ABCDEFGHIJKLMNO
1
Group:High coverage districtsLow coverage districtsNotes
2
3
Cohort size and incentive amounts
4
Full cohort of those eligible for the program1,0001,000Arbitrary
5
Percent of cohort in each group68%32%We calculate the proportion of the birth cohort in Sindh eligible for IRD's mCCT program that resides in districts with high baseline vaccination coverage vs. low baseline coverage. In our cost-effectiveness analysis, the children in high-coverage districts only benefit from ZM, while the children in low-coverage districts benefit from both ZM and IRD's mCCT program. This is because IRD will only be providing incentives in low-coverage districts, but will be implementing ZM in all of Sindh.
6
Incentive amount per visit (current USD)$0.00$1.26No incentive in high coverage. 200 PKR in low coverage.
7
Cohort in each group677323Calculation
8
9
Vaccine efficacy
10
Vaccine efficacy for vaccine-preventable disease from meta-analyses0.710.71Calculated based on weighted average of RR for each vaccine in schedule and that vaccine's contributions to all vaccine-preventable disease deaths. See "Vaccine efficacy and deaths among unvaccinated" tab for details.
11
Adjustment for lower vaccine efficacy in Pakistan (includes "biomarkers adjustment")0.940.94We include a smaller adjustment, compared to New Incentives, because we do not have reason for concern about biomarkers and because mCCTs trial includes corroborating biomarkers evidence.
12
Adjustment for all-cause mortality effect1.181.18Set so that 0.5 deaths from non-vaccine-preventable disease are averted for every death averted from vaccine-preventable disease. (This is similar to what we assume in our CEA for New Incentives.) This is based on evidence that reductions in all-cause mortality from some vaccines are larger than what would be expected based on reductions in deaths due to vaccine-preventable diseases alone.
13
Adjustment for coverage in trials0.950.95Rough guess.
14
Vaccine efficacy for vaccine-preventable disease from meta-analysis, adjusted0.830.83Calculation
15
16
Deaths from vaccine-preventable disease
17
Unadjusted probability of death from vaccine-preventable diseases among vaccinated and unvaccinated children in Sindh0.6%0.6%Calculated based primarily on IHME GBD data. Takes probability of death for children under 5 from vaccine-preventable diseases, accounting for deaths before vaccination and etiological fraction of certain pathogens. We assume this value from IHME includes both vaccinated and unvaccinated children. See "Probability of death" tab for details. Assume Sindh is in line with Pakistan average.
18
Percent vaccinated at time of IHME data in Sindh76.7%76.7%Calculated based on coverage of different vaccines (BCG, DTP1, etc.), weighted by their contribution to deaths from vaccine-preventable diseases. We use Sindh-wide estimates because the probability of death data we use are assumed to be the same as Sindh-wide average. See "Vaccine efficacy and deaths among unvaccinated" tab for details.
19
Probability of death from vaccine-preventable diseases for unvaccinated1.5%1.5%Calculation (see cell note for explanation of formula)
20
Adjustment for higher/lower child mortality in Sindh/targeted area among unvaccinated0.901.20Rough guess. We assume the child mortality rate is higher in low-coverage districts, which may have lower overall health beyond vaccine access. For low-coverage districts, we use an adjustment similar to the adjustment we used for child mortality rates in North West Nigeria in our New Incentives CEA. These values are set so that average (weighted by population) is 1 across districts.
21
Adjusted probability of death from vaccine-preventable diseases for unvaccinated1.4%1.8%Calculation
22
23
Effect of program on vaccination rates
24
Increase in vaccination rates from ZM, excluding mCCTs (percentage points), weighted by vaccines' contributions to deaths2.0%2.6%Effect size based on the trial.
25
Increase in vaccination rates due to mCCTs (percentage points), weighted by vaccines' contribution to deaths0.0%8.8%Effect size estimate comes from (1) effect size observed in trial plus some adjustments for how this would look at scale and (2) effect size we would expect based on New Incentives. We have high uncertainty about this parameter.
26
Total increase in vaccination rates due to ZM and mCCTs (percentage points), weighted by vaccines' contribution to deaths0.0200.114Calculation
27
28
Benefits from child deaths averted
29
Ratio of the reduction in vaccine-preventable disease mortality to the reduction in vaccine-preventable disease100%100%Assumption
30
Child deaths averted in cohort0.160.57Calculation
31
Units of value from child deaths averted in the cohort18.666.3Calculation
32
33
Additional benefits
34
Units of value from deaths averted for individuals older than 55.8620.86See "Deaths at older age groups" tab.
35
Units of value from development benefits per counterfactually vaccinated infant0.230.30We benchmark development benefits based on New Incentives. See "Development benefits" tab.
36
Units of value from development benefits3.211.2Calculation
37
Units of value from consumption benefits per vaccination paid by mCCTs0.000.02We benchmark consumption benefits based on New Incentives. See "Consumption benefits" tab.
38
Units of value from consumption benefits0.004.31Calculation
39
Units of value from additional benefits9.036.4Calculation
40
41
Benefits from inclusion/exclusion and downside adjustments
42
Total adjustment for additional benefits and negative or offsetting impacts22%22%This captures additional benefits and negative or offsetting impacts we haven't included in the model elsewhere. We assume similar values as we do for New Incentives. We may revise these with further work. See "Inclusion/exclusion" tab for details.
43
Downside adjustments for organizational quality, risk of wastage, quality of monitoring and evaluation, confidence in funds being used for intended purpose-7%-7%We assume similar values as we do for New Incentives. We may revise these with further work.
44
45
Total units of value31117Calculation
46
47
Costs
48
Unweighted increase in vaccination rate from ZM and mCCTs0.030.12Calculation. We use unweighted for costs, since the cost is the same for each vaccine in the sequence.
49
Number of infants counterfactually vaccinated2340Calculation
50
Cost of the program per eligible infant (current USD), mCCTs$1.25$9.50We include incentive costs plus other costs to add mCCTs on top of ZM platform and cost of ZM platform itself, based on conversation with IRD. See "Costs" tab.
51
Government cost for additional full immunization$16.29$16.29See "Costs" tab.
52
Gavi cost per additional full immunization$19.91$19.91See "Costs" tab.
53
Total costs to IRD per cohort$849$3,072Calculation
54
Total marginal costs to government per cohort (i.e., from additional infants induced to vaccinate)$371$657Calculation
55
Total marginal costs to Gavi per cohort (i.e., from additional infants induced to vaccinate)$453$803Calculation
56
Total costs per cohort$1,673$4,532Calculation
57
58
Cost effectiveness
59
Units of value generated per dollar spent, before accounting for leverage/funging0.01880.0258Calculation
60
Total units of value from GiveDirectly's cash transfer program generated per dollar0.00344
61
Cost effectiveness (in multiples of cash transfers), before leverage/funging, by group5.57.5Calculation
62
Cost effectiveness (in multiples of cash transfers), before leverage/funging6.9Calculation
63
Cost effectiveness (in multiples of cash transfers), after accounting for leverage/funging, by group6.48.9Calculation
64
Cost effectiveness (in multiples of cash transfers), after leverage/funging8.3Calculation
65
% change in cost-effectiveness due to leverage/funging20%Calculation
66
67
Cost per additional child vaccinated (for reference; not used in calculations)$73$112Calculation
68
69
70
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