Nutrition International — Vitamin A Supplementation Renewal, Chad (May 2023)

Note: This page summarizes the rationale behind a GiveWell-recommended grant to Nutrition International. Nutrition International staff reviewed this page prior to publication.

Published: July 2023

Summary

What we think this grant will do

With this grant, Nutrition International will provide financial and technical assistance to the government of Chad to support implementation of vitamin A supplementation (VAS) campaigns. This grant extends previous support we provided for the program for an additional year (through 2025).

Vitamin A deficiency is a common condition in low- and middle-income countries that can cause stunting, anemia, dry eyes (the leading cause of preventable childhood blindness), susceptibility to infection, and death. VAS involves treating all children aged approximately six months to five years in areas at high risk for vitamin A deficiency with high-dose vitamin A supplements, usually twice per year.

We think this grant will lead to an increase in the number of children with vitamin A deficiency receiving VAS, which will, in turn, lead to reduced mortality and other health and income-related benefits during the rest of these children's lives.

For more about what we think this grant will do, see the "Planned activities and budget" section below.

Why we made this grant

• We believe that Nutrition International's program is highly cost-effective. Intuitively, this is because we think:
  • Nutrition International's program will increase the number of children with vitamin A deficiency that receive vitamin A supplements. We are not aware of other mechanisms (e.g., government funded health systems or existing VAS campaigns) by which a meaningful proportion of children are already receiving VAS in Chad. We think this is due to lack of funding and technical guidance, and we think Nutrition International will address those barriers by providing financial and technical support for VAS campaigns.
  • We believe that vitamin A supplementation is likely to reduce child mortality. Evidence from randomized controlled trials (RCTs) of VAS suggests it reduced all-cause mortality rates for children in the trial contexts by 24%. After adjusting for how those results might translate to the context of Nutrition International's program, we roughly estimate that the program reduces mortality by approximately 12%.
  • Rates of child mortality are high in Chad. Based on the 2019 Global Burden of Disease Project model, we estimate 17.8 deaths per 1,000 child years in children aged 6 to 59 months in Chad, in the absence of VAS programs.
  • VAS is relatively cheap and straightforward to deliver. We estimate a round of vitamin A supplementation costs $1-$2 per child covered, due to the low costs of supplementation1 and the opportunity to frequently deliver VAS alongside other health interventions such as vaccination campaigns.
  • We think there are funding gaps for VAS campaigns in Chad. On a global level, there is a need for more funding for VAS, especially as opportunities to co-deliver VAS with other interventions like polio vaccination campaigns are decreasing. For Chad specifically, we are not aware of another funder who would be very likely to fill this funding gap in the near-term. Even after modeling the possibility that we are crowding out funding from other sources, this grant remains cost-effective.
  Our best guess is this grant is approximately 20 times as cost-effective as cash transfers, which is above our current bar of 10x cash. A sketch of our cost-effectiveness analysis and level of uncertainty is in the table below:

  	Best guess	25th-75th percentile range for key parameters	Cost- effectiveness over that range ("x cash")
  Grant size	$6,000,000
  Child mortality benefits
  Cost per child receiving VAS due to Nutrition International's program	2.31	$1.70-$2.80	17x-28x
  Number of children reached from grant to Nutrition International	2,597,403
  Percentage of those children who would have received VAS without Nutrition International's program	15%	5%-35%	16x-23x
  Annual child mortality rate	1.8%	1.0%-3.5%	12x-41x
  Reduction in child mortality from VAS	12%	6%-16%	10x-28x
  Cost per child death averted	$1,289
  Moral weight for each child death averted	119
  Cost-effectiveness from child mortality benefits only (x cash)	27x
  Breakdown of benefits from primary benefit streams
  Child mortality benefits	93%
  Development effects	7%
  Additional upside and downside adjustments	70%
  Supplemental charity-level adjustments	82%
  Supplemental intervention-level adjustments	167%
  Downstream costs to government incurred as a result of funding to NGO (leveraging)	98%
  Chance that other funders would have funded the program in our absence (funging)	52%
  Final cost-effectiveness (after all adjustments)	21x

  $6m / (2.6m * (1-15%) * 1.8% * 12%)

  Reducing deaths of children under the age of five is the largest source of benefits we model for Nutrition International's program, but we believe the program has developmental benefits as well. This section summarizes how much different benefit streams contribute to the overall effect of the program. Note that this section only includes benefit streams we directly model. We adjust for additional benefits and downsides to the program in the following section.

  We believe there are likely additional benefits and downsides of the program beyond those modeled above. This section summarizes our best guess of how those additional considerations contribute to the overall impact of Nutrition International's program.

  Adjustments of <100% indicate we believe those additional factors reduce the overall impact of the program, while adjustments >100% indicate additional factors that we believe increase the program's overall impact.

  It's often difficult to explicitly model different kinds of wastage or other charity-level issues directly in our cost-effectiveness analysis. These adjustments take into account our best guess of how different charity-level factors (e.g., false monitoring results, possible changes in charity priorities, etc.) affect our best guess of true cost-effectiveness.

  We only directly model some of the possible effects of the intervention. Many effects, such as averted treatment costs, are excluded to keep the model tractable.

  This row sums the speculative guesses we use to estimate the impact of outcomes we haven't explicitly modeled. The "Supplemental intervention-level adjustments" tab of our cost-effectiveness analysis includes more details on how these adjustments are calculated and some brief justifications.

  "Leverage" refers to GiveWell funding causing other entities to spend more on the program than they otherwise would have. We assume this funding would have been spent on programs that create value. We account for this by subtracting the counterfactual value of spending caused by GiveWell funding.

  In this case, we think a grant to Nutrition International will cause the government of Chad to devote in-kind resources to support VAS campaigns. We think that without our funding the government of Chad would devote those resources to other health programs. Our downward adjustment reflects the funding the government of Chad spends as a result of GiveWell's funding and the value that would have been created by programs it otherwise would have supported with that funding.

  "Funging" refers to GiveWell funding causing other entities to spend less on the program than they otherwise would have. See this blog post for a description of our approach to leverage and funging.

  Since our initial cost-effectiveness estimate of 27x cash only incorporates the benefits of reducing child mortality, we divide that estimate by 93% to account for the additional benefits attributed to development effects. We then multiply by our total estimate for additional upside and downside adjustments, to account for the impact of additional factors not directly modeled. Thus, (27 / 93%) * 70%.

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

• We believe this grant will allow us to learn more about Nutrition International and inform future decision-making. We started supporting Nutrition International in January 2022 and believe this grant will provide the opportunity to learn more about the organization and its VAS program. What we learn through this grant could inform the amount of funding we direct to Nutrition International in the future. The value of this learning is not included in our cost-effectiveness estimate above.

For more about why we made this grant, see "The case for the grant" section below.

Main Reservations

• What is the child mortality rate? Record keeping on child deaths in the locations where GiveWell has supported VAS is not reliable. Our analysis uses estimates of child mortality from the Global Burden of Disease tool, which rely on a number of modeling assumptions. We have not triangulated these against other data sources or expert opinion on child mortality in Chad relative to other countries. Our estimate of the 25th-75th percentile for child mortality is 1.0%-3.5%, which implies a cost-effectiveness of 12x-42x. We don’t expect to update this parameter as a result of this grant.
• How effective is vitamin A supplementation at reducing child mortality? There is conflicting evidence on the effectiveness of VAS at reducing child mortality.2 The estimate we use in our cost-effectiveness analysis (that VAS reduces child mortality by 12%) relies on studies conducted between 1983 and 2009, most of which use data collected more than 30 years ago.3 We are unsure how relevant these results are to today's context, given changes in vitamin A deficiency rates and the proportion of deaths due to vitamin A-susceptible diseases. Our estimate of the 25th-75th percentile for VAS efficacy in reducing child mortality is 6%-16%, which implies a cost-effectiveness of 11x-29x. We don’t expect to update this parameter as a result of this grant.

For more information, see the "Risks and reservations" section below.

Funding for the grant

This grant was funded by $2.5 million in donations to Effective Altruism Funds' Global Health and Development Fund, about $1.9 million in donations to GiveWell's All Grants Fund, and about $1.6 million in donations to Effektiv Spenden for GiveWell's All Grants Fund.

Planned activities and budget

This $6 million grant will enable Nutrition International to support implementation of biannual vitamin A supplementation (VAS) campaigns in Chad targeting children aged 6 to 59 months. Vitamin A deficiency is a common condition in low- and middle-income countries that can cause stunting, anemia, dry eyes (the leading cause of preventable childhood blindness), susceptibility to infection, and death.4 Vitamin A supplementation involves treating all children aged approximately 6 to 59 months in areas at high risk for vitamin A deficiency with high-dose vitamin A supplements.

The budget for this grant includes:

• Approximately $4.3 million for program activities. Program activities include campaign activities, vitamin A capsule costs, and post-event coverage (PEC) surveys to be conducted after the campaigns.
• Approximately $900,000 for program management. Program management costs include support for country staff, global staff, and travel.
• Approximately $800,000 for indirect costs.5

This grant extends our previous support to Nutrition International's program for an additional year (through 2025) and provides top-up funding to account for an increase in the budgeted costs for 2023 and 2024.6 Funding Nutrition International’s VAS campaigns for three years into the future is consistent with the three-year funding runway we typically provide to grantees. See the breakdown of this grant funding by year in the table below:

The case for the grant

• We estimate that this program exceeds our bar for cost-effectiveness. More below.
• We do not believe this program would be fully funded absent GiveWell's support. More below.
• We have a positive impression of Nutrition International's work so far. More below.

Cost-effectiveness

Based on our cost-effectiveness analysis (CEA) of Nutritional International, we believe its VAS program in Chad meets our bar for directing funding. We estimate that Nutrition International's VAS program in Chad is 20 times as cost-effective as unconditional cash transfers.8 When we created a cost-effectiveness model of this program for our January 2022 grant, we relied heavily on the work we had done creating a cost-effectiveness model of Helen Keller International's VAS program.9 Since that grant, we have made various updates to the Helen Keller International cost-effectiveness analysis, which we have now incorporated into the Nutrition International model as well.10 While considering this grant, we incorporated the following updates into our cost-effectiveness model of Nutrition International:

• We updated our internal validity adjustment: We base our estimate of the impact of vitamin A supplementation on child mortality on the effect size reported in Imdad et al. 2017, a meta-analysis of vitamin A supplementation (VAS) trials. In 2022, we revisited our internal validity adjustment for that meta-analysis, and concluded that its results likely overestimate the true impact of VAS on mortality.11 Consequently, we changed our internal validity adjustment from 85% to 74%, which negatively impacted our cost-effectiveness estimate for this program, as it reduces our estimate of the number of deaths averted by vitamin A supplementation.12
• We updated our external validity adjustment: Our external validity adjustment is intended to account for differences between the contexts of the trials included in Imdad et al. 2017 and the contexts of programs we're considering funding. Specifically, we base our external validity adjustments for VAS grants on estimates of the child mortality rate, the prevalence of vitamin A deficiency (VAD), and the proportion of child mortality caused by infectious diseases, measles, and diarrhea in a given location. In 2022, we updated two aspects of our methodology for calculating the external validity adjustment for VAS programs: we slightly increased the amount of weight we put on the proportion of deaths caused by infectious diseases (compared to that caused by measles and diarrhea), and we fixed an error in our calculations that was resulting in inflated external validity adjustments for some countries (including Chad).13 In considering this grant, we also updated our estimate of VAD prevalence in Chad.14 Together, these updates changed our external validity adjustment for VAS in Chad from 81% to 58%.15 This lowers our cost-effectiveness estimate for Nutrition International's program, as it implies that the results from the VAS trials are not as generalizable to the current program context in Chad as we previously estimated.
• We updated our adjustment for treatment costs averted: In our Helen Keller International CEA (and our CEAs for other life-saving interventions), we include a supplemental adjustment for “treatment costs averted from prevention” because we think it’s likely that reduced disease morbidity and mortality resulting from its program leads to reductions in medical costs borne by governments and recipient households. Modeling treatment costs averted from prevention led us to update our estimate from a rough guess of 6% to 20%.16 Incorporating this update into our Nutrition International CEA increased our cost-effectiveness estimate of the program, as it implies that the program results in greater benefits to governments and recipient households than we previously estimated.

Altogether, these updates decreased our cost-effectiveness estimate of Nutrition International's program in Chad from 30x as cost-effective as cash transfers to 20x cash. This estimate is still well above our current bar for directing funding (10x cash).

Funding landscape

Funding for VAS campaigns globally

Based on conversations with funders and implementers supporting VAS campaigns, our understanding is that the primary sources of funding for VAS other than GiveWell are domestic governments, Global Affairs Canada, and UNICEF. Global Affairs Canada, via grants to Nutrition International and UNICEF, seems to be the primary other funder of VAS in countries with the highest child mortality rates, and we do not expect its future support to be significantly higher than its past support, though this is uncertain.17 We also expect funding for VAS from UNICEF, beyond its funding from Global Affairs Canada, to be limited, though we don't have a strong understanding of how UNICEF's country offices decide where and when to support VAS.18

Another consideration in the global funding landscape for VAS is whether VAS campaigns are able to be co-delivered with other campaigns, such as polio vaccination campaigns. VAS can be delivered more cheaply when it piggybacks on another campaign, and historically polio campaigns were the chosen vehicle.19 As polio campaigns decreased in frequency over the last few years, costs of VAS campaigns increased.20 However, polio campaigns may increase again due to an increase in polio cases, which may again change VAS implementation possibilities.21 Nutrition International currently expects to be able to integrate one of its biannual VAS campaigns with a vaccination campaign each year.22 The cost of the program could substantially increase if Nutrition International is unable to find opportunities for integration and instead delivers two standalone VAS campaigns per year.23

Given the global picture of limited and possibly slightly shrinking funding for VAS, paired with increasing costs to deliver VAS as a standalone intervention rather than as an integrated campaign, we believe it is unlikely that this grant will cause much funding to be crowded out of the VAS funding landscape in highly cost-effective locations.

Funding for VAS campaigns in Chad

Our understanding of the funding landscape for VAS campaigns in Chad has not changed much since our last grant investigation. Nutrition International informed us that UNICEF has limited funding for VAS campaigns in Chad, though UNICEF may be receiving new funding from Global Affairs Canada.24

In making this grant, we may be crowding out some funding for VAS campaigns in Chad that would have come from other funders. However, we do not believe another funder would fill the full funding gap for VAS campaigns in Chad. Even after modeling the probability of crowding out some funding in our cost-effectiveness analysis, this grant remains highly cost-effective. As with our previous grant to Nutrition International, we continue to roughly estimate that each $1 we contribute crowds out $0.40 from other philanthropic sources.25 We believe most of the risk is that we may crowd out funding from UNICEF and vaccination campaigns.

Positive impression of Nutrition International

We made our first grant to Nutrition International in January 2022, to support implementation of biannual VAS campaigns in Chad from 2022 to 2024. Based on conversations with Nutrition International, we believe it achieved its coverage targets (i.e., approximately 75% of children targeted to receive VAS were reached during the VAS campaigns) in the first grant year.26 From discussion with Nutrition International and a cursory review of its coverage survey report for the June 2022 campaign, we do not have any atypical concerns with its monitoring approach or the quality of the survey results.27

Risks and reservations

• Child mortality rates: Record keeping on child deaths in the locations where GiveWell has supported VAS can be unreliable.28 Our analysis uses estimates of child mortality from the Institute for Health Metrics and Evaluation's Global Burden of Disease tool; these estimates rely on a number of modeling assumptions.29 If baseline child mortality is actually lower than estimated, then VAS would avert fewer deaths than we estimate. If child mortality is higher than estimated, we could be underestimating the impact of this grant.
• Uncertainties about the effect of VAS on child mortality: Our estimate of the impact of vitamin A supplementation on child mortality is based on the effect size reported in Imdad et al. 2017, a meta-analysis of vitamin A supplementation (VAS) trials. The most heavily weighted study in that analysis was published in 2013 and did not find a statistically significant effect on mortality.30 The five other studies that receive the most weight in the analysis are about 30 years old.31 Both these facts raise questions about the impact of VAS in current contexts. We have an external validity adjustment in our cost-effectiveness analysis to adjust for differences between trial contexts and the contexts in which the VAS programs we fund will be implemented, and refining our external validity adjustments has been an ongoing area of research. Our most recent updates further increased the size of our external validity adjustment for Chad (see above), indicating that the meta-analysis results are less generalizable to the current program context than we previously estimated.
• Uncertainty about the proportion of children reached by the program that would have received VAS without the program: We roughly estimate that 15% of children reached by Nutrition International's program would have received VAS from another source.32 This estimate is based on our understanding that VAS is sometimes available during visits to healthcare clinics in other countries in sub-Saharan Africa.33 We may overestimate cost-effectiveness if either (a) Nutrition International's support leads to providing VAS to more children who would have received VAS without its support than we estimate, or (b) our adjustment for double treatment does not fully account for the extent to which the additional children reached may be systematically at lower risk of illness and death than those who are reached by the routine health system.34 If the number of children who would not have received VAS without support is less than the number we estimate or if these children are at higher risk than those reached by the routine health system, we may be underestimating cost-effectiveness.
• Uncertainties about Nutrition International as a grantee: We made our first grant to Nutrition International in January 2022. As when we start working with any new grantee, we have more uncertainty about Nutrition International’s work than that of our long-term partners because we do not yet have a history of working with the organization. Our sense is that Nutrition International's role in supporting VAS differs from the role typically played by our long-standing VAS grantee, Helen Keller International.35 We will not have a more comprehensive understanding of the program's performance until we have received multiple years of grant implementation, spending, and coverage data. We expect to learn more about Nutrition International's strengths and weaknesses as we follow its work in Chad.

Plans for follow up

• We will ask Nutrition International to have check-in calls every three to six months to provide updates on the work funded by this grant. We plan to use those calls to ask the following questions:
  • Have VAS campaigns been happening every six months? What portion of the country have they reached?
  • Have there been opportunities to co-deliver VAS with immunization campaigns or other types of campaigns? Has Nutrition International been able to take advantage of any such opportunities?
  • Who, if anyone, has supported the VAS campaigns besides Nutrition International?
  • What coverage data has Nutrition International collected? What are potential sources of bias due to how that data was collected?
  • How does actual spending compare to the projected budget, and what are key drivers of any differences?
  • What is our updated estimate of cost per child per supplementation round based on data on the number of children reached and total spending?
  • What efforts has Nutrition International undertaken to improve coverage, and how have these efforts gone?
• We will ask for annual updates on spending, the size of target populations, and methods and findings of coverage surveys. We will use this information to update our estimate of the cost per child per VAS round in Chad.

Internal forecasts

Our process

This was a fairly light-touch grant investigation. Our process for this grant relied heavily on (a) our prior work on modeling the cost-effectiveness of vitamin A supplementation, and (b) our discussions with Nutritional International. For this particular grant:

• We had two calls with Nutrition International to get updates on the first year of the program and to discuss the proposed budget for this grant.
• We reviewed Nutrition International's proposed budget to confirm it aligns with our expectations of program costs.
• We conducted a cursory review of the coverage survey report for the June 2022 VAS campaign (the first campaign supported by our previous grant to Nutrition International).
• We updated our cost-effectiveness analysis of the program as described above.

Sources

• 1

  We estimate that procuring and shipping a vitamin A capsule costs $0.06. See here.

• 2

  See the discussion in our vitamin A supplementation report.

• 3

  See the study dates listed in this spreadsheet.

• 4

  "Vitamin A deficiency (VAD) impairs body functions and may cause death. Adverse health consequences may also include xerophthalmia (dry eyes), susceptibility to infection, stunting, and anaemia (Sommer 1996; Rice 2004).... VAD is common in low- and middle-income countries." Imdad et al. 2017, p. 7.

• 5

  See the grant budget here.

• 6

  After conducting year 1 of the program, Nutrition International identified more children requiring VAS than originally projected based on population estimates. This has resulted in an increase in program budgets for future years relative to the original budget proposed prior to Nutrition International establishing operations in Chad. Nutrition International, conversation with GiveWell, April 14, 2023 (unpublished).

• 7

  See our calculations here.

• 8

  Note that (a) our cost-effectiveness analyses are simplified models that are highly uncertain, and (b) our cost-effectiveness threshold for directing funding to particular programs changes periodically. As of June 2023, our bar for directing funding is about 10x as cost-effective as cash transfers. See GiveWell’s Cost-Effectiveness Analyses webpage for more information about how we use cost-effectiveness estimates in our grantmaking.

• 9

  See 2021 GiveWell cost-effectiveness analysis version 3, sheet "Helen Keller International." We discuss our model of cost-effectiveness for Helen Keller on this page.

• 10

  See this page for information on the changes we made to our cost-effectiveness analyses in 2022.

• 11

  See a discussion of our reasoning here.

• 12

  See our previous cost-effectiveness analysis of Nutritional International's VAS program here, and our latest version here.

• 13

  See our discussion of those updates here.

• 14

  Previously, we estimated a VAD rate of 29% in Chad. We now estimate a VAD rate of 27.66%.

• 15
  • See our external validity calculations before these updates here, and our current external validity calculations here.
  • Note that the change in the external validity adjustment is largely driven by the change we made to our methodology for calculating the adjustment for non-independence of changes in VAD and child mortality rates. Our external validity adjustment for VAS accounts for differences in both of these factors relative to trial contexts. The non-independence adjustment is intended to account for the fact that differences in vitamin A deficiency prevalence and infectious disease burden between two contexts are likely to be correlated (i.e., a reduction in VAD prevalence is likely correlated with a reduction in the proportion of mortality attributed to vitamin A susceptible diseases), so including the full change in both factors over-penalizes VAS in current contexts. Our previous methodology accounted for non-independence by applying a 30% increase to our external validity adjustments across the board. We believe this method over-inflated external validity adjustments for countries like Chad that have a greater proportion of mortality attributed to vitamin A susceptible diseases than trial contexts (see here), despite having lower VAD rates.

• 16

  See the value in our previous cost-effectiveness model here, and in our current cost-effectiveness model here. For more information about the model we use for our cost of illness averted adjustment, see this write-up.

• 17

  This understanding comes from unpublished calls with UNICEF and Global Affairs Canada.

• 18

  "In 2021, Helen Keller International significantly increased its support for Vitamin A Supplementation (VAS) in multiple countries following the large reduction in funds available from one of the main VAS supporters--UNICEF. With support from GiveWell, Helen Keller was able to close part of the funding gap and distributed ~50 million capsules through semi-annual VAS campaigns in 2021. More recently, reductions in UNICEF VAS funding in 2022 were less severe overall, but remain significant in countries such as Kenya, Nigeria or Niger. Although UNICEF thought it could provide greater support to Guinea, Mali and Côte d’Ivoire in 2022, their actual level of support remains lower than in the previous years. The likelihood of increased UNICEF support in these countries in the near future is low. Filling UNICEF’s funding gaps has led Helen Keller to spend higher amounts of GiveWell funds than initially budgeted, causing estimated funding gaps in each country beginning in 2024 and creating 'room for more funding' for these countries.” Helen Keller International, Room for more funding report, 2022, p. 2.

• 19
  • “When polio campaigns take place, the World Health Organization and the Global Polio Eradication Initiative cover the cost of the core teams, and VAS stakeholders 'only' have to support the addition of one distributor. When there is no polio campaign, VAS partners usually support two distributors per campaign, resulting in significantly higher costs.” Helen Keller International, Room for more funding report, 2021, p. 7.
  • For past Helen Keller VAS campaigns that were linked with polio campaigns, see "Table 2. Polio campaign organized in Helen Keller supported countries between 2018 and 2021 and links with VAS distribution" Helen Keller International, Room for more funding report, 2021, p. 7.

• 20

  "The frequency of polio campaigns organized in Helen Keller countries continues to decline (see Table 2).… In situations where VAS could not be 'piggy-backed' on polio campaigns, we had to increase funding support for VAS in these countries." Helen Keller International, Room for more funding report, 2021, p. 7.

• 21

  "The Covid-19 pandemic continues to significantly impact health systems and the health status of populations. In the last two years, access to routine immunization for children has reduced dramatically, leaving millions of children unprotected. As a result, cases of wild polio have resurfaced in Sub-Saharan Africa, and cases of measles are increasing rapidly. Polio campaigns are likely to be organized by WHO in 2022 and 2023, once again changing the campaign landscape in many countries." Helen Keller International, Room for more funding report, 2022, p. 2.

• 22

  Nutrition International, conversation with GiveWell, April 14, 2023 (unpublished)

• 23

  Nutrition International estimates an integrated campaign costs $400,000 to $600,000, whereas a standalone VAS campaign costs $1.2 to $1.3 million. Nutrition International, conversation with GiveWell, April 14, 2023 (unpublished)

• 24

  Nutrition International expects the new Global Affairs Canada funding would be limited and intended for targeted activities (i.e., it is not expected to support broad campaigns). We are unsure when this funding would become available. Nutrition International, conversation with GiveWell, February 13, 2023 (unpublished)

• 25

  See here in our cost-effectiveness analysis.

• 26
  • Nutrition International informed us that, according to its coverage surveys, the first VAS campaign that received GiveWell funding (the June 2022 campaign) achieved an average coverage rate of 76%. The January 2023 campaign achieved an average coverage rate of 74%. Nutrition International, conversation with GiveWell, February 13, 2023 (unpublished) and Nutrition International, conversation with GiveWell, April 14, 2023 (unpublished).
  • We have conducted a cursory review of the coverage survey report for the June 2022 campaign. We have not yet received the report from the January 2023 campaign.

• 27

  Nutrition International employs both in-country supervision and remote oversight of survey data collection, including daily checks of survey data. However, Nutrition International acknowledged that there remains a possibility of recall bias in the survey results, given that surveys are run a few months after the VAS campaigns. Nutrition International, conversation with GiveWell, April 14, 2023 (unpublished)

• 28

  For example, the Institute for Health Metrics and Evaluation explains that mortality data can be flawed: "This last piece – the underlying cause of death – is what GBD researchers use to produce mortality statistics. Understanding the underlying cause of death in a population allows public health officials to develop interventions that target the root cause. But picking out the underlying cause of death can be difficult. Many countries in the world do not have functioning vital registration systems, and even for those countries who do, there are often high levels of misclassification and vague reporting of causes of death on death certificates." Alexander, "Determining causes of death: How we reclassify miscoded deaths," 2018.

• 29
  • GBD's mortality estimates for Chad model mortality based on Demographic and Health Surveys and Multiple Indicator Cluster Surveys. Institute for Health Metrics and Evaluation, "Global Burden of Disease Study 2019 (GBD 2019) data input sources tool"
  • "GBD researchers use models to determine the relationship between the known data points that we have, and then estimate the range of possible values (researchers call this the “uncertainty interval”).... GBD models now consider thousands of data sources and millions of data points to produce measures of death and disability for hundreds of diseases and risks in countries around the world. Incorporating this many sources, and producing this many results, would not be possible without models." Gall, McKee, and Redford, "The power of models," 2018.

• 30
  • Under a random-effects model, the DEVTA 2013 study receives approximately 14% of the weight in the analysis. See here. Note that this study receives an even greater weight under a fixed-effect model (see Imdad et al. 2017, Figure 3, p. 17) but we use the results of the random-effects model in our analysis.
  • The Deworming and Enhanced Vitamin A (DEVTA) was a large VAS trial in India, with results published in 2013. The results estimate that VAS reduced child mortality by 4% and cannot rule out the possibility that VAS did not affect child mortality at all (the 95% confidence interval ranged from a 3% increase in child mortality to an 11% decrease): “Deaths per child-care centre at ages 1.0–6.0 years during the 5-year study (the primary trial endpoint) were 3.01 retinol versus 3.15 control (absolute reduction 0.14 [SE 0.11], mortality rate ratio [RR] 0.96, 95% CI 0.89–1.03, p=0.22), suggesting absolute risks of death between ages 1.0 and 6.0 years of approximately 2.5% retinol versus 2.6% control. Although this finding suggests that overall child mortality was 4% lower in vitamin A than in control blocks, this 4% reduction includes the possibility of no benefit and the possibility of appreciable benefit (95% confidence limit for reduction 11%).” Awasthi et al. 2013, p. 1473.

• 31
  • The five studies are: Ross et al. 1993, West et al. 1991, Herrera et al. 1992, Daulaire et al. 1992, and Sommer et al. 1986. For more details on these studies, see our page on VAS here.
  • See the weight accorded to each study under a random-effects model here.

• 32

  See this adjustment in our cost-effectiveness analysis.

• 33

  For example, Helen Keller International notes, "Vitamin A supplements in Kenya are administered to children aged 6-59 months through various delivery approaches. Throughout the year, children can access VAS in primary health care facilities, but this routine coverage only accounts for around 20 percent of children, essentially children below 12 months, as many caregivers do not bring their children to the health facilities after the end of the immunization contact points at one year of age." Helen Keller International, Room for more funding report, 2021, p. 21.
  We do not have information regarding the proportion of children receiving VAS through health facilities in Chad.

• 34

  The additional children reached may be older than those that receive VAS through the routine health system (if younger children receive VAS while visiting clinics for immunizations during their first year of life). We would expect these older children to be at a lower risk of illness and death; thus, we may be overestimating the mortality averted by VAS if Nutrition International is primarily reaching older children.

  • In 2019, the all-cause mortality rate among children 1-4 years old in Chad was 1.26%, compared to a mortality rate of 7.14% for children under the age of one. Institute for Health Metrics and Evaluation, GBD results tool, 2019 all-cause mortality rates in Chad
  • "Vitamin A supplements in Kenya are administered to children aged 6-59 months through various delivery approaches. Throughout the year, children can access VAS in primary health care facilities, but this routine coverage only accounts for around 20 percent of children, essentially children below 12 months, as many caregivers do not bring their children to the health facilities after the end of the immunization contact points at one year of age." Helen Keller International, Room for more funding report, 2021, p. 21.

• 35

  See our January 2022 grant page for more information.