementation begins	50%
At least 44% of under-5 households of PTB patients notified in the past year in treatment districts have initiated TPT (our best guess of TPT coverage in year 1)	1 year after implementation begins	50%
At least 52% of households of PTB patients notified in the past year with children under-5 in treatment districts have initiated TPT (our best guess of TPT coverage in year 2)	2 years after implementation begins	50%
CHAI will exit from supporting the government to implement this this program in both Bihar and Uttar Pradesh	Dec 2029	25%
Conditional on CHAI requiring additional funding to continue supporting the government to implement this program in Bihar or Uttar Pradesh, GiveWell will provide a renewal grant to continue supporting this program in the relevant state(s)	Dec 2029 / Dec 2030	40% / 50%
our estimate of the cost-effectiveness of the program, as modeled in our CEA, will increase by more than ‘1x cash’	1 year after evaluation results are available	40%
our estimate of the cost-effectiveness of the program, as modeled in our CEA, will decrease by more than ‘1x cash’ by December 2027, or whenever evaluation results are available	1 year after evaluation results are available	40%
conditional on the cRCT being run successfully, the point-estimate from the cRCT of the effect of the program on TPT coverage will be X (note: current best guess is 22pp), where X is: <15pp 15-25pp 25-35pp >35pp	When evaluation results are available	25% 45% 25% 5%

Our process

• This opportunity came to us through the CHAI Incubator.
• We researched and published an intervention report on community case management for child contacts of TB patients in September 2024.
• We regularly communicated with CHAI to better understand programmatic activities, evaluation plans, and market-shaping proposal, while conducting a cost-effectiveness analysis.
• We spoke with government stakeholders, academic experts, and representatives of large TB funders to better understand the risks and benefits of this program.

Relationship disclosures

Neil Buddy Shah was hired in April 2022 as CHAI’s CEO. Previously, he was GiveWell’s Managing Director.

Sources

Document	Source
Ayieko et al. 2014	Source
Bonnet et al. 2023	Source (archive)
CHAI staff, conversation with GiveWell, April 11, 2024 (unpublished).	Unpublished
CHAI, "About Us"	Source (archive)
CHAI, “Homepage”	Source (archive)
CHAI, “Market Shaping.”	Source (archive)
CHAI, "Proposal to GiveWell," September 2023 (unpublished).	Unpublished
CHAI, "Quality and Market shaping learning agenda," 2024 (unpublished).	Unpublished
CHAI, "TB HCM program budget," July 5, 2024.	Source
CHAI, "TB HCM program proposal," May 1, 2024	Source
Dr. BK Mishra, State TB Officer, Bihar in conversation with GiveWell, January 2024. (unpublished)	Unpublished
Dr. Vinod Kumar, WHO consultant to Uttar Pradesh’s State TB program in conversation with GiveWell, January 2024 (unpublished)	Unpublished
GiveWell, "Clinton Health Access Initiative – CHAI Incubator," August 2022	Source
GiveWell, "Internal forecasts"	Source
GiveWell, "Intervention Report on Household contact management for tuberculosis," 2024.	Source
GiveWell, "Mass Screen-and-Treatment Programs for Tuberculosis"	Source
GiveWell, "Our Top Charities"	Source
GiveWell, copy of TB HCM RFMF analysis from implementer	Source
GiveWell, Learning from CHAI's TB HCM pilot in India (unpublished)	Unpublished
GiveWell, Market-shaping opportunity in CHAI TB grant, 2024 (unpublished).	Unpublished
GiveWell's CEA of a community-based TB HCM program in India	Source
IDInsight, “Concept note: Impact Evaluation of CHAI’s Household Contact Management for Childhood Tuberculosis Program in India,” May 2024.	Source
IMPAACT4TB	Source (archive)
Joint Effort for Elimination of Tuberculosis (JEET), “About”	Source (archive)
Ni-kshay, TB Notification report	Source (archive)
Project JEET	Source
Rethink Priorities, An overview of market shaping in global health: Landscape, new developments, and gaps, September 1, 2023.	Source (archive)
Reuben Swamickan, Deputy Director, USAID India in conversation with GiveWell, January 2024. (unpublished)	Unpublished
Unitaid, "New Child-Friendly Formulation of Rifapentine for Short Course Tuberculosis Prevention Treatment Now Available as Unitaid and IMPAACT4TB Launch an Early Market Access Vehicle."	Source (archive)
Unitaid, "Short-course Treatment Regimens to Prevent TB: 3HP and 3RH"	Source
WHO, “WHO consolidated guidelines on tuberculosis,” 2020,	Source (archive)
World Health Organization, “Financing for TB prevention, diagnostic and treatment services,” 2023.	Source (archive)
World Health Organization, “Tuberculosis (Fact sheet),” 2023.	Source (archive)
World Health Organization, India, “Tuberculosis,” 2023.	Source (archive)

• 1
  • “A total of 1.3 million people died from TB in 2022 (including 167 000 people with HIV). Worldwide, TB is the second leading infectious killer after COVID-19 (above HIV and AIDS).” World Health Organization, “Tuberculosis (Fact sheet),” 2023.
  • “India has the highest burden of TB with two deaths occuring every three minutes from tuberculosis (TB).” World Health Organization, India, “Tuberculosis,” 2023.

• 2
  • "CHAI was founded in 2002 with a transformational goal: help save the lives of millions of people living with HIV/AIDS. Today, we support government priorities across many areas, including other infectious diseases, non-communicable diseases, women and children’s health, health workforce and financing, assistive technologies for disabilities, and the intersection of climate and health." CHAI, "About Us"
  • “CHAI operates in over 35 countries around the world. More than 125 countries have access to CHAI-negotiated price reductions for medicines, diagnostics, vaccines, devices, or other life-saving health products and services.” CHAI, “Homepage”

• 3

  See Ni-kshay, TB Notification report, data from 01/01/2024 to 17/09/2024:

  • Bihar: 54,738 + 84,423 = 139,161
  • UP: 291,458 + 173,838 = 465,296
  • Bihar + UP = 139,161 + 465,296 = 604,457
  • Total: 1,145,408 + 657,464 = 1,802,872
  • (Bihar + UP)/Total = 604,457/1,802,872=0.335=33.5%

• 4
  • “Once identified as a household contact, children still need to be brought in to a health facility for TPT initiation. This leads to several “missed” children at this step or delays in linkage to TPT treatment.” CHAI, "TB HCM program proposal," May 1, 2024, p. 1.
  • “Through our field visits to non JEET districts as well as interactions with district staff, our overall impression is that contact tracing and treatment initiations are facility-based. Moreover, some of the facility-based initiation practices observed were not ideal. For instance, we observed index TB patients were being handed over TPT drugs for HHC after cursorily asking for details of household members. Our proposed approach of working with existing CHWs to undertake HH contact tracing is aimed at ensuring long term sustainability by providing CHWs with necessary tools to plan, monitor and independently undertake contact tracing." Correspondence from CHAI to GiveWell, January 26, 2024 (unpublished).

• 5

  “3HP is a short-course TPT regimen that combines two antibiotics active against TB, INH and RPT. 3HP is taken once a week for 12 weeks (12 doses in 3 months).” Unitaid, "Short-course Treatment Regimens to Prevent TB: 3HP and 3RH."

• 6
  • “Until recently, 6 months isoniazid was the more widely used regimen for TPT (isoniazid preventive treatment or IPT)" Unitaid, "Short-course Treatment Regimens to Prevent TB: 3HP and 3RH"
  • "Daily isoniazid monotherapy: The efficacy of daily isoniazid monotherapy for six months (6H) or more in different populations and settings has been shown in a number of systematic reviews" WHO, “WHO consolidated guidelines on tuberculosis,” 2020, p. 18.

• 7

  “Because a fruit-flavoured, dispersible medicine can make the difference between a child taking their medicine or not, they can be transformative in the lives of children and caregivers affected by disease.” Unitaid, "New Child-Friendly Formulation of Rifapentine for Short Course Tuberculosis Prevention Treatment Now Available as Unitaid and IMPAACT4TB Launch an Early Market Access Vehicle."

• 8
  • See the “Increase in TPT coverage rates” subsection of GiveWell, “Intervention Report on Household contact management for tuberculosis,” 2024.
  • See cells B14:C16 in GiveWell's CEA of a community-based TB HCM program in India. Total number of under-5 deaths directly averted = 974, and deaths averted via prevention of disease progression = 718. 718/974=75%.

• 9
  • See the “Increase in treatment rates for TB disease” subsection of GiveWell, “Intervention Report on Household contact management for tuberculosis,” 2024.
  • See cells B14:C16 in GiveWell's CEA of a community-based TB HCM program in India. Total number of under-5 deaths directly averted = 974, and deaths averted via prevention of disease progression = 718. 718/974=~75%. The remaining ~25% of total number of under-5 deaths directly averted are attributed to deaths averted via prevention of disease progression.

• 10

  See the “Additional benefits of preventing TB in early childhood” subsection of GiveWell, “Intervention Report on Household contact management for tuberculosis,” 2024.

• 11

  See the “Second-order considerations in our cost-effectiveness analysis” subsection of GiveWell, “Intervention Report on Household contact management for tuberculosis,” 2024.

• 12
  • “At the two-year point, CHAI expects to successfully transition treatment district implementation to Bihar and UP officials. … CHAI will maintain operational continuity by providing light-touch TA support which may include: 1) making sporadic supervisory visits; 2) organizing state level review meetings; and 3) providing ongoing supply chain support to ensure drug availability. CHAI could facilitate additional training, if required.” Correspondence from CHAI to GiveWell, May 13, 2024 (unpublished)
  • We think that CHAI’s decision to leverage existing CHWs for household visits is preferable to hiring a parallel set of health workers. This is because it increases the chances of the program being sustained by the government after CHAI leaves and because this seems like a more scalable and cost-effective implementation model than one which requires additional workers to be hired. This view was also expressed to us by multiple external stakeholders, including government officials. See here.

• 13

  See the budget for grant activities below.

• 14

  "Leveraging the combined strengths of the NTP and general health system, the program will engage Accredited Social Health Activist (ASHA) and TB Health Volunteers (TB-HVs) for patients in the public sector to optimize their bandwidth and prioritize TPT. Moreover, to address the gaps in the private sector CHAI will contract PPSAs to expand TPT coverage amongst patients seeking care in the private sector. These three cadres of CHWs will go to households for contact tracing, management, and care, ultimately bridging the gap between healthcare facilities and communities." CHAI, “TB HCM program proposal,” May 1, 2024, p. 2.

• 15
  • "In addition to this household-based primary intervention, all household TB contacts who are symptomatic will be referred to the base health facility for additional diagnostic screening." CHAI, “TB HCM program proposal,” May 1, 2024, p. 2.
  • "All household contacts that screen positive would get a fast-track referral to the health facility. This is different from an individual with symptoms visiting at a facility to seek care, where there may be significant delays in diagnosis because the doctor may not suspect TB upfront." Correspondence from CHAI to GiveWell, November 9, 2023 (unpublished)

• 16
  • “Leveraging the combined strengths of the NTP and general health system, the program will engage Accredited Social Health Activist (ASHA) and TB Health Volunteers (TB-HVs) for patients in the public sector to optimize their bandwidth and prioritize TPT. Moreover, to address the gaps in the private sector CHAI will contract PPSAs to expand TPT coverage amongst patients seeking care in the private sector. These three cadres of CHWs will go to households for contact tracing, management, and care, ultimately bridging the gap between healthcare facilities and communities.” CHAI, "TB HCM program proposal," May 1, 2024, p. 2.
  • CHAI explained that it will conduct extensive training and ensure adequate supervision of CHWs. Correspondence from CHAI to GiveWell, November 27, 2023 (unpublished)

• 17

  CHAI’s proposed approach of working with existing CHWs includes providing CHWs with necessary tools to plan, monitor and independently undertake contact tracing. Correspondence from CHAI to GiveWell, January 26, 2024 (unpublished)

• 18

  This will be done by a central call center contracted by CHAI: “We plan to contract the same call centre engaged by NTP for this project which we believe will also enable long term sustainability of the intervention. The call centre staff will be based centrally. However, their work will support district level teams.” Correspondence from CHAI to GiveWell, March 28, 2024 (unpublished)

• 19

  “We do want to note that the budget does include dedicated District Program Officers (DPOs) who will jointly monitor, supervise and train CHWs with health system staff.” Correspondence from CHAI to GiveWell, October 2023 (unpublished).

• 20

  CHAI’s district level staff will “support Block Program Managers and District TB Officers to conduct periodic review meetings by compiling data, setting agenda, defining talking points etc., engage with District Chief Medical Officer and other officials to prioritize TB HCM, review performance data and suggest opportunities for improvement to district / block officials, engage with key district officials in control districts to support evaluation activities and obtain necessary permissions.” Correspondence from CHAI to GiveWell, March 28, 2024 (unpublished).

• 21

  See “The roles and responsibilities of CHAI, Co-Investigators and IDinsight for different M&E components” table, under “CHAI role.” CHAI, "TB HCM program proposal," May 1, 2024, p. 4-5.

• 22

  CHAI’s state level staff will “support the state in streamlining payment of incentives already budgeted by the government to CHWs, support quantification and supply chain strengthening for TPT drugs, support state program officers to conduct review meetings by compiling data, defining agenda, and prioritizing key discussion points; engage with state level leadership to prioritize TB HCM in the state, monitor activities of the CHAI district level staff and coach them to improve performance.” Correspondence from CHAI to GiveWell, March 28, 2024 (unpublished).

• 23
  • “At the two-year point, CHAI expects to successfully transition treatment district implementation to Bihar and UP officials. In CHAI’s experience, key stakeholders in Bihar and UP can make a robust case for adoption based on interim performance results and feedback from trusted, local primary investigators and implementation staff. This local evaluation would also serve as the foundation for the decision to scale-up beyond intervention and control districts.

  CHAI will maintain operational continuity [in treatment districts] by providing light-touch TA support which may include: 1) making sporadic supervisory visits; 2) organizing state level review meetings; and 3) providing ongoing supply chain support to ensure drug availability.” Correspondence from CHAI to GiveWell, May 23, 2024 (unpublished).

  • “During years three and four, you will have a very clear vision as to how well the intervention transition has gone. And you can then adjust. We're making assumptions about how effective the transition will be. … [B]ecause we're doing a sequenced rollout, we can use the first cohort [treatment districts] to see what that transition looked like, how difficult it was. And in some ways, the first cohort should actually be the easiest because they're the most compelled, right? They're the one who did the pilot. That's where the champion, sort of the state championship initiative and energy comes from. And if they need a lot of help, then you can say, well, wow, the place we did the intervention needed a lot of help. Then the places that are newer, who are potentially lower TB risk, lower capacity, they may need more. That's something that you have two years to make an assessment of before that final phase. So we're giving you our best guesstimate. But there's optionality of ‘you'll get real data real time during years three and four’ as you correctly describe.” CHAI staff, conversation with GiveWell, May 16, 2024 (unpublished).

• 24

  Note that the outcomes of this market-shaping are unlikely to affect TPT delivery funded by this grant, as these outcomes are likely to manifest after the program starts being implemented. CHAI is planning to support procurement of sufficient TPT for treatment districts at current prices. See the “Market shaping activities could generate additional benefits” subsection for more detail.

• 25

  “Once the caregiver has accepted, the adherence challenges are minimal as compared to active TB therapy: Ease - Dispersible and can be put into juice, etc. Short time frame - 1 month - 3 months depending on the selected regime.” Correspondence from CHAI to GiveWell, July 17, 2024 (unpublished).

• 26

  “In December 2023, the government completed the procurement of >1 million courses of Rifapentine. With this recent development, we expect similar levels of 3HP availability in both intervention and control districts. Pediatric formulation of 3HP is expected to be approved by WHO by September 2024. However, we do not expect that [the] government of India will be able to procure pediatric 3HP formulation by the expected start of the project (early 2025). Thus, we would now like to suggest that all pediatric 3HP courses required for this proposed pilot are procured through CHAI. In the control districts we expect under 5 HHC to receive the adult 3HP formulation, which will be crushed and administered. Although, crushing is not recommended due to lack of studies on bioequivalence, it currently practiced in India even for the 6 month TPT regimen.” Correspondence from CHAI to GiveWell, December 2023. (unpublished).

• 27

  “PPSAs will be incentivized $4 for every under 5 HHC initiated on TPT. We expect the PPSA to utilize this amount to augment staff capacity to provide the additional service of HCM for private sector patients.” Correspondence from CHAI to GiveWell, November 27, 2023. (unpublished).

• 28

  “While there will be a slight increase in [TB-HV] workload due to TPT initiation at the community level, this may be partially compensated by TB-HVs no longer being responsible for TPT initiation for private sector patients, who account for 50% of patient load in Bihar and 28% in Uttar Pradesh.” Correspondence from CHAI to GiveWell, November 27, 2023. (unpublished)

• 29

  “Since TBHVs are full-time staff and ASHAs are already incentivized by the government to support HH contacts to complete TPT, we haven’t budgeted any additional resources under the project.” Correspondence from CHAI to GiveWell, October 18, 2023. (unpublished)

• 30

  More specifically, in the districts assigned to the treatment group, CHAI will implement district-level activities to support health workers to visit the households of TB patients with children under 5 to screen their contacts and provide TPT and referrals. In the districts assigned to the control group, CHAI will start implementing these activities two years later, after the evaluation period.

  • "To minimize conflict of interest and prove that the program had the impact as intended, IDinsight India has been hired to conduct an impact evaluation and accompanying process evaluation on the program over the first 2 years of the 4-year grant. For the impact evaluation, the primary outcome of interest is the impact of the CHW-based model on TPT initiation rates amongst eligible children-U5. The impact evaluation will thus be a cluster randomized controlled trial comparing the impact of the TB HCM approach on TPT coverage amongst U5, compared to standard TPT delivery which is currently a facility-based model.” CHAI, “TB HCM program proposal,” May 1, 2024, p. 3-4.
  • See the “Timeline” chart on p. 5. CHAI, “TB HCM program proposal,” May 1, 2024.
  • We propose a cluster randomized evaluation, randomized at the district level, to rigorously measure the impact of CHAI’s program. IDInsight, “Concept note: Impact Evaluation of CHAI’s Household Contact Management for Childhood Tuberculosis Program in India,” May 2024, p. 2.

• 31

  GiveWell, Learning from CHAI's TB HCM pilot in India (unpublished)

• 32
  • The cluster RCT is intended to run for 2 years at which point all remaining districts will launch the CHW-based program.” CHAI, "TB HCM program proposal," May 1, 2024, p. 4.
  • “A pilot M&E will be conducted over a period of 2 years [from Jan 2025 to Dec 2026, tbc]. Subsequently, program implementation will proceed for subsequent 2 years [Jan 2027 to Dec 2028, tbc] from across all pilot sites.” CHAI, "TB HCM program proposal," May 1, 2024, p. 5.

• 33

  “The planned program includes 26 high-TB notification districts Bihar and 44 high-TB notification districts Uttar Pradesh, for a total of 70 districts. Collectively, these states represent 26% of the population and 32% of TB prevalence in India.“ CHAI, "TB HCM program proposal," May 1, 2024, p. 2.

• 34

  “At the two-year point, CHAI expects to successfully transition treatment district implementation to Bihar and UP officials. … CHAI will maintain operational continuity by providing light-touch TA support which may include: 1) making sporadic supervisory visits; 2) organizing state level review meetings; and 3) providing ongoing supply chain support to ensure drug availability. CHAI could facilitate additional training, if required.” Correspondence from CHAI to GiveWell, May 13, 2024 (unpublished)]

• 35

  See CHAI’s full budget, including the breakdown between years, here.

• 36

  See the “The risk of tuberculosis disease and death for under-5 household contacts” subsection of GiveWell, “Intervention Report on Household contact management for tuberculosis,” 2024.

• 37

  See the “The increase in treatment coverage generated by a community-based approach” subsection of GiveWell, “Intervention Report on Household contact management for tuberculosis,” 2024.

• 38

  See the “The effectiveness of TPT for household contacts under 5” subsection of GiveWell, “Intervention Report on Household contact management for tuberculosis,” 2024.

• 39

  See the “The cost of buying TPT drugs and implementing a community-based program” subsection of GiveWell, “Intervention Report on Household contact management for tuberculosis,” 2024.

• 40

  Note that at the time of making this grant, 3HP dosing guidelines were only available for children age 2 years and above, but the WHO has been reviewing dosing for younger children and CHAI expects these guidelines to be put into practice by countries in the second half of 2024:

  • “WHO currently recommends the use of 3HP in children 2 years of age and above. Dosing for younger children will be reviewed by the WHO Technical Advisory Group on dosing in Q1 2024.” Unitaid, “New Child-Friendly Formulation of Rifapentine for Short Course Tuberculosis Prevention Treatment Now Available as Unitaid and IMPAACT4TB Launch an Early Market Access Vehicle,” December 22, 2023.
  • “The Tuberculosis Trials Consortium Study 35 (TBTC S35)[1] that aims to provide 3HP regimen dosing information for children under 2 is anticipated to conclude this year. The study results are expected to be available early next year. WHO’s guidance based on the findings is anticipated by July 2024. … Following WHO’s guidance, we plan to seek IRB approval to administer 3HP to children under 2, including 0-4 months age group- anticipating national guideline revision may take longer duration after WHO’s guidance is available.” Correspondence from CHAI to GiveWell, November 27, 2023 (unpublished)

• 41
  • “3HP is a short-course TPT regimen that combines two antibiotics active against TB, INH and RPT. 3HP is taken once a week for 12 weeks (12 doses in 3 months).” Unitaid, "Short-course Treatment Regimens to Prevent TB: 3HP and 3RH."
  • “Evidence shows that short-course TB preventive treatment regimens are cost-effective; people taking shorter drug regimens are up to three times more likely to complete their course of TPT than those on longer regimens—leading to better outcomes and more lives saved.” Unitaid, “New Child-Friendly Formulation of Rifapentine for Short Course Tuberculosis Prevention Treatment Now Available as Unitaid and IMPAACT4TB Launch an Early Market Access Vehicle,” December 22, 2023.
  • The paediatric formulation … is water-soluble and raspberry-flavoured, making it easy to administer and more acceptable to children.” Unitaid, “New Child-Friendly Formulation of Rifapentine for Short Course Tuberculosis Prevention Treatment Now Available as Unitaid and IMPAACT4TB Launch an Early Market Access Vehicle,” December 22, 2023.

• 42

  See the “Additional benefits of preventing TB in early childhood” subsection of GiveWell, “Intervention Report on Household contact management for tuberculosis,” 2024.

• 43

  See the “Increase in treatment rates for TB disease” subsection of GiveWell, “Intervention Report on Household contact management for tuberculosis,” 2024.

• 44

  See adjustments made in cells A205:E237 in GiveWell's CEA of a TB HCM program in India.

• 45

  “CHAI sees further market shaping potential to further reduce the price of pediatric rifapentine … minimizing supply chain disruptions” GiveWell, Market-shaping opportunity in CHAI TB grant, 2024. (unpublished).

• 46

  “CHAI’s proposal to create a co-packaged product will enable countries to buy both products together, minimizing supply chain disruptions (we’ve noted numerous instances where one of the therapies in the regimen is unavailable).” GiveWell, Market-shaping opportunity in CHAI TB grant, 2024. (unpublished).

• 47

  “The co-packaged product would also effectively increase demand for the peds TPT products and provide suppliers with the opportunity to increase sales volumes allowing for manufacturing efficiencies and profitability at lower margins.” GiveWell, Market-shaping opportunity in CHAI TB grant, 2024. (unpublished).

• 48

  Over 3 million children under the age of 15 are in need of TPT each year, while roughly 1 million per year are currently receiving pediatric TPT. Correspondence from CHAI to GiveWell, December 7, 2023 (unpublished)
  Assuming CHAI’s market-shaping efforts affect half of the current market, and that the regimen currently costs ~$7 on average for eligible children under the age of 15, these figures imply that i) a 20% price reduction would save governments and other funders around $700,000 per year to maintain current volumes (20%*$7=$1.4 in savings per child, impacting 50%*1 million=500,000 children for a total savings of $1.4*500,000 children = $700,000) and ii) could increase demand such that hundreds of thousands of additional child contacts receive TPT for at least a few years. This scale of benefit seems worth the $213,000 for CHAI to conduct the market-shaping.

• 49

  “To minimize conflict of interest and prove that the program had the impact as intended, IDinsight India has been hired to conduct an impact evaluation and accompanying process evaluation on the program over the first 2 years of the 4-year grant.” CHAI, "TB HCM program proposal," May 1, 2024, p. 3.

• 50

  See the “Research questions” section of IDInsight, “Concept note: Impact Evaluation of CHAI’s Household Contact Management for Childhood Tuberculosis Program in India,” May 2024 (p. 2-3).

• 51

  “To estimate the impact of the HCM program, we will conduct a cluster randomized controlled trial RCT, randomizing at the district level. In Bihar, we will randomly assign 13 districts to receive the program and 13 districts to serve as controls. In UP, we will randomly assign 22 districts to receive the program and the remaining 22 districts to serve as controls.
  We will randomize districts to receive CHAI’s HCM program using stratified random assignment
  with strata defined by JEET implementation status and any other variables which the CHAI team
  believes are most predictive of the primary outcome variables. This could include whether the
  district is rural or urban, estimates of the baseline levels of the primary outcome variable from
  Ni-Kshay / JEET admin data, or an estimate of CHW effectiveness from NFHS (or other) data.” IDInsight, “Concept note: Impact Evaluation of CHAI’s Household Contact Management for Childhood Tuberculosis Program in India,” May 2024 (p. 3)

• 52

  “This evaluation aims to answer the following primary research questions … What is the effect of CHAI’s HCM for Childhood TB Program on TPT initiation for eligible children under 5?” IDInsight, “Concept note: Impact Evaluation of CHAI’s Household Contact Management for Childhood Tuberculosis Program in India,” May 2024 (p. 3)

• 53

  “To collect data on primary outcomes, we will [following approval from the National Tuberculosis Elimination Program] sample active TB patients from the Ni-Kshay registry, locate patients’ households using data from Ni-Kshay, and interview household members. … Our survey respondents will be the index patient, head of household or other adult member of the household (other than the index patient), and the primary caregiver of the U5 child(ren) in the household (if different from either of the above). … During the household survey of index patients, we will supplement info on recall of TPT initiation of U5s with observation of strips of 3HP tablets to verify U5s’ intake of 3HP doses. For now, we assume that caregivers are distributed tablets for a month at a time based on the last call. … The
  survey will also include additional probes around challenges with consumption, methods of disposal, etc. to indirectly verify recall of doses administered.” IDInsight, “Concept note: Impact Evaluation of CHAI’s Household Contact Management for Childhood Tuberculosis Program in India,” May 2024 (p. 4-5)

• 54

  See the “Research questions” section of IDInsight, “Concept note: Impact Evaluation of CHAI’s Household Contact Management for Childhood Tuberculosis Program in India,” May 2024 (p. 2-3).

• 55

  “We will collect both quantitative and qualitative data to answer the evaluation questions at one of
  three timepoints:

  • Baseline: Data will be collected prior to the roll-out of the program. Based on CHAI and GiveWell’s feedback, this will be done in the full sample of evaluation districts.
  • Process evaluation: Rather than conducting a quantitative midline survey, we will conduct a qualitative process evaluation ~9 months after the program is launched in the evaluation areas.
  • Endline: Data will be collected ~19 months after the program is launched in the evaluation areas to measure our outcomes.” IDInsight, “Concept note: Impact Evaluation of CHAI’s Household Contact Management for Childhood Tuberculosis Program in India,” May 2024 (p. 4).

• 56

  IDinsight plans to sample 5,636 TB patients and expects 40% to be living with a child under 5 and 70% to respond to the survey. This implies an analytical sample of 1,578 (=5,636*0.4*0.7) child contacts aged under 5 if each of these households has exactly 1 under-5 child contact:
  “At baseline, we will sample 5,636 active TB patients indexed approximately 30-60 days prior to the survey. At the endline, we will sample 5,636 active TB patients indexed approximately 30-60 days
  prior to the survey and an additional 5,636 active TB patients indexed approximately 120-150 days prior to the survey for a total of 11,272 active TB patients. … These figures represent the total number of active patients to be sampled and not our final analytical sample (which will be significantly smaller as many patients won’t live with children under 5 or respond to the survey). If the share of patients who do not respond or who do not live with children under 5 is higher than anticipated we would increase our sample size to account for this. The table below provides more information on the inputs used in these sample size calculations.” IDInsight, “Concept note: Impact Evaluation of CHAI’s Household Contact Management for Childhood Tuberculosis Program in India,” May 2024 (p. 8, 9, 22).

• 57

  “Note that while the MDE for Bihar and UP is 14 and 15 percentage points respectively, the MDE for
  both states combined is approximately 9.74 percentage points.” IDInsight, “Concept note: Impact Evaluation of CHAI’s Household Contact Management for Childhood Tuberculosis Program in India,” May 2024 (p. 24).

• 58

  Bonnet et al. (2023) report that a community-based child TB contact management program in Uganda and Cameroon had effects of 35pp (=81.9%-47.3%) on the percent of eligible contacts screened and 18pp (=79.7%-61.7%) on TPT completion:
  “Among all child contacts declared, 1548 (81·9%) of 1889 in the intervention group and 475 (47·3%) of 1005 in the control group were screened for tuberculosis. 1400 (48·4%) child contacts were considered to be in the TPT target group: 941 (49·8%) of 1889 in the intervention group and 459 (45·7%) of 1005 in the control group. In the TPT target group, TPT was commenced and completed in 752 (79·9%) of 941 child contacts in the intervention group and 283 (61·7%) of 459 in the control group (odds ratio 3·06 [95% CI 1·24–7·53]).”

• 59

  The cost-effectiveness of the program would be 4x, rather than 8x, if the effect on population-level TPT coverage was 10 percentage points rather than 22 percentage points. This would occur if the proportion of households visited by CHWs dropped below 30 percent, or if the proportion of household contacts who initiate TPT when visited by a CHW dropped below 60 percent.

• 60
  • “In addition to the main randomized controlled trial, we will conduct a qualitative process evaluation to take place midway between baseline and endline. The process evaluation will provide insights into the experiences and perceptions of the program from caregivers of U5s, CHWs, and other program implementation staff (including government stakeholders and CHAI staff) at the district and state level.” IDInsight, “Concept note: Impact Evaluation of CHAI’s Household Contact Management for Childhood Tuberculosis Program in India,” May 2024 (p. 3).
  • “Semi-structured key informant interviews: We will conduct in-depth interviews with various program stakeholders including household members, CHWs, CHW supervisors, district officials, state officials in both treatment and control areas, as well as CHAI district staff. Data from these interviews will be used to understand the experience and perception about the intervention from caregivers of U5s on TPT regimen and those involved in its implementation.” IDInsight, “Concept note: Impact Evaluation of CHAI’s Household Contact Management for Childhood Tuberculosis Program in India,” May 2024 (p. 6-7).

• 61
  • “We will survey a random sample of TB-HVs and PPSAs to collect data on CHW workload and other variables. By comparing reported HCM and non-HCM workload for TB-HVs and PPSAs in the treatment and control groups between baseline and endline, we may obtain experimental estimates of the impact of the program on non-HCM workload and other variables.” IDInsight, “Concept note: Impact Evaluation of CHAI’s Household Contact Management for Childhood Tuberculosis Program in India,” May 2024 (p. 5).
  • “To estimate the impact of the program on ASHA workers’ workload, we will conduct a time-use survey of ASHA workers performing household contact management. In the time-use survey, surveyors will shadow ASHA workers as they visit households of active TB patients for the purpose of household contact management. This data will allow us to estimate the time required to perform tasks related to the HCM program and thus crude estimates of the impact of the HCM program on ASHA workloads. We will also perform time use surveys of TB-HVs and PPSAs as they perform HCM tasks. This data will allow us to double check and validate the experimental estimates of the impact of the HCM program on these cadres’ workloads. This may be useful if TB-HVs and PPSAs have an incentive to over or under report the impact of HCM tasks on their workload.” “Concept note: Impact Evaluation of CHAI’s Household Contact Management for Childhood Tuberculosis Program in India,” May 2024 (p. 6).

• 62

  Additional room for more funding for Bihar and Uttar Pradesh if the program is not fully transitioned to government ownership is estimated here.

• 63

  "Market shaping — in the context of global health — comprises interventions to create well-functioning markets through improving specific market outcomes (e.g., availability of products) with the end goal of improving public health. Market shaping interventions tend to be catalytic, timebound, and have a strong focus on influencing buyer and supplier interactions." Rethink Priorities, An overview of market shaping in global health: Landscape, new developments, and gaps, September 1, 2023.

• 64

  “Results: 135 global agreements to lower prices 50 – 90 % and improve access to drugs, diagnostics, and other health products in low-income countries.” CHAI, “Market Shaping.”

• 65

  CHAI, "Quality and Market shaping learning agenda," 2024 (unpublished).

• 66
  • “A total of 1.3 million people died from TB in 2022 (including 167 000 people with HIV). Worldwide, TB is the second leading infectious killer after COVID-19 (above HIV and AIDS).” World Health Organization, "Tuberculosis (Fact sheet)," 2023.
  • “The current Global Plan, for 2023–2030, estimates much higher funding needs, of US$ 15–32 billion per year in LMICs; this includes funding for implementation of a new TB vaccine after 2027. The political declaration adopted at the second UN high-level meeting on TB, held in September 2023, includes funding targets to mobilize US$ 22 billion per year by 2027 for TB diagnostic, treatment and prevention services, and US$ 35 billion per year by 2030; a target of US$ 5 billion per year by 2027 was set for investment in TB research.” World Health Organization, “Financing for TB prevention, diagnostic and treatment services,” 2023.
  • “Funding available for TB prevention, diagnostic and treatment services in LMICs falls far short of the globally estimated need and the UN global targets, and has fallen since 2019. In 2022, the total was only US$ 5.8 billion (Fig. 4.3). This is less than half of the amounts estimated to be required in 2022 in the Global Plan (2018–2022) and the global target set at the UN high-level meeting on TB in 2018.” World Health Organization, “Financing for TB prevention, diagnostic and treatment services,” 2023.

• 67

  We consulted these stakeholders when drafting this intervention report but did not ask them for feedback on the entire report.

• 68
  • “JEET was being implemented across 11 states (including Bihar and UP), with one third of the implementation districts in Bihar and UP.” Correspondence from CHAI to GiveWell, March 28, 2024 (unpublished).
  • JEET 1.0 ran from 2018-2021, and JEET 2.0 ran from 2021-2024. See CHAI, “TB HCM program proposal,” May 1, 2024, p. 6-7.

• 69

  Our understanding is that Project JEET was focused on private sector patients, but because it was funded through a Global Fund grant we believe that CHAI had to maintain strong relationships with public sector officials in order for the project to have been implemented successfully.

• 70

  “CHAI India, with support from The Global Fund, (2021-2024) has implemented HCM interventions by
  mobilizing the family members of index patients to the health facility. Screening, treatment initiation and
  referral were triggered from the health facility. First-hand, field experience has helped CHAI gain a
  nuanced understanding of the challenges of health facility based HCM. Since HCM has now been taken up by the health system, this provides an opportunity to decentralize the service delivery to community-
  based initiation through CHWs. … This proposal introduces a dynamic community-based intervention that capitalizes on the strengths of both NTEP and general health system teams, supported by CHWs. By aligning with national guidelines and fostering collaborative efforts, we anticipate significant strides in enhancing TB prevention and care, ultimately contributing to India's overarching goal of eliminating TB by 2025” CHAI, “Proposal to GiveWell,” September 2023, (unpublished)

• 71

  This experience includes:

  • Its collaboration with Indian research organizations as part of IMPAACT4TB to generate local evidence on short-course TPT for the National Technical Working Group
  • Its recent work with Delhi State TB Program to generate evidence and introduce guidelines regarding the use of computer-aided diagnosis of chest x-rays in detecting TB
  • Its ongoing immunization program in Bihar, Uttar Pradesh and Madhya Pradesh, which involves strengthening data systems and governance/review mechanisms and prototyping community-based interventions to reduce the number of ‘zero dose’ children.

  See the “Previous CHAI experience in India” section of CHAI, “TB HCM program proposal,” May 1, 2024, p. 5-8.

• 72

  See our grant page for the CHAI Incubator here for more details.

• 73

  “We did a parallel, cluster-randomised, controlled trial across 20 clusters (consisting of 25 district hospitals and primary health centres) in Cameroon and Uganda, which were randomised (1:1) to receive a community-based approach (intervention group) or standard-of-care facility-based approach to contact screening and management (control group).” Bonnet et al. 2023.

• 74

  “The community-based approach consisted of symptom-based tuberculosis screening of all household contacts by community health workers at the household, with referral of symptomatic contacts to local facilities for investigations. Initiation of TPT (3-month course of rifampicin–isoniazid) was done by a nurse in the household, and home visits for TPT follow-up were done by community health workers.” Bonnet et al. 2023.

• 75

  “In both groups, the TPT regimen consisted of 3 months of daily rifampicin and isoniazid and using the dispersible, fixed-dose combination formulation (75 mg rifampicin and 50 mg isoniazid for contacts weighing <25 kg), procured by the CaP-TB project, as the national programme did not yet recommend this regimen when the study started.” Bonnet et al. 2023.

• 76

  The main way in which CFR might be overestimated in these studies is if they captured only the most severe (and hence more easily diagnosed) forms of TB. But there are also ways in which these studies might underestimate CFR. See “The risk of tuberculosis disease and death for under-5 household contacts” subsection of GiveWell, “Intervention Report on Household contact management for tuberculosis” for more details.

• 77

  We have already tried to account for the effect of BCG coverage on the under-5 CFR with a 20% downwards adjustment, but it is possible we are still underestimating its effect. See “The risk of tuberculosis disease and death for under-5 household contacts” subsection of GiveWell, “Intervention Report on Household contact management for tuberculosis” for more details.

• 78

  See the “Additional benefits of preventing TB in early childhood” subsection of GiveWell, “Intervention Report on Household contact management for tuberculosis” for more details.

• 79

  See the “Additional benefits of preventing TB in early childhood” subsection of GiveWell, “Intervention Report on Household contact management for tuberculosis” for more details.

• 80

  Ayieko et. al. 2014, a meta-analysis on the efficacy of isoniazid prophylactic therapy in prevention of tuberculosis in children, identified only two studies focused primarily on the effectiveness of TPT for infants, and found significant differences in the impact found in those studies compared to studies that included primarily older children.

  • “Upon pooling the studies, INH was found to provide substantial protection against TB among children aged ≤15 years (RR = 0.65, 95% CI 0.47, 0.89 p = 0.004), however, there was marked heterogeneity in the efficacy of INH (Chi 2 18.56, df = 7, I 2 = 62.3% p = 0.01). We conducted a sensitivity analysis to explore the observed heterogeneity. Most of the heterogeneity was noted to be due to the two studies by Madhi et al.(also referred to as the infant studies in this article) [19], primarily due to two reasons. First, the results of the two studies by Madhi et al. were notably different from the results of the other studies, in that these two studies yielded point estimates much closer to the null compared to the results of the other studies, with confidence intervals that included the null, while the rest of the studies had point estimates and confidence intervals that showed that INH conferred a protective effect. Second, both the random effects and the fixed effects models gave these two studies more weight because of their relatively large size. One plausible explanation for the null results in the Madhi studies may be due to overdiagnosis of TB given that only few of the cases were confirmed microbiologically and many cases only met minimal clinical criteria [19]. After excluding the two studies by Madhi et al., the heterogeneity disappeared (I2 = 0.0%, p = 0.799) and our analysis yielded an even greater effect estimate (RR = 0.41 (95% CI 0.31, 0.55) p <0.001). One possible explanation for the discrepancy between findings from Madhi et al. and from the remaining studies has to do with the age of the participants at initiation of INH. Whereas subjects in the Madhi et al. trials were infants (median age 96 days, range 91–120 days) at the time of INH initiation, the remaining studies involved mostly older children, suggesting that age may modify the effect of INH prophylaxis on the development of TB” Ayieko et. al. 2014, p. 6.

• 81

  3HP safety and dosing information for children under 2 are forthcoming this year, according to CHAI; this review is also mentioned by Unitaid here.

• 82

  See the “The effectiveness of TPT for household contacts under 5” subsection of GiveWell, “Intervention Report on Household contact management for tuberculosis” for more details.

• 83

  This phenomenon should be captured in studies of the effectiveness of TPT if they are conducted in high-burden settings (where risk of reinfection is high) and had long follow-up periods. However, it’s unclear whether many of the studies of TPT we base our effectiveness estimate meet these two conditions. This may be more of a concern with shorter-course regimens than longer-course regimens, since the direct protective effect of the drugs may only last as long as they are being taken. Based on some preliminary unpublished research regarding the extent to which this is a concern across different diseases and interventions, we think it isn’t a particularly large concern for pediatric TB programs. This is because much of the mortality risk from TB in childhood seems to be concentrated in the early years. If a program can prevent deaths in these years it is likely to prevent the child’s death at least until they become an adult. See the “The effectiveness of TPT for household contacts under 5” subsection of GiveWell, “Intervention Report on Household contact management for tuberculosis” for more details.

• 84

  See the “The effectiveness of TPT for household contacts under 5” subsection of GiveWell, “Intervention Report on Household contact management for tuberculosis” for more details.

• 85

  CHAI reported to us that they have so far had formal conversations with:

  • The State TB Officer in Bihar (we understand this to be the most senior state-level position in the National Tuberculosis Program in India)
  • The WHO consultant to the TB Program in Uttar Pradesh
  • Two Director-level National TB Program officials Correspondence from CHAI to GiveWell, March 28, 2024 (unpublished)

• 86

  CHAI staff, conversation with GiveWell, April 11, 2024 (unpublished).

• 87

  CHAI's past work in Bihar and Uttar Pradesh includes immunization programs, and past TB programs such as JEET 2.0. For more detail, see the “Previous CHAI experience in India” section of CHAI, “TB HCM program proposal,” May 1, 2024, p. 5-8 and Joint Effort for Elimination of Tuberculosis (JEET), “About” (the map shows WJCF (CHAI's India affiliate) implementing JEET in Bihar and UP (by color).

• 88

  CHAI reports having a long history of working with and for the Indian government. CHAI staff, conversation with GiveWell, April 11, 2024 (unpublished).

• 89
  • It is aligned with India’s current guidelines on Programmatic Management of TB Preventive Treatment: “India released the Guidelines on Programmatic Management of TB Preventive Treatment (PMTPT) in July 2021. These guidelines introduced a comprehensive "cascade of care" approach, to expand TPT to HHCs of all ages of pulmonary TB patients and other high-risk groups (HRGs). The strategy involves effective screening, TPT evaluation, treatment initiation, and systematic follow-up. The proposed program by CHAI is aligned with guidelines laid down in the National Guidelines.” Correspondence from CHAI to GiveWell, March 28, 2024 (unpublished)
  • State TB officials in Bihar and UP are aware Project JEET is coming to an end this year and would like to ensure that gains in TPT coverage from those efforts are not lost: “CHAI has been implementing Project JEET from Aug 2021 to March 2024. As JEET is being wound down, UP and Bihar are keen that the gains made towards increasing TPT coverage under JEET is not lost completely. State officials recognize the need to work with existing staff (including CHWs) to improve TPT coverage. Moreover, CHAI has extensive experience implementing Technical Assistance projects in UP and Bihar. From our experience we have observed the states are keen to work with partners for technical assistance to improve health program performance.” Correspondence from CHAI to GiveWell, March 28, 2024 (unpublished)

• 90
  • Reuben Swamickan, Deputy Director, USAID India in conversation with GiveWell, January 2024. (unpublished)
  • Dr. BK Mishra, State TB Officer, Bihar in conversation with GiveWell, January 2024. (unpublished)
  • Dr. Vinod Kumar, WHO consultant to Uttar Pradesh’s State TB program in conversation with GiveWell, January 2024 (unpublished)

• 91
  • CHAI confirmed that they will not be able to directly manage or exert authority over CHWs: “Though we may liaise and build the capacity of CHWs, we don’t have direct management over CHWs.” Correspondence from CHAI to GiveWell, March 28, 2024 (unpublished)
  • CHAI also flagged CHW behavior change challenges as key risks in their program theory of change: “Due to limited number of TB households with U5 contact per ASHA, they may not prioritize TB HCM despite provision of incentives … TB HVs [may] prefer to conduct facility based HCM due to convenience.” Correspondence from CHAI to GiveWell, March 28, 2024 (unpublished)

• 92

  Our best guess of 60% is an average of our best guesses of the proportion of households that will be reached in year 1 (55%) and year 2 (65%) of the pilot.

• 93

  CHAI shared its approach to ensuring CHWs have the time, skills, tools and financial incentives in Correspondence from CHAI to GiveWell, March 28, 2024 (unpublished). Some examples are below:

  • In general:
    • “Undertake trainings, provide tools, establish processes to streamline tasks, and intensify monitoring
    • “CHAI will further support TBHVs to optimize their tasks through M&E tools, and call centre support.”
    • “CHAI district staff helps prioritize the supportive supervision … to poor performing areas.”
  • TBHVs:
    • “District CHAI staff to undertake field visits to monitor whether household visits are taking place. Call centre based calling to index TB patients to validate household visits by CHWs.”
    • “We will monitor actual practice through field visits and call centre. Any deviations would lead to CHAI staff working with their immediate supervisors. If it persists we will escalate it to supervisors / district officials.”
    • “Key to success is ensure robust governance mechanisms are in place at all levels. Thus, if a certain district doesn’t take action it gets escalated to the state level. An additional lever would be CHAI teams ability to influence change through immediate supervisors and any informal interactions with CHWs.”
  • ASHAs:
    • “Monitor ASHA performance through project M&E tools, as well as supervisory visits by CHAI district staff to identify ASHAs / localities where there is need for additional trainings / supervision.”
    • “Implement a Reward and recognition program to recognize best performing ASHA, and best performing blocks to motivate ASHAs and their supervisors”
  • PPSA-hired CHWs:
    • “Train PPSA staff, provide necessary tools, establish processes to streamline tasks, and ensure intensive monitoring of TPT activities.”
    • “We will also provide additional financial incentives [to finance and encourage TPT initiation activities]”

• 94

  See detail on CHAI’s relevant previous experience here.

• 95

  See the “Increase in TPT coverage rates” subsection of GiveWell, “Intervention Report on Household contact management for tuberculosis” for more details.

• 96

  GiveWell, Learning from CHAI's TB HCM pilot in India (unpublished)

• 97

  CHAI, "Quality and Market shaping learning agenda," 2024 (unpublished).