Preventing diarrheal infections with household water treatment: lessons from simulation models

TitlePreventing diarrheal infections with household water treatment: lessons from simulation models
Publication TypeThesis
Year of Publication2012
AuthorsEnger, KScott
Thesis Typephd

Diarrheal disease kills two million children per year in developing countries. Diarrhea is controlled in industrialized countries by systems to remove sewage and distribute clean water. These systems are difficult to fund, build, and maintain in developing countries, so simpler technologies are promoted: e.g., household water treatment ({HWT)}, handwashing, and latrines. These technologies, however, require consistent effort by individuals for proper use and maintenance, defined here as 'compliance'. Measuring compliance is difficult, and often neglected. It is important to understand how the extent and pattern of compliance within communities affects the prevention of diarrhea by {HWT}, while accounting for biases in field trials and characteristics of natural transmission systems, such as the presence of multiple pathogens, transient spikes of contamination, and multiple transmission routes. This question was answered by: reanalyzing and generalizing results from a {HWT} field trial, using a quantitative microbial risk assessment ({QMRA)} model to adjust for bias (chapter 3); examining the joint effects of {HWT} antimicrobial efficacy and compliance on prevention of diarrhea (chapter 4); and using a model of diarrheal infection transmission incorporating multiple routes of infection to further examine efficacy and compliance issues with {HWT} (chapter 5). The {QMRA} model of the field trial found that compliance greatly affected {HWT} effectiveness: with low compliance, 10% of diarrhea was prevented; with high compliance, 90% was prevented. It also estimated source water pathogen concentrations source water that were consistent with measurements from other developing countries. The model found that the effect of an imperfect placebo device used during the field trial depended on the assumed level of compliance during the field trial. The {QMRA} model was modified to examine how {HWT} compliance and antimicrobial effectiveness jointly altered diarrheal disease risk. Given perfect compliance, increasing antimicrobial effectiveness always lowered risk. If compliance was incomplete, increasing antimicrobial effectiveness eventually ceased to lower risk, except in a few scenarios with high incidence, high compliance, or large water contamination spikes. The pattern of compliance by communities also influenced risk; e.g., risk was lower if 90% of people used {HWT} perfectly and 10% never used {HWT}, than if 100% of people used {HWT} 90% of the time. A preliminary transmission model simulated a community in which infected people shed pathogens, which were ingested by other people via exposure to land, drinking water, their household environment, or visits from other households. It found similar results to the {QMRA} models regarding compliance. Transmission of diarrheal pathogens by household visits appeared unimportant compared to other routes; however, visits consisted of exchanges of pathogens between household environments. Other types of visits (e.g., shared child care) might lead to greater transfer of pathogens and a greater influence of visits on illness. The model also inferred that viruses and protozoa were attenuated (removed from the system, e.g., by decay or sequestration) {\textasciitilde}10 times faster than bacteria. Future sensitivity and uncertainty analyses will highlight important aspects of the model and its parameters that contribute to its results. The amount and pattern of compliance strongly affects diarrhea prevention by {HWT.} Further research should be conducted on improving compliance with {HWT} in developing countries.