Improving global health: A closer look at WASH interventions
Q&A with Andrew Brouwer and Joe Eisenberg
Assistant professor of Epidemiology; Professor of Epidemiology
Diarrheal diseases continue to pose a significant global public health challenge, particularly for children under 5 in many communities worldwide. Despite substantial investments in water, sanitation, and hygiene (WASH) interventions, large-scale trials have struggled to achieve the expected health improvements. A recent study led by researchers at the University of Michigan School of Public Health offers new insights into why these interventions sometimes fall short, and how to approach the developments of more effective public health strategies.
In this interview, Andrew Brouwer, assistant professor of Epidemiology, and Joe Eisenberg, professor of Epidemiology, discuss the use of advanced simulation models to understand the complex dynamics of WASH interventions.
Can you walk us through the original WASH Benefits trials in Bangladesh and Kenya? What motivated this research, and what made the initial trial results so surprising?
WASH Benefits Bangladesh and its sister trial WASH Benefits Kenya, were two large, well-designed and well-implemented randomized controlled trials intended to assess whether combined WASH (and combined WASH and nutrition) interventions were more effective at reducing diarrheal disease and linear growth faltering than single interventions. These studies were conducted from 2012-16 and the results published in 2018.
WASH Benefits Bangladesh found only modest reduction in diarrhea in most intervention arms, no effect on linear growth, and no additional benefit to combined interventions. WASH Benefits Kenya did not even find a reduction in diarrhea. The trial results were surprising and set off many important discussions about why the effects were less than expected and how to move forward as a field.
Our recent work was meant to contribute to this discussion by revisiting the WASH Benefits Bangladesh results using mechanistic models that can evaluate various hypotheses addressing why the trials estimated modest to no health benefits. We believe that our work can help lead to better-targeted public health WASH interventions, policy recommendations, and field trials.
Your research highlighted the importance of community-wide interventions. Can you explain how the percentage of households implementing WASH strategies affects overall health outcomes, and how local conditions influence intervention success?
We found that increasing community coverage—the percent of households within a community that receive the intervention—improved the effectiveness of the interventions. This result is an important finding as it means that, for example, the extent to which your neighbors have improved sanitation facilities matters to your health, and therefore interventions should be focused at the neighborhood or community level rather than the individual or household level.
We also found that the effectiveness of the interventions was reduced when pre-existing WASH conditions were poorer or when the baseline disease burden was higher—it’s harder to intervene when there’s more disease pressure. This fact was underscored by the WASH Benefits Kenya trial, which had 27% diarrheal prevalence in the control arm compared to 5% in Bangladesh. We think this is a significant reason why the Kenya trial did not find any health benefits associated with the WASH interventions.
Diarrheal diseases seem to spread through multiple pathways. Could you elaborate on these transmission routes and explain why targeting just one pathway is often insufficient?
One reason that it is difficult to control the spread of enteric pathogens that cause diarrheal disease is that they can exploit numerous environmental transmission pathways. For example, they can be spread through contact with contaminated water, food, contaminated surfaces and objects (called fomites), and animals, among other things. Therefore, depending on the environmental and social context and the specific pathogens circulating, different pathways may be more dominant than others. Since any given intervention may only target one or a few of these pathways, the optimal control options for diarrheal disease will likely vary from site to site.
Our analysis echoes these challenges. We found that matching the intervention with the specific transmission pathways that caused the most disease for a given setting was a major factor in how effective an intervention or set of interventions were at reducing disease. This result emphasizes the importance of site-specific interventions that target those pathways that are exploited by local pathogens. We need to move beyond the water, sanitation and hygiene interventions traditionally considered when trying to reduce diarrheal disease burden, and more explicitly account for other pathways such as food and animals.
The trials showed that combined interventions didn't perform as expected. What insights did your modeling reveal about the potential effectiveness of multi-pronged WASH approaches?
Because transmission occurs along multiple pathways, combined interventions targeting multiple pathways should be more effective than single interventions targeting single routes. The initial analysis of the WASH Benefits Bangladesh data did not find this to be the case, but our model analysis suggests that combined interventions should still be expected to be more effective.
Based on your research, what recommendations would you provide to public health organizations and policymakers looking to invest in WASH interventions?
Our work underscores the importance of developing intervention strategies that focus on high community coverage and that target multiple transmission pathways. Similar to the concept of herd immunity, the importance of attaining high community coverage highlights the significance of the concept we call herd protection, or the percent coverage that can interrupt transmission.
We also found that it is more difficult to decrease the disease burden in areas with greater diarrheal disease prevalence or lower quality preexisting WASH infrastructure. This result emphasizes that high-burden areas will likely need multiple, mutually reinforcing strategies to achieve health gains.
More broadly, our work highlights the challenges of focusing on randomized controlled trials alone. A randomized controlled trial is a powerful design to test whether an intervention is effective or not. However, their results are difficult to generalize to other contexts and are not able to answer why an intervention did successfully reduce disease burden.
Moving forward, the WASH community will benefit from the integration of practitioners with local experience, intervention trialists with the experience in designing high-quality studies, and modelers who can project potential study outcomes.
Paper cited: “Understanding the Effectiveness of Water, Sanitation, and Hygiene Interventions: A Counterfactual Simulation Approach to Generalizing the Outcomes of Intervention Trials,” Environmental Health Perspectives. https://doi.org/10.1289/EHP15200
Additional authors of the study include: Mondal Zahid and Marisa Eisenberg, University of Michigan School of Public Health; Benjamin Arnold, University of California, San Francisco; Sania Ashraf and Mahbubur Rahman, International Centre for Diarrhoeal Disease Research; Jade Benjamin-Chung and Stephen Luby, Stanford University; John Colford Jr. and Amy Pickering, University of California Berkeley; Ayse Ercumen, North Carolina State University; Alicia Kraay, Gates Foundation; Matthew Freeman, Emory University
Media Contact
Destiny Cook
PR and Communications ManagerUniversity of Michigan School of Public Health734-647-8650