In a rural district of eastern Ethiopia, trained nurses ran twelve sessions with families: how to wash hands with soap, how to store water, how to use a latrine, where to put waste. Fieldworkers came back every two weeks to record who had been sick. When the numbers were tallied, diarrhea among children under five had fallen by 35 percent compared with households that got no sessions.
That trial is one of 15 gathered in a new systematic review from a team at Brazil's Fundação Oswaldo Cruz, led by Arthur Santos Lima and Viviane Sampaio Boaventura, published in PLOS Neglected Tropical Diseases. The researchers screened 1,917 records and kept only randomized controlled trials that tested community-based education aimed at the parents and caregivers of children under five. What survived spans 33 years and three continents: 44,794 caregivers in Honduras, Brazil, Bolivia, the Gambia, Ghana, Tanzania, Ethiopia, China, India, Pakistan, Nepal, Bangladesh and the United States. The smallest trial enrolled 51 families in the slums of Dhaka. The largest covered 16,301 people in western Honduras.
Diarrhea is the third leading cause of death in children under five, the authors note, with roughly 1.7 billion childhood cases a year worldwide. Most of it traces back to unsafe water, poor sanitation and hands that carry fecal bacteria to food. Education is the cheap lever: teach a mother to wash before cooking, to boil water for a toddler, to mix oral rehydration solution correctly, and in principle you prevent illness without laying a single pipe.
Twelve of the 15 trials reported something positive. In Karachi, Pakistan, fieldworkers who visited weekly with soap, slide shows and videotapes saw diarrhea incidence fall 53 percent. In the Gambia, a program built around performing arts and public meetings, teaching families to reheat stored food and dry utensils on clean surfaces, cut reported diarrhea sharply at six months, and some of that gain was still visible 32 months later, after the program itself had ended. In Handan, China, hygiene training delivered through parent sessions, posters and radio dropped hand contamination with coliform bacteria from 9.5 percent to 2.0 percent, a rare case where a study measured the biology rather than asking a parent to remember.
Where the lessons stopped working
The more interesting findings are the failures. In rural Tanzania, a large trial promoting handwashing and sanitation did change behavior: more latrines got built, less open defecation. Diarrhea prevalence did not budge in any group, including the one that got both campaigns. In a periurban stretch of Bolivia, education alone produced only a modest, non-significant reduction. Households given a hollow fiber water filter saw diarrhea prevalence fall by roughly 79 percent, and adding the education campaign on top did not improve on the filter. Nearly 72 percent of households there were drinking heavily contaminated water delivered by tanker truck. No amount of instruction fixes that.
The review's authors are careful about what they can and cannot claim. They ran no meta-analysis, because follow-up ran anywhere from 2 to 32 months and the trials measured outcomes in incompatible ways: incidence, prevalence, incidence density, knowledge scores, self-efficacy scales. They assessed every study for bias and found several rated high or unclear risk, usually because the randomization was described too vaguely or nobody was blinded. Most trials relied on a mother recalling the past seven or fourteen days, which invites people to report what they think the visiting fieldworker wants to hear. The team also did not run a formal GRADE assessment of overall evidence certainty, which they flag as a real limitation of their own work.
One Ghana trial shows how tangled the picture gets. A cholera outbreak hit the district mid-study, prompting authorities to hand out water treatment tablets. Diarrhea in the comparison group fell from 20.1 percent to 7.0 percent on its own, which may well have hidden whatever the health volunteers were accomplishing.
Why it matters
Caregiver education is one of the few child health tools that a poor district can afford at scale, and this review is the first to look specifically at it rather than folding it into broader water and sanitation programs. Its answer is a qualified yes: the approaches that worked best combined several channels at once, home visits plus visual materials plus sustained behavior change campaigns, delivered often enough to stick. A subgroup analysis in Ghana found benefits only where volunteers reached at least 70 percent of the community and spent at least 30 minutes per visit. Intensity is not a detail.
The review also carries a warning worth heeding by anyone designing such a program. Behavior improved in trial after trial; health did not always follow. In Bangladesh, an intervention cut diarrhea by 22 percent, yet children in both groups gained weight identically and stayed at 76 percent of the standard weight-for-age. Knowing how to wash your hands does not give you clean water, a latrine, or enough food. The authors' conclusion is that education belongs inside water and sanitation systems rather than as a substitute for them, and that the trials themselves need to get better: standardized protocols, objective measures instead of recall, longer follow-up.