SCI Foundation — Support for Deworming and Worm Burden Mapping (January 2022)

Published: March 2022

Planned activities and budget

• Support for deworming programs in six countries for one year and three countries for two years ($16.4 million, 81% of the grant).1 These are countries where we model the cost-effectiveness as above our bar for making a grant (discussion below). The grant will support deworming in Madagascar, Burundi, and Mauritania for two years (following SCI Foundation's fiscal year: April 2022-March 2023 and April 2023-March 2024). We plan to determine whether to renew that funding by the end of the first year. The grant will also support deworming in Côte d'Ivoire, Democratic Republic of Congo (DRC), Liberia, Malawi, Sudan, and Tanzania for one year (April 2022-2023). Over the next year, we plan to gather more information to determine whether and how to continue supporting these programs.
• Fund worm burden mapping in DRC and Madagascar ($2.9 million, 14% of the grant).2 SCI Foundation will collect fecal and urine samples from children selected from across DRC and Madagascar and test the samples for presence and intensity of schistosomiasis and soil-transmitted helminth eggs. These results will be used to inform where and how often to conduct deworming in the future.
• Support one year of deworming in three locations as part of GiveWell’s exiting from these locations ($0.9 million, 4% of the grant).3 Our cost-effectiveness analysis of deworming in Uganda, Niger, and Zanzibar finds that these programs do not meet our bar for cost-effectiveness for ongoing support.4 To aid in the transition from prior GiveWell and UK government support, this grant includes one year of funding for these programs.

The case for the grant

There is a different case for each part of the grant: (a) directly cost-effective impact through support for deworming in nine countries, (b) generating data that may be helpful in assessing the cost-effectiveness of programs and in designing more cost-effective programs, and (c) providing an exit period for three locations.

The grant as a whole was time-sensitive. SCI Foundation was previously supporting the majority of its work in Côte d'Ivoire, DRC, Liberia, and Niger with funding from the UK government, which ended its funding for deworming in 2021, earlier than previously expected.5 Funding recommended by GiveWell has historically been most of the rest of SCI Foundation's funding.6 Our most recent recommendation of funding to SCI Foundation7 was in November 2020, at which time we wrote, "SCI's current and expected funding, along with this grant of $3.1 million, will allow SCI to maintain its current work through 2022. Since this spending will be driven by available funding much more than ongoing revenue, SCI will be facing a large budget shortfall in 2023. We plan to decide in the next year whether we believe that SCI's use of additional funding in 2022 and 2023 is competitive with other opportunities to which we could direct funding." SCI's need for funding was then accelerated by the UK government's decision to end its funding for deworming in early 2021, leaving a large funding gap for 2022.

A note on how we use cost-effectiveness estimates in our grantmaking

After assessing a potential grantee's room for more funding, we may then choose to investigate potential grants to support the spending opportunities that we do not expect to be funded with the grantee's available and expected funding, which we refer to as "funding gaps." The principles we follow in deciding whether or not to fill a funding gap are described on this page.

The first of those principles is to put significant weight on our cost-effectiveness estimates. We use GiveDirectly's unconditional cash transfers as a benchmark for comparing the cost-effectiveness of different funding gaps, which we describe in multiples of "cash." Thus, if we estimate that a funding gap is "10x cash," this means we estimate it to be ten times as cost-effective as unconditional cash transfers. As of this writing, we have typically funded opportunities that meet or exceed a relatively high bar: 8x cash. We also consider funding opportunities that are between 5x and 8x cash.

Note that our cost-effectiveness analyses are simplified models that do not take into account a number of factors. For example, in the case of deworming, we have not yet attempted to model the possibility that current programs will contribute to eliminating the targeted parasites and thus could allow programs to cease treatment in the future while having lasting benefits. There are limitations to this kind of cost-effectiveness analysis, and we believe that cost-effectiveness estimates such as these should not be taken literally due to the significant uncertainty around them. We provide these estimates (a) for comparative purposes to other grants we have made or considered making, and (b) because working on them helps us ensure that we are thinking through as many of the relevant issues as possible.

Directly cost-effective impact

Based on our cost-effectiveness analyses of deworming in Burundi, Madagascar, Mauritania, Côte d'Ivoire, DRC, Liberia, Malawi, Sudan, and Tanzania, we believe these programs are in the range of cost-effectiveness of programs we expect to direct funding to, as of 2022. Our estimate is that these programs range from 9 to 36 times as cost-effective as GiveDirectly's program.8

Differences in cost-effectiveness across countries in our cost-effectiveness analyses of deworming programs are generally driven by:

• Worm burden adjustment: For every deworming grant, we ask the potential grantee to provide data on the prevalence and/or intensity of infections for each species of schistosoma and soil-transmitted helminths in the locations where it would support deworming. We focus on the prevalence of moderate-intensity infections and of heavy-intensity infections for each species, rather than the average infection intensity for each species. We then apply an adjustment to our cost-effectiveness estimates of deworming programs to account for differences between the prevalence and intensity of worm infections in the geographies targeted by our deworming grantees and the prevalence and intensity of worm infections among the population studied in Miguel and Kremer 2004, the randomized controlled trial (RCT) on which we base our estimate of deworming's impact on consumption. We recently made an update to the worm burden adjustments we were using for a number of SCI-supported programs that increased our estimate of cost-effectiveness in these locations.9 A key uncertainty for this grant is the extent to which the data we are using in the model corresponds to the specific sub-national areas that will be targeted for deworming under this grant.
• Cost per child dewormed: We have been using a single estimate for this in our cost-effectiveness analysis, based on results from all SCI-supported programs that we had data from.10 In practice, cost per child may vary across countries.

We have decided to fund two years of the program in Madagascar, Burundi, and Mauritania because we believe that these are the locations where there is the strongest case that SCI is particularly well-positioned to support the programs. We will assess whether to provide a grant to SCI to extend funding past March 2024 by the start of SCI's fiscal year in April 2023.

For the other six countries, we are interested in reassessing the size, cost-effectiveness, and possible implementing partners for the funding gaps. We have included funding in this grant for SCI Foundation to continue these programs for a year while we consider the options for the future. When choosing between grantees for the same program, our decisions are often driven by our qualitative assessments of grantees and the extent to which grantees design programs that align with the evidence that our cost-effectiveness analyses are based on.

Worm burden mapping

The main case for funding worm burden mapping in DRC and Madagascar is to improve decisions about these two programs in the future. These are two large programs for which we have estimated high cost-effectiveness and are thus fairly likely to continue supporting at some level in the future. As worm burden can vary widely within a country, the mapping data may make it possible to target our future support to areas where it will be above our bar for cost-effectiveness.

SCI told us that it would expect the largest updates from new mapping surveys relative to currently available data in DRC, Madagascar, Liberia, and Côte d'Ivoire. Of these, we decided to start with supporting surveys in DRC and Madagascar because they would inform the largest of these programs. In both DRC and Madagascar, the mapping data that is currently in use is not based on a single national survey with consistent methods, but rather several surveys conducted at different times in different parts of the country with some differences in methodology. In DRC, surveys in use were conducted between 1976 and 2014. In Madagascar, there is a similar situation to DRC with baseline data being informed by different sources at multiple time points and with variable methods.11

We also expect to learn from this process and be able to make more informed decisions in the future about the value of these types of surveys. The methods that SCI Foundation and others are now using for mapping surveys have evolved. Mapping surveys based on the new protocol have been or are being conducted in Ethiopia and Zimbabwe. The DRC and Madagascar surveys will add to this evidence base on how remapping changes treatment strategies and cost-effectiveness and will give us direct experience of following the process. In parallel, the World Health Organization has rolled out new guidance and recommendations which change which populations require deworming, treatment frequency, and interventions required.12

Exit funding

We estimate that deworming in Niger is about as cost-effective as cash transfers and in Uganda and Zanzibar is less cost effective than cash transfers.13 See discussion above about the cost-effectiveness level that we look for in grants we recommend. We previously supported programs in Uganda and Zanzibar because, as we wrote before, "We focus on SCI's overall cost-effectiveness because, based on our past experience with SCI and our understanding that SCI has access to a substantial amount of non-GiveWell-driven flexible funding, we don't currently believe that we can either predict or drive how SCI will use marginal funding." Niger was supported with funding from the UK government.14

The differences in our estimates of cost-effectiveness across countries are driven largely by the worm burden adjustment. It is possible that we will learn more about worm burden levels in these locations and decide in the future to provide additional support.

We have made a rough guess that one year of funding for these programs is sufficient for a manageable exit period. In that time, other funders may decide to fund some or all of this work, or SCI may decide to put its unrestricted funding toward continuing the programs. It's possible that this exit will lead to programs being disrupted.

Risks and reservations

• Our typical process is to reduce the size of grant we recommend by the amount of funding that the grantee has available and would otherwise put toward the activities funded by the grant. We have not done that in this case. Our default is to fund a program out three years in the future. As we have not done that here, we want SCI to have the flexibility to deploy its $5.7 million in available funding toward activities in the second and third years. (This grant covers most planned activities in the first year; some costs for mapping surveys were not included.) SCI also holds $7.6 million in reserve funding.15
• We have not revisited our worm burden estimates in light of the shifting funding landscape. Funding now may be supporting different parts of each country than it has in the past, which may cause us to misestimate cost-effectiveness.
• We have only a high-level understanding of the methods that will be used in the worm burden surveys. We might be able to add features to the surveys that would make them more useful for our future grant decisions, but we have elected not to spend time on that at this stage. It's also not clear to us how the data will be used to change plans for who is provided with deworming treatment and how often and whether that will make the programs more or less cost-effective.

Plans for follow up

SCI provides annual updates on spending, number of people reached, and coverage surveys from the last year. SCI also shares informal updates by email and in conversation. In particular, we plan to ask more about how the funding landscape and plans for whom to provide deworming to and on what frequency are changing in each country.

Worm burden mapping activities in DRC and Madagascar are expected to extend into the April 2023-March 2024 fiscal year,16 so we'd guess that we'll see results in 2024.

Our process

We had a pre-existing cost-effectiveness analysis for each of the countries supported by this grant. We did not revisit the analysis as part of investigating this grant.

We had several conversations with SCI about its funding needs, and it provided budgets and information on available funding. We also spoke with Deworm the World and Sightsavers about this grant.

For internal review, a Program Officer who was not otherwise involved in the grant investigation gave feedback on the plan for investigating the grant, and a Senior Researcher reviewed the internal grant proposal.

Sources

Document	Source
GiveWell, 2022 GiveWell cost-effectiveness analysis — version 1	Source
GiveWell, "GiveDirectly," 2020	Source
GiveWell, "Qualitative Assessments of Top Charities," 2020	Source
GiveWell, "Recommendation to Open Philanthropy for Grants in November 2020," February 2021	Source
GiveWell, "Recommendation to Open Philanthropy for Grants to Top Charities," 2019	Source
GiveWell, "SCI Foundation," 2021	Source
GiveWell, SCI grant budget— January 2022	Source
GiveWell, "Sightsavers — Support for Deworming in Kaduna, Zamfara, Niger, Kano, Adamawa, and Katsina in 2021-2023 (October 2021)"	Source
GiveWell, Summary of GiveWell's 2020 Worm Burden Adjustment Model (public)	Source
Miguel and Kremer 2004	Source

• 1

  See this spreadsheet, "Deworming funding above GiveWell's cost-effectiveness bar" row.
  

• 2

  See this spreadsheet, "Worm burden mapping" row.
  

• 3

  See this spreadsheet, "Exit funding" row.
  

• 4

  See cost-effectiveness estimates for these programs, in terms of multiples of cash transfers, here.

• 5

  Email from SCI, December 2, 2021 (unpublished). See here for general discussion of UK government cuts for deworming.

• 6

  See here.

• 7

  Other than the annual "incentive grant" we have historically asked Open Philanthropy to make, which is intended to ensure that top charities receive a minimum amount as compensation for regular sharing of information.

• 8

  See this spreadsheet, "SCI Foundation" sheet, "Cost-effectiveness in multiples of cash transfers, after all adjustments" row.

• 9

  From conversations with SCI, we came to believe that worm burden surveys in many of the areas SCI supports used sampling methods that would be less likely to cause bias (such as random sampling) than the methods we previously thought had been used. See this spreadsheet.

• 10

  See this spreadsheet, "Cost per person dewormed (per year) — SCI" row.

• 11

  Conversation with SCI, January 24, 2022 (unpublished).

• 12

  Conversation with SCI, January 24, 2022 (unpublished).

• 13

  See cost-effectiveness estimates for these programs, in terms of multiples of cash transfers, here.

• 14

  Email from SCI, December 2, 2021 (unpublished).

• 15

  See here.

• 16

  See this spreadsheet, "Source: Group 1" and "Source: Group 2" sheets, "Reassessment survey" row.