GiveWell's CEA of vitamin A supplementation (public)

	A	C	E	F
1	Guide to interpreting this cost-effectiveness analysis

2	Sheet descriptions
3
4			Main CEA	This sheet contains the primary calculations that result in final cost-effectiveness estimates. This sheet draws on inputs from the "Inputs" sheet and inputs calculated in the other supplemental sheets listed below.
5			Counterfactual mortality	This sheet contains our calculations of how high mortality rates would be in the absence of VAS campaigns.
6			External validity	This sheet contains our calculations of the external validity adjustments we apply to the effect of VAS on mortality, based on differences between the populations treated in studies and the populations treated by grantee programs.
7			Leverage/Funging	This sheet contains our calculations of the impact that crowding in and crowding out funding by other contributors has on the benefits of the program.
8			Inputs	This sheet contains all of the manually entered input values informing the CEA, along with the sources and reasoning supporting them. This sheet also contains inputs sourced from the "GBD estimates" sheet.
9			GBD estimates	This sheet contains disease burden and population datasets downloaded from the Institute of Health Metrics (IHME)'s Global Burden of Disease (GBD) project.
10			Simple CEA	This sheet contains a simplified version of the CEA that is designed to be comparable to our CEAs of other programs.
11			Sensitivity analysis	This sheet contains the outputs of a sensitivity analysis macro that tests the cost-effectiveness ranges that would result from updating certain inputs to their 25th or 75th percentile values.
12
13	Terminology key
14
15		Terms specific to this CEA
16			VAS	Stands for vitamin A supplementation; describes an intervention in which vitamin A supplements are delivered to a population in an effort to reduce rates of vitamin A deficiency.
17			VAD	Stands for vitamin A deficiency; describes a condition resulting from insufficient dietary intake of vitamin A that is associated with increased morbidity and mortality.
18			Stunting	Describes a condition in which a child has a height-for-age measurement more than two standard deviations below average. We use rates of stunting as one of several proxies for rates of vitamin A deficiency.
19			Wasting	Describes a condition in which a child has a weight-for-age measurement more than two standard deviations below average. We use rates of wasting as one of several proxies for rates of vitamin A deficiency.
20			Anemia	Describes a condition in which a person does not have enough healthy red blood cells or hemoglobin in their blood to transport oxygen through their body. We use anemia rates as one of several proxies for rates of vitamin A deficiency.
21			Poverty headcount	Refers to the percentage of a population that has income or consumption levels lower than a given poverty benchmark.
22			Fixed-effects	A fixed-effect meta-analysis assumes that the true impact of the intervention is consistent in every study in the analysis. Differences between study results are assumed to be solely down to chance. The main estimate from the analysis can be interpreted as the typical intervention effect.
23			Random-effects	A random-effects meta-analysis assumes that the true impact of the intervention varies and follows some distribution. The main estimate from the analysis can be interpreted as the average impact of the intervention across heterogenous studies with different findings.
24
25		Terms used across GiveWell's CEAs
26			Internal validity	Describes an adjustment we make to the treatment effect of an intervention to account for the possibility that the treatment effects found in studies may not represent the true effect the intervention had on the populations studied.
27			External validity	Describes an adjustment we make to the treatment effect of an intervention to account for differences in the program implementation or populations treated in studies from the program implementation or populations treated by grantee programs.
28			Leverage	Describes a situation where a grantee's spending on a program causes other organizations or governments to contribute more to the program than they otherwise would have. In most cases, accounting for leverage increases cost-effectiveness
29			Funging	Describes a situation where a grantee's spending on a program causes other organizations or governments to contribute less to the program than they otherwise would have. In most cases, accounting for funging decreases cost-effectiveness.
30			In-kind contributions	Describes non-financial (e.g., staff time, office space) contributions to a program that might otherwise have been used for other activities or programs.
31			Counterfactual	In most cases, describes the state of the world that would exist if we did not provide funding to a grantee for a program. When discussing the "counterfactual value of other actors' spending," we are referring to how much benefit another organization's or government's spending would generate if it were spent on something other than the grantee program.
32			Moral weights	To compare cost-effectiveness across different programs, we use ‘moral weights’ to quantify the benefits of different program impacts (e.g. increased income vs reduced deaths). We benchmark the value of each benefit to a value of 1, which we define as the value of doubling someone’s consumption for one year.
33			Philanthropic actors	Non-governmental organizations (NGOs) providing funding to philanthropic programs.
34
35	Unit and source key
36
37		Unit labels used in the CEA
38			#	Number
39			per 1k	Rate per 1,000
40			per 100k	Rate per 100,000
41			$	U.S. dollars
42			ln($)	The natural logarithm of a monetary value
43			%	Percentage
44			ppt	Percentage points
45			UoV	Units of value: an arbitrary unit GiveWell uses to compare the moral value of different types of outcomes, such as saving lives or increasing income
46			xcash	Cost-effectiveness in terms of multiples of GiveDirectly's unconditional cash transfer program
47
48		Source labels used in the CEA
49
50			Input	A value pulled from the "Inputs" sheet
51			Calc	A value calculated using other values in the same sheet
52			Supp	A value pulled from one of the sheets hosting supplemental calculations
53			Main	A value pulled from the "Main CEA" sheet
54			Feed	A value pulled from an earlier section within the same sheet
55
56	GiveWell analyses informing this model
57
58		Analyses specific to this CEA
59			GiveWell's analysis of Helen Keller International's cost per supplement
60			GiveWell's analysis of coverage estimates from RCTs of VAS from Imdad et al. 2017
61			GiveWell's analysis of baseline VAS coverage in regions of Madagascar
62			GW's analysis of baseline mortality for Imdad et al. 2017
63			GW's analysis of recent vitamin A deficiency surveys
64			GW's analysis of disease-specific mortality effects of VAS
65			GW's analysis of baseline VAS coverage
66			GW's analysis of subnational mortality rates for VAS
67			GiveWell, VAS supplementary analysis, 2024
68
69		Analyses referenced across GiveWell's CEAs
70			GW's moral weights and discount rate
71			GW's supplemental intervention-level adjustments
72			GiveWell's CEA for GiveDirectly's unconditional cash transfers
73			GW's analysis of the counterfactual value of other actors' spending
74			GW's analysis of the counterfactual value of Global Fund spending
75