Reducing Childhood Lead Exposure in Ghana

How to cite

Timmis, Hannah; Crawfurd, Lee; Osei, Carl Stephen; Salifu, Usamatu; Sandefur, Justin. 2026. "Reducing Childhood Lead Exposure in Ghana." AEA RCT Registry. AEARCTR-0019098. https://doi.org/10.1257/rct.19098-3.2

Download: BibTeX · RIS — for reference managers without a browser translator. Zotero and most others pick up this page's citation metadata directly.

Hannah Timmis* , Lee Crawfurd , Carl Stephen Osei , Usamatu Salifu , Justin Sandefur

* Primary investigator

Registry ID
AEARCTR-0019098
DOI
https://doi.org/10.1257/rct.19098-3.2
First registered
Last updated
Status
On going
Trial dates
2026-03-08 – 2026-11-30
Countries
Private
JEL codes
I18, Q53, C26
Registry record
http://www.socialscienceregistry.org/trials/19098

Abstract

Childhood lead exposure remains a major public health problem in low- and middle-income countries, where around half of children have elevated blood lead levels that can cause lifelong harm to their health and development. A key challenge in addressing this problem is the limited evidence on which lead sources contribute most to child exposure, and therefore which interventions are likely to be effective. This study addresses this gap by experimentally evaluating the effects of removing two commonly contaminated household products — cookware and traditional eyeliner — on children's blood lead levels. We randomly assign 300 households in Ghana's Northern Region, a known lead exposure hotspot, to receive lead-free replacements for contaminated cookware, contaminated eyeliner, both products, or neither, and compare changes in children's blood lead levels across treatment and control households using venous blood samples collected at baseline and endline. The findings will identify the contribution of these two products to child lead exposure, informing the prioritisation of lead-remediation strategies and the design of a second, larger-scale randomised trial to estimate the causal effects of reduced lead exposure on children's cognitive development.

Intervention

Households are randomly assigned to receive lead-free replacements for contaminated cookware, contaminated eyeliner ("chilo"), both, or neither. Treatment households have their cookware and/or eyeliner tested for lead using a portable X-ray fluorescence (XRF) analyser; items exceeding reference thresholds are replaced with lead-free equivalents and removed from the household. Compliance is monitored through monthly follow-up visits.

Experimental design

A randomised controlled trial testing whether providing lead-free replacements for contaminated cookware and/or traditional eyeliner reduces children's blood lead levels. 300 children aged 4–6 in Ghana's Northern Region, selected for elevated BLL and regular use of both products, are randomised into four arms: control, cookware replacement, eyeliner replacement, and both. BLL is measured via venous blood sample at baseline and approximately three months after product delivery. Treatment effects are estimated using two-stage least squares, instrumenting product use with random assignment.

Design details

Randomization method
Randomisation conducted by computer. Within each locality (stratum), children were ranked by screening BLL, divided into blocks of six, and randomised in a 2:2:1:1 ratio across the four arms; remainders not divisible by six were randomised independently using the same allocation probabilities.
Randomization unit
Individual child/household.
Clusters
Not applicable (individual-level randomisation, not clustered).
Observations
300 children
Arms
101 households control, 102 households cookware replacement, 49 households eyeliner replacement, 48 households cookware and eyeliner replacement.
Primary outcomes
Child blood lead level (BLL), measured in µg/dL via venous blood sample analysed by ICP-MS, at endline (approximately three months after product delivery).
Secondary outcomes
Contaminated cookware use and contaminated eyeliner use, each constructed as the child's exposure-weighted use of contaminated products (lead concentration × frequency of use over the previous seven days). These serve as first-stage/mechanism outcomes in the IV analysis.
Sponsors
Name: University of Oxford Location: Research Governance, Ethics & Assurance, Boundary Brook House, Churchill Drive, Headington, Oxford Url: https://www.ox.ac.uk/research/support/governance-and-committees
Partners
Name: Pure Earth Partner_type: ngo Url: https://www.pureearth.org/

Cited by

No citation data available for this registration yet.

Keywords