Millions of people in developing countries lack access to improved drinking water sources. Unsafe water typically carries pathogens that can cause severe health problems, including diarrhea. Children are particularly vulnerable; almost two million children under age five die from diarrheal diseases annually. Diarrhea can cause severe dehydration and a depletion of essential nutrients. Frequent incidence of diarrhea can negatively affect a child’s nutritional status  , and poor nutritional status during childhood has been shown to negatively impact long-term physical, educational, and economic outcomes.  While drinking water interventions have been implemented worldwide, they vary widely in cost and have not been adequately evaluated, resulting in an urgent need for low-cost, high-impact solutions.
Solar Drinking Water Disinfection (SODIS) is a water purification technology used by two million people in over 30 countries. SODIS is a low-cost, point-of-use disinfection system that kills harmful pathogens by exposing water in plastic (PET) bottles to sunlight. SODIS has been found to be highly effective both in the lab and in highly controlled experiments conducted at the household level. However, this study is one of the first to evaluate SODIS at the community-level in a more credible low-monitoring context.
The evaluation took place in rural Quechua communities in Bolivia. Surveyed households typically obtained water from unimproved sources, storing it in plastic buckets or open containers. Only 15 percent of surveyed households had latrines in their homes and the large majority reported practicing open defecation.
To create a randomized evaluation 22 communities were matched and paired based on children’s diarrhea incidence and then assigned to either the comparison or treatment group via a public lottery event. The 11 treatment communities (225 households; 376 children) received instructions on how to properly use SODIS, free plastic bottles, and 4,385 household visits and 210 community events to promote the use of SODIS. The 11 comparison communities (200 households; 349 children) did not receive SODIS and were sufficiently far away from treatment communities to avoid spillovers. During the 51-week observational period, community field workers visited both treatment and comparison households weekly to collect diaries kept by mothers and primary caregivers to document incidents of diarrhea, cough, fever, and eye-irritation. Field workers also assessed SODIS usage by recording the number of plastic bottles exposed to sunlight, how many bottles were ready for use, and general perceptions about household usage.
Results and Policy Implications
Researchers found that households in the comparison group experienced 4.3 incidents of gastrointestinal illness per child per year, while the treatment group households experienced 3.6 incidents per child per year. The risk ratio of diarrhea between treatment and comparison groups was 0.81 (a 19 percent relative reduction) and not statistically significant at the 95 percent confidence level.
There was also no change found in the median length of a diarrheal episode (3 days), the longitudinal prevalence  of diarrhea, or the incidence of severe diarrhea or dysentery. Additionally, in a random sample of children’s stool samples, no difference was found in the pathogen load between treatment and comparison groups, further suggesting that the program effect was not significant.
According to researchers, the effect of SODIS was likely dampened by low compliance rates. Although approximately 80 percent of households reported using SODIS, field workers observed a mean usage of only 32 percent on any given day. Further, only 43 percent of households used SODIS in more than one-third of the weeks during the observation period. Only 14 percent used SODIS in more than two-thirds of the weeks. Additionally, lack of improved sanitation facilities and open defecation contributed to the presence of fecal contamination in 60 percent of household yards, resulting in additional transmission pathways for pathogens.
Despite free provision of water bottles, intensive training and education on its use, and continual promotion of the technology, the study found no evidence that SODIS led to a substantive reduction in diarrhea among children. The researchers indicate that larger studies in differing environments and contexts would be useful to fully understand the impact of SODIS. They also state that additional studies are needed to develop a better understanding of the discrepancy between laboratory and field results of SODIS interventions, as well as studies that directly compare the effects of SODIS with alternative water treatment methods.
 Checkley, W.; Buckley, G.; Gilman, R. H.; Assis, A. M.; Guerrant, R. L.; Morris, S. S.; Mølbak, K.; Valentiner-Branth, P.; Lanata, C. F.; Black, R. E.; Malnutrition, C. & Network, I. 2008. “Multi-country analysis of the effects of diarrhea on childhood stunting.” International Journal of Epidemiology. 37, 816-830.
 Black, R. E.; Allen, L. H.; Bhutta, Z. A.; Caulfield, L. E.; de Onis, M.; Ezzati, M.; Mathers, C.; Rivera, J.; Maternal and Child Undernutrition Study Group. 2008. “Maternal and child undernutrition: global and regional exposures and health consequences.” Lancet. 371, 243-260.
 Victora, C. G.; Adair, L.; Fall, C.; Hallal, P. C.; Martorell, R.; Richter, L.; Sachdev, H. S.; Maternal and Child Undernutrition Study Group. 2008. “Maternal and child undernutrition: consequences for adult health and human capital.” Lancet. 371, 340-357.
 Longitudinal prevalence refers to the number of days a child suffers diarrhea divided by the number of days of observation. It is a critical indicator for evaluating the overall burden of diarrhea.
Photo Credit: Thomas Chupein