ABCDEFGHIJKLMNOPQRSTUVWXYZ
1
CostNotes
2
Intervention cost per pregnant woman treated$7.33
3
Hypothetical cohort10,000Arbitrary. Changing this value does not affect the bottom line estimates.
4
Cost to treat cohort$73,300
5
6
Maternal anemia reduction
7
Proportion of cohort with anemia100%
Estimate hack: "Over 50% of pregnant women in LMICs suffer from anaemia" Salam 2015, Pg 1
8
RR of anaemia with intervention0Assumed upper bound of 28% reduction (based on custom meta-analysis)
9
Internal validity adjustment100%
Generally strong internal validity across included RCTs, as vetted by other meta-analysis researchers and qualitatively/briefly by GiveWell
10
External validity adjustment100%
Reduced 20% (percentage is very arbitrary) due to iheterogenity in programs (e.g., iron supplementation) and effects (I^2=83%)
11
RR of anemia with intervention adjusted for internal and external validity0.00
12
Proportion of population 'cured' of anemia100%
13
YLD burden of a year of moderate anemia0.0522013 GBD weights from Salomon et al (2015) - Table 2.
Chris has some additional thoughts on DALY weights for anemia here.
Further thoughts on DALY weights for anemia here: https://docs.google.com/document/d/10m2ZcYfmjjl3fFU-6kw6bXyjFz-YV57HKz3eULw309o/edit#heading=h.gi5pd1pem8ui
14
Total anemia YLDs averted in cohort520.0
15
16
Cost per anemia YLD averted$141
17
Anemia YLDs averted per $10,00070.94
18
19
Newborn anemia reduction
20
Proportion of cohort with anemia100%Assuming 1:1 with mother
21
RR of anaemia with intervention0Assuming 1:1 with mother
22
Internal validity adjustment90%
Generally strong internal validity across included RCTs, as vetted by other meta-analysis researchers and qualitatively/briefly by GiveWell
23
External validity adjustment80%
Reduced 20% (percentage is very arbitrary) due to iheterogenity in programs (e.g., iron supplementation) and effects (I^2=83%)
24
RR of anemia with intervention adjusted for internal and external validity0.28
25
Proportion of population 'cured' of anemia72%
26
YLD burden of a year of moderate anemia0.0522013 GBD weights from Salomon et al (2015) - Table 2.
Chris has some additional thoughts on DALY weights for anemia here.
Further thoughts on DALY weights for anemia here: https://docs.google.com/document/d/10m2ZcYfmjjl3fFU-6kw6bXyjFz-YV57HKz3eULw309o/edit#heading=h.gi5pd1pem8ui
27
Total anemia YLDs averted in cohort374.4
28
29
Cost per anemia YLD averted$196
30
Anemia YLDs averted per $10,00051.08
31
32
Newborn cognitive benefits
33
Proportion of newborn with anemia at baseline100%Assuming 1:1 with mother
34
Change in intelligence quotient from intervention33%
0.25 SD in Sachdev et al (2005), 0.33 SD in Low et al (only for anemic children), ~0.25 for other cognitive tests in Christian et al (2010)
35
Percentage increase in wages/consumption for every 1 point increase in IQ1.00%
Same methodology as in the GiveWell's iodine CE estimate, using information from DCP2, 'Economic Benefits of Long-Term Improvements in IQ' section here: https://www.ncbi.nlm.nih.gov/books/NBK11783/
36
Speculative adjustment to assess the long-term benefit of many years of iron10.00%
We see increases in IQ after supplementation for less than 1 year. It could be that there's a linear effect (e.g. one year of supplementation raises IQ by 5 points, 2 years raises IQ by 10 points). That's extremely implausible though. More likely, going from low haemoglobin levels to normal haemoglobin levels leads to an IQ increase of a couple points. I would guess that those cognitive benefits remain as long as someone is non-anemic and mostly disappear if haemoglobin levels return to the initial low levels. More details here: https://docs.google.com/document/d/15ifHdYl-mI6ZF_SgksADD3myusxya3ZBYrPHLKt9_ao/edit?usp=sharing
37
Percentage increase in wages/consumption from increase in IQ from supplementation0.50%Definitional—1 sd of IQ = 15 points
38
Increase in annual ln(consumption) for beneficiaries0.0049
39
40
Internal validity adjustment70%Weak to moderate quality evidence base
41
External validity adjustment40%
Discount by ~50% for arguments about why short-term benefits may not extend to long-term; discount by ~20% for arguments related to external validity (anemia was more severe in studies on effects of IQ than in this target population). More details here: https://docs.google.com/document/d/10m2ZcYfmjjl3fFU-6kw6bXyjFz-YV57HKz3eULw309o/edit#heading=h.n5l1r9eaukrc and here: https://docs.google.com/document/d/12Wl1EKFoKa1GpiDNhw1umfpUGObHY1vUZGWZB6buZms/edit
42
43
Adjusted increase in annual ln(consumption) for beneficiaries0.0014
44
Average increase in annual ln(consumption) in cohort (including non-beneficiaries)0.0014
45
Average number of years between 'entering fortification program' and beginning of long term benefits16.40
Assumes earnings begin at 16 years and a midpoint intervention age of -4.5 months (halfway through pregnancy)
46
Discount rate3.00%
Copied directly from "Early life growth intervention BOTEC", could customize for future iterations
47
Duration of long term benefits of fortification (in years)30Median duration of benefits for deworming was 30 in GiveWell CEA, Nov 2016
48
49
Present value of sum of average lifetime income benefits (including non-beneficiaries)0.0167
50
Multiplier for resource sharing in households (beneficiaries will eventually have children)200%Consistent with the multiplier used in GW's 2018 V1 deworming CEA
51
Sum of lifetime income benefits per individual in cohort, accounting for resource sharing0.033
52
Total lifetime income benefits for full cohort, in terms of ln(income)333.78
53
Units of increase in ln(income) per $10,00045.54
54
55
Reaching a common unit for comparison
56
Moral weights
57
Value of 1 YLD1Arbitrary starting point
58
Value of doubling consumption for one person for one year1
https://docs.google.com/spreadsheets/d/1zLmPuddUmKsy3v55AfG_e1Quk-ngDdNzW-FDx0T-Y94/edit#gid=1362437801&range=B3
59
Value of increasing ln(consumption) by one unit for one person for one year1.44
60
61
Units of value per $10,000
62
Units of value from maternal anaemia YLDs averted70.9
63
Units of value from newborn anaemia YLDs averted51.1
64
Units of value from income benefits from newborn cognitive effects65.7
65
Overall units of value per $10,000187.7
66
67
Comparison to cash
68
GiveDirectly units of increase in ln(consumption) for each $10,000 donated25.2
See Row 21, Sheet "Cash", here: https://docs.google.com/spreadsheets/d/1mUbO8D5hWKn8nM3a2ZtD6zrRFP83LN5HBQzxXOarbM4/edit#gid=1680005064
69
Units of value per $10,000 from GiveDirectly36.4
70
Deworming for pregnant women is X times as cost-effective as GiveDirectly5.2
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100