Twenty-Three Cases of Diarrhoea After the Rain: Finding the Source
Between 10 and 16 June, 23 children under ten in a 45-household settlement in Kanchanpur developed watery diarrhoea; the first cases appeared two days after 90 mm of rain fell in one afternoon. Most families use a 6 m deep hand-pump tubewell set in a concrete apron that is cracked on the uphill side. A pit latrine stands 9 m from the tubewell, and its soil is sandy. Eight households on the far side of the settlement use a 30 m deep dug well, and none of their children is ill. The health post has oral rehydration salts and zinc tablets but no water-testing kit; the sub-health post has chlorine solution, and a social worker wants to chlorinate the tubewell immediately because 'the water looks clear'. A mother says she boils water only for her eight-month-old baby, not for the older children.
1. Which water source is most likely responsible, and what is the transmission route? Use the evidence, not general advice.
Read the worked answer
The hand-pump tubewell. Three separate lines of evidence converge on it:
- Timing - cases began two days after 90 mm of rain, which matches the incubation period of common bacterial enteric infections and is exactly what a cracked apron would allow: runoff flushing contamination down the casing.
- Construction - a shallow 6 m well with a crack on the uphill side has no effective sanitary seal; surface water reaches the water directly. A 9 m separation from a pit latrine in sandy soil is below the usual 15 m minimum, and sand transmits water quickly.
- The natural comparison - the eight households using the 30 m dug well on the far side have no cases. That is what a source-specific outbreak looks like, rather than a general food or person-to-person outbreak.
Route: faecal–oral, waterborne - pathogen from the latrine → sandy soil → runoff and shallow groundwater → tubewell → drinking water → mouth. 'The water looks clear' is irrelevant: pathogens are invisible at these concentrations, and clarity is not a test.
2. The social worker wants to pour chlorine into the tubewell at once. Assess this plan.
Read the worked answer
Partly right, wrongly sequenced.
- It does not stop recontamination. With a cracked apron, the next rain re-infects the well. The effect is temporary and may create false confidence.
- The dose is unknown. Chlorine demand depends on turbidity and organic matter. A guessed dose may be too weak to work or strong enough to make the water undrinkable.
- It cannot be verified. You want roughly 0.2–0.5 mg/L free residual after 30 minutes of contact, which requires a test kit the health post does not have.
- It is not universal. Chlorine is much less effective against Cryptosporidium and against pathogens sheltered inside particles.
- It is awkward. Dosing a hand-pump well without a proper seal is hard to do safely.
Correct sequence: protect the source first (repair the apron, seal the casing), test before and after, then shock-chlorinate if the test supports it. For households in the meantime, a rolling boil for one minute with covered cooling and safe storage is the most reliable option available, because it does not depend on dose calculations the team cannot check.
3. Put the actions in priority order and say how you would know within two weeks whether they worked.
Read the worked answer
- Break the route today - advise all 45 households to boil drinking water or use a safe household method with safe storage, and stop drawing from the tubewell where an alternative exists.
- Treat and count - ORS and zinc at the health post, refer any child with sunken eyes, no urine, blood in stool or repeated vomiting; keep a line list by household and date of onset.
- Repair the source - rebuild the apron sloping away from the well, cut a drainage channel to divert runoff, seal the casing, and later relocate the latrine at least 15 m away and downhill.
- Test and disinfect - sample before and after repair; shock-chlorinate once the well is protected.
- Educate - handwashing with soap at critical times, covered storage, no open defecation.
Knowing whether it worked: plot the epidemic curve - no new case with onset more than about one incubation period (2–7 days for common bacterial diarrhoeas) after the intervention should appear; check free chlorine residual in household storage; repeat the water test. If cases continue despite safe water, look for a food handler or person-to-person spread rather than blaming the well again.