Around 2 million children globally die each year due to preventable causes like diarrhoea and respiratory illness, because they lack sufficient water, sanitation and hygiene (WASH) amenities, like a toilet at home, piped water or a handwashing basin. Every single one of these deaths is a personal tragedy, leaving a family bereaved.

There is great demand for rigorous evidence on the impacts of health and environmental interventions on childhood mortality. For example, the UN Water Conference, held in March 2023, called for “scaling up action to address the water crisis and ensure equitable access to water for all”.

Mortality comprises 99% of the global disease burden measured in disability-adjusted life years (DALYs) due to respiratory infection and 90% of the diarrhoea DALYs. But measuring mortality accurately in intervention research is challenging. For example, a large enough sample must be collected of children who have died, which is difficult to do in any single intervention trial. The approach adopted thus far in health impact evaluations, and the systematic reviews based on them, has been to measure reported illness as a more accessible outcome, rather than mortality. But data on reported illness have been shown to be biased.

By combining data from multiple health impact evaluations, which individually could not measure mortality as a primary outcome, researchers from the London School of Hygiene & Tropical Medicine (LSHTM) provide the first rigorous and unbiased evidence on the impacts of WASH interventions on endemic disease mortality in low- and middle-income country (LMIC) contexts. The researchers compiled data from 165,000 people including 2,600 deaths that were previously reported in research publications’ participant flow charts, but which had been sitting there unanalysed until now.

It is an established finding that study participants, while misreporting information like symptoms of a common infection, do not misreport death. This might be because death of a child is a devastating premature event. The crucial advantage of this approach, therefore, is that reported mortality is less prone to bias than other reported measures.

The study team found answers to longstanding questions about how to prevent childhood mortality in normal circumstances (outside of disease outbreaks like cholera):

• When people have more water to wash with, they can wash properly, which significantly improves the survival chances of their children, reducing mortality in childhood due to any cause, primarily due to respiratory infection, diarrhoea and undernutrition, by one-third. 
• When sanitation is available to most people in a community, it lessens infection transmission by reducing children’s interactions with faeces from open defaecation, reducing diarrhoea mortality in childhood by around one-half. 
• WASH interventions can be highly pro-poor investments because there are significantly bigger effects of interventions provided in resource-poor settings, where access to WASH is very limited and mortality rates consequently highest, than in settings where access to resources was better established.

The analysis suggested that water affects mortality by enabling domestic hygienic practices around handwashing, food preparation and cleanliness. In contrast, there was no effect of interventions to improve the quality of water for drinking. The researchers also found no effects of WASH interventions on mortality among adults and children aged over 5, who are thought to have developed sufficiently robust immunity to infection.

The approach used in this study, which is applicable to many other topics in global health, is reliant on journals ensuring study authors report participant flows, which is a condition of publication in reputable journals in health. These standards urgently need to be adopted by journals in economics, which have published a substantial number of trials of WASH interventions in LMICs, none of which transparently reported the participant flows on which mortality could be measured.

Read the paper.