Update: The meta-analysis from Michael Kremer, Stephen Luby, Ricardo Maertens, Brandon Tan, and Witold Więcek mentioned in this blog, was published as a working paper on the University of Chicago’s Becker Friedman Institute for Economics’ website on March 30, 2022. You can read it here.
Evidence Action has been operating our Dispensers for Safe Water program at scale for over eight years, reaching over four million people across rural Uganda, Malawi, and Kenya. This program provides communities with free and reliable access to safe water, a proven method of reducing diarrheal disease—the second leading cause of mortality for children under five. Our delivery of Dispensers for Safe Water draws on our ongoing assessment of the strength of the evidence base around the impact of water treatment on child mortality, and the program’s resultant cost-effectiveness.
We are excited to share a growing evidence base that points to a much greater impact of safe water than previously estimated. Given new research demonstrating significant decreases in child mortality as a result of safe water, we are actively exploring both expansion of our Dispensers for Safe Water footprint and new water treatment options via our Accelerator.
Robust scientific experiments require large sample sizes, especially if researchers are looking to detect an impact on rare outcomes, such as death. Because of this, existing evidence on water treatment and child survival is mostly non-experimental or focuses on intermediate proxy outcomes such as caregiver-reported diarrhea. The recent analyses, however, focus on directly analyzing reductions in child mortality resulting from water treatment interventions—and their findings suggest that safe water has a prominent role to play in improving child survival.
A new working paper from Johannes Haushofer, Michael Kremer, Ricardo Maertens, and Brandon Joel Tan shows that four years of water treatment in rural Kenya reduced all-cause under-five mortality by 63%. This independent research included communities served by our Dispensers for Safe Water program as one of two treatment groups, with the researchers estimating that the program is 20 times more cost-effective than the WHO’s “highly cost-effective” threshold.
We are also encouraged by a forthcoming (not yet peer-reviewed) meta-analysis from Michael Kremer and colleagues which incorporates these results with those from a number of randomized controlled trials of other water treatment interventions. This meta-analysis finds a significant decrease of about 30% in mortality of children under five as a result of water treatment.
We are therefore evaluating the growing evidence around safe water to update our cost-effectiveness estimates for our Dispensers for Safe Water program. We are undertaking this analysis thoughtfully to ensure we fully incorporate the broadened evidence base, consistent with our long-standing approach of regularly reviewing the evidence for our programs.
The Haushofer et al study is of particular relevance for us as the research was specific to chlorine dispensers, and we are encouraged by their findings. However, the study was not originally intended to measure mortality. The forthcoming meta-analysis, meanwhile, is making use of a broader evidence base across different interventions and geographical contexts. This is particularly helpful given potential differences in treatment effects across time and space due to fluctuations in the disease environment. Most importantly, when looked at holistically, the new evidence points to a much greater ability of water treatment to save lives of very young children—and the cost-effectiveness of doing so.
We are also excited about the potential for other water treatment interventions to have similar impacts on reducing diarrheal disease and averting child mortality, particularly in communities that don’t have access to safe water, but where dispensers aren’t necessarily the best solution. These communities tend to be urban and peri-urban (Dispensers for Safe Water operates in rural areas), with access to piped water that is untreated—and hence unsafe to drink. We are currently testing an intervention in our Accelerator called in-line chlorination, a simple method to consistently deliver chlorine into piped water systems, to understand its cost-effectiveness and potential for scale.
We have been confident in the evidence supporting Dispensers for Safe Water’s impact and cost-effectiveness for many years, and reach millions of people annually through our last-mile supply chain and deep community connections, with support from key donors and partners. Excitingly, this emerging evidence shows that our program is potentially much, much more impactful and cost-effective than we thought; in light of this, we are actively exploring program expansion – and complementary safe water interventions—to reach many millions more.