Year: 2019
Type: editorial
Abstract: (See the Major Article by Rogawski McQuade et al., on pages 1379–86.) Improvements in water, sanitation, and hygiene (WASH) have been considered key public health measures in low- and middle-income countries [1], but recent controlled trials, including an important one reported by Rogawski et al in this issue of The Journal of Infectious Diseases [2, 3], raise questions about current approaches. Because diarrhea is caused by microbes that rely on fecal-oral transmission from infected to susceptible individuals, WASH interventions would appear to be a primary means of prevention. In low- and middle-income countries, these interventions have generally included one or all of the following: improving the water supply, typically by creating a community hand pump well; improving the water quality by treating water for possible contaminants; enhancing sanitation, primarily by providing sanitary latrines; and educating individuals about handwashing and other hygienic behaviors [1, 4]. Previous reviews estimated that each of these interventions could reduce the incidence of diarrhea by 15%–40%, but these estimates were based primarily on studies the authors deemed to be of poor quality and on observational methods rather than controlled trials [4, 5]. Recent systematic reviews using more-stringent inclusion criteria for study design have not found the same benefits. A Cochrane systematic review by Clasen et al concluded that studies of improved water-source interventions were of low quality and yielded insufficient evidence to conclude that they reduce diarrhea [6]. A later meta-analysis of 5 water-source intervention studies found no effect on childhood diarrhea (relative risk, 0.98; 95% confidence interval, .73–1.32) [7]. Recent systematic reviews of the effects of improved sanitation (ie, construction and promotion of latrines and reducing open defecation) [6, 7], identified only 2 controlled studies [8, 9]. Both had some improvement in sanitary behaviors in intervention communities, but neither found a reduction in diarrhea prevalence in children. In another recently reported randomized trial of community-led sanitation, there were substantial improvements in the use of latrines and a reduction in open defecation, but there was still no decrease in diarrhea prevalence [10]. Interventions to improve water quality, primarily chlorination or filtration, at the point of use and to promote handwashing have had consistent effects in reducing diarrhea in small trials [7], but scaling up these interventions has been difficult [11, 12]. It has also been proposed that WASH interventions may reduce growth faltering in infants and young children by reducing diarrhea and its adverse effects on growth [13] and by decreasing oral microbial exposure, which has been hypothesized to result in environmental enteric dysfunction [14]. Thus, there have been high expectations that WASH interventions could reduce stunting and have synergistic benefits with nutritional supplementation to improve growth in early childhood. Cluster-randomized trials in Bangladesh [15] and Kenya [16] evaluated several interventions separately and as a combination—provision of chlorinated drinking water, improvement in sanitation (through the creation of latrines), and promotion of handwashing with soap—and the provision of nutritional supplements with or without the combined WASH interventions, for their effects on diarrhea prevalence and stunting. A trial in Zimbabwe randomly assigned clusters of households to receive integrated WASH interventions, nutritional supplementation, both, or neither [2]. All 3 trials achieved substantial changes in WASH behaviors, and the trial in Zimbabwe had a notably high fidelity of intervention delivery (eg, 99% of households in the WASH groups had improved latrines installed vs 45% in the control group), very high reported and observed hygiene behaviors, and a focus on preventing infants from ingesting soil or other potentially contaminated materials. Only in Bangladesh was there a small reduction in childhood diarrhea, and no trial found an effect of WASH interventions on stunting, although the nutritional supplementation enhanced linear growth [2, 15, 16]. The current article by Rogawski et al provides additional important information from the trial in Zimbabwe [3]. Consistent with the lack of effect in this trial on overall diarrhea prevalence, there was no effect of the WASH interventions on pathogen-attributable diarrhea or on the prevalence or number of bacterial or viral enteropathogens at 6 or 12 months, although there was a small decrease in the number of parasites. The value of this study is enhanced by the use of state-of-the-art probe-based quantitative polymerase chain reaction assays for 29 enteropathogens. These highly sensitive methods allow attribution of diarrhea to specific pathogens and quantification of the number of pathogens per gram of stool. These methods detect pathogens in a high percentage of diarrheal stool specimens (in 94% of specimens, ≥1 pathogen was detected), and coinfection with multiple pathogens was very common, although the attribution to specific pathogens was in only a fraction (47%) of the diarrhea cases. The evidence is limited because it relies on testing routine stool samples collected at only 6 and 12 months of age and on only 30% of the diarrhea cases in a subsample of HIV-unexposed children. While this limitation does not permit the full characterization of either pathogen-attributable diarrhea or differences in enteropathogen prevalence during the 18-month postpartum study period, there is no suggestion from the available data that there is an effect of the WASH interventions. It appears from these results that the WASH interventions in the Zimbabwe trial did not change the enteropathogen exposure among infants and toddlers, consistent with the lack of effect on diarrhea or stunting [2]. It is still possible that there were effects on the intestinal microbiome that will be assessed in future analyses. Nevertheless, it seems likely that, in this setting, an integrated package of WASH interventions delivered with high fidelity to current recommendations [1] were insufficient to reduce enteropathogen transmission to infants. Although the intervention package was comprehensive and focused on protecting young children, there may have been less success with interrupting some important transmission routes. For example, contaminated complementary foods likely had a role in the period from 6 to 12 months, when enteropathogen infections increased. Because bacterial enteropathogens can quickly multiply in complementary foods kept at warm ambient temperatures, even a small inoculum can increase to numbers likely to cause infection and diarrhea [17]. It is possible that, with more-intensive promotion of behavioral changes, the included interventions would have had greater effects, as has been seen in small-scale trials of handwashing and chlorination of household water. However, the delivery by well-trained and supervised study-specific village health educators of WASH behavior change modules in 15 household visits during pregnancy and the first year of life [2] is likely already beyond what would be feasible in most low- and middle-income country programs with multipurpose village health workers. It is also possible that, in settings with a high level of environmental fecal contamination, relatively low-cost WASH technologies, such as improved household latrines or protected community-level water sources, both of which are currently being implemented, are inadequate to interrupt transmission of enteropathogens. Although there are other possible health and economic benefits from expanded use of current WASH interventions, the results of these recent trials raise doubts about the basis for previous cost-benefit analyses and the justification for the considerable additional expense of continuing to expand access and promote their use [1]. Perhaps more-advanced WASH infrastructure will be needed, but a review of the effect of household (“on-plot”) water supply concluded that the evidence was divided, only suggestive of a benefit against diarrhea, based on observational and cross-sectional studies [18]. It is undoubtedly the case that a combination of improvements in environmental conditions and hygienic behaviors will be necessary to sufficiently reduce the number of possible routes of enteropathogen transmission and protect susceptible children [19]. The limitations of currently implemented WASH interventions should make us think again about what level of WASH services are needed and how they can be cost-effectively implemented. Potential conflicts of interest. Both authors: No reported conflicts. Both authors have submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest. Conflicts that the editors consider relevant to the content of the manuscript have been disclosed.
Authors: Robert E Black, Christa Fischer Walker
Institution: Johns Hopkins University
Language: English
Subfield: Nutrition and Dietetics
Field: Nursing
Domain: Health Sciences
Open Access status: bronze
OpenAlex ID: W2942575694