GiveBetter x SF

Helen Keller Intl

Whole unrestricted gift with a representative vitamin A core

Research time: ~17 min on GPT-5.6 Sol Medium
  • Research — organization and evidence review.
  • Modeling — cost-effectiveness analysis.
  • Historical estimate for research done before time tracking.

Published: September 8, 2026.

Donate

Summary

What do they do? Helen Keller International (EIN 13-5562162) works in nutrition, vitamin A delivery, neglected tropical diseases and vision, including US vision services. It is distinct from Helen Keller Services. More

Why this approach interests us

  • In high-burden settings, inexpensive vitamin A delivery can reduce childhood illness and mortality.

Our main reservations

  • The unrestricted allocation and survival conversion are judgments. A very large randomized trial found no clear mortality reduction, and modern effectiveness varies by setting.

What do you get for your dollar?

We charge the whole unrestricted gift and quantify a judgmental 30% vitamin A core. Other health benefits remain unquantified. No SF or Bay health is credited—not because Helen Keller has no US programs, but because their marginal local health is unverified. Inspect the whole-gift model.

global best estimate
$8,015 / 10 QALYs — Whole gift cash; conditional representative-core model.
sf / bay credit
Zero modeled — US vision programs exist; local incremental health is unverified.
gift allocation
30% represented core — Remaining 70% health-unquantified, not assumed worthless.
full resources
Unestimated — Donor cash is not a complete societal cost.

1. What do they do?

Helen Keller International (EIN 13-5562162) works in nutrition, vitamin A delivery, neglected tropical diseases and vision, including US vision services. It is distinct from Helen Keller Services.

Support the whole organization

General gifts are flexible. Its dedicated vitamin A page restricts gifts to that work; that is a different donation boundary.

Represent one health pathway

The model assigns 30% of the marginal gift to vitamin A-equivalent health, leaving other outcomes unquantified.

Compare against existing care

Government programs, donated medicine and other grants already fund substantial services; existing recipients are not automatically additional.

Scope of this review. The core share is an analyst prior, not the audited vitamin A share or a second overhead haircut. All gift dollars remain in the numerator.

2. Monitoring and information sharing

About one million children in northern India; 36 retinol versus 36 control blocks

DEVTA cluster randomized trial, Lancet 2013. Mortality ratio .96 (95% CI .89–1.03), p=.22; no clear all-cause mortality reduction.

Our assessment. Primary indexed evidence verified; challenges universal 20–30% mortality claims but does not exclude smaller benefits in different settings.

2022–24 vitamin A opportunities; summary updated August 2026

Historical GiveWell grant modeling. $3,500 per modeled life saved.

Our assessment. Not a trial QALY result or ordinary-gift quote. Allocation and future survival are our separate judgments.

Contemporary modeled vitamin A settings

GiveWell research update, April 2026. Modern mortality reduction modeled at 1–11%, versus 19% trial average.

Our assessment. New disease, dosing and publication-bias analyses matter. A 25-times moral-weighted benchmark claim cannot be converted directly into dollars per life or QALYs.

Historical expansion to new Nigerian states

GiveWell 2025 Nigeria grant lookback. Evaluator says it would likely have made a smaller grant, excluding two states later exited.

Our assessment. Geography and monitoring can change effectiveness; historical average is not guaranteed everywhere.

3. Qualitative assessment

In high-burden settings, inexpensive vitamin A delivery can reduce childhood illness and mortality.

Key reservations

  • General gifts are not automatically restricted vitamin A funding.
  • Nutrition expenditure does not identify vitamin A's marginal allocation.
  • The large mortality-null trial and current-context uncertainty remain substantive.
  • US programs exist, but no verified marginal SF/Bay health share is assigned.
  • Scenarios are not confidence bounds.

Benefits not included in our estimate

  • Other nutrition, neglected-disease, eye-health and US vision effects not quantified here.
  • Separate income, education or blindness gains on top of the mortality bridge.
  • Unverified SF/Bay spillovers or headquarters-based benefit allocation.
  • Healthcare savings treated as donor cash and unreconciled societal resources.

4. What do you get for your dollar?

A $10,000 unrestricted gift yields an estimated 12.4769 global QALYs.

The historical anchor is $3,500 per modeled life saved. Our finite conversion assumes 80% early and 20% older deaths, 40/20-year remaining trajectories and a 2.5-year event delay. These are explicit priors, not a measured VAS death-age mix, national life table or GiveWell QALY estimate.

A better life is our comparison unit of 10 additional quality-adjusted life years (QALYs), potentially spread across people. These are uncertain estimates, not measured returns or verified donation offers.

How we calculate the estimate

WHOLE-GIFT GLOBAL MODEL: ($10,000 × .30 / $3,500) × 14.55634 gift-time QALYs per modeled death
12.47686 QALYs; $100,000 / 12.47686 = $8,014.83 per 10 QALYs

Model inputs and assumptions
Represented vitamin A core
30% (range: 5–60% joint scenarios; zero-core null). FY2025 nutrition includes vitamin A and other programs. Its broad expense category is not the next unrestricted dollar's allocation. Analyst judgment.
Historical mortality benchmark
$3,500 per modeled life (range: $2,500–$7,000 scenario assumptions). GiveWell's August 2026 summary uses 2022–24 grants, not a current unrestricted offer. Evaluator model.
Early / older death mix
80% / 20% (range: All early / all older). Provisional harmonized conversion; actual native age composition remains unresolved. Analyst judgment.
Finite survival and utility
40 / 20 years; .85 / .80 utility (range: Three-year tails). Annual survival .99/.98, 3% discount. No generic extra persistence haircut. Analyst judgment.
Gift-to-event delay
2.5 years (range: 1–5 years). Pipeline approximation, not observed current deployment. Analyst judgment.
Relative core-opportunity yield
1 (range: Half yield; zero yield). Transfers the already-adjusted historical anchor without stacking new generic coverage, efficacy or financing penalties. Analyst judgment.

What would an ordinary gift change?

  • central: $8,015 / 10 global QALYs. 12.476864 global QALYs per $10,000. Assumption scenario, not a confidence interval.
  • favorable-joint: $2,516 / 10 global QALYs. 39.746029 global QALYs per $10,000. Assumption scenario, not a confidence interval.
  • pessimistic-positive: $685,041 / 10 global QALYs. 0.145977 global QALYs per $10,000. Assumption scenario, not a confidence interval.
  • half-yield: $16,030 / 10 global QALYs. 6.238432 global QALYs per $10,000. Assumption scenario, not a confidence interval.
  • all-early: $7,364 / 10 global QALYs. 13.579378 global QALYs per $10,000. Assumption scenario, not a confidence interval.
  • all-older: $12,396 / 10 global QALYs. 8.066805 global QALYs per $10,000. Assumption scenario, not a confidence interval.
  • short-tail: $53,020 / 10 global QALYs. 1.886068 global QALYs per $10,000. Assumption scenario, not a confidence interval.
  • zero-core: No positive price / 10 global QALYs. 0.000000 global QALYs per $10,000. Assumption scenario, not a confidence interval.
  • zero-yield: No positive price / 10 global QALYs. 0.000000 global QALYs per $10,000. Assumption scenario, not a confidence interval.
  • zero-yield-shared-harm: No positive price / 10 global QALYs. 0.000000 global QALYs per $10,000. New harm diagnostic, not an observed clinical rate.
  • zero-yield-independent-harm: No positive price / 10 global QALYs. -0.100000 global QALYs per $10,000. New harm diagnostic, not an observed clinical rate.
  • gross-resource-double: $16,030 / 10 global QALYs, gross resource stress. 12.476864 global QALYs per $10,000. Adds a hypothetical $10,000 associated-resource allowance; donor-only result remains $8,015.

Uncertainty. Null core allocation or yield gives no finite positive price. Independently caused harm can remain when benefits are zero. Shared harms scale with provision; source-covered clinical harms are not deducted twice. No unmodeled lifetime, income or separate blindness credit is added.

5. Funding and previous grants

Current marginal allocation, funding room and age-specific mortality yield need better evidence. This is not a guaranteed offer.

A current unrestricted funding gap or allocation is not verified. Dedicated VAS giving is restricted and cannot be substituted silently. Government services, existing grants and capsules funded through Nutrition International and Canada form the baseline. Full resource costs remain unknown; annual in-kind accounting is not the marginal cost of this gift.

This review does not establish a verified marginal funding offer or a complete history of grants.

Donate

Opens the organization’s giving page. A general donation may not fund the specific activity modeled here; confirm allocation with the recipient.

6. Sources

  1. FY2025 Annual Report, financial table and identity. Helen Keller Intl. Primary organizational report; PDF pages 16–17. Published: April 2026 upload; fiscal year ended June 30, 2025; retrieved: September 8, 2026.
  2. Support Our Vitamin A Work. Helen Keller Intl. Primary restricted donation boundary. Published: Undated; retrieved: September 8, 2026.
  3. How We Produce Impact Estimates. GiveWell. Primary evaluator model. Published: Updated August 2026; 2022–24 grants; retrieved: September 8, 2026.
  4. DEVTA: vitamin A every six months in one million preschool children. Lancet / PubMed. Primary randomized trial; indexed abstract. Published: 2013; retrieved: September 8, 2026.
  5. Scrutinizing One of Our Longest-Funded Programs. GiveWell. Primary evaluator methodological update. Published: April 2, 2026; retrieved: September 8, 2026.
  6. Lookback: Helen Keller vitamin A expansion in Nigeria. GiveWell. Primary evaluator retrospective; indexed summary. Published: 2025; retrieved: September 8, 2026.
  7. Vitamin A Supplementation technical analysis. GiveWell. Primary evaluator payer and delivery analysis. Published: April 2024 technical page; mixed-vintage inputs; retrieved: September 8, 2026.

Annual expenses: years and sources

Average annual expenses (three consecutive fiscal years): Not available. Organization size is separate from the modeled cost-effectiveness of a donation.

Helen Keller Intl

Three years of organization-level expenses have not yet been verified for this report. No other organization or fiscal sponsor budget has been substituted.