Identifying the Source of a Diarrhoeal Outbreak in Three Villages
A district health office in Takeo province investigates a two-week outbreak of acute watery diarrhoea. Village A has 45 cases among 500 residents; Village B has 12 cases among 400 residents; Village C has 30 cases among 600 residents. Village A relies on river water and shallow hand-dug wells; Village B uses a protected spring; Village C uses river water and rainwater jars. Latrine coverage is 40% in A, 85% in B, and 55% in C. A quick case-control study finds that 80% of cases drank river water in the past week, compared with 30% of controls. However, all 12 cases in Village B are in adults who commute daily to work in Village A; no cases have occurred among children or elderly people in Village B. The health office has limited chlorine tablets and one health education team. The team must decide where to intervene first and whether the river is the true source.
1. Calculate the attack rate in each village and compare. What does this suggest about the source?
Read the worked answer
Attack rate in Village A = 45/500 = 0.090, or 9.0%. Attack rate in Village B = 12/400 = 0.030, or 3.0%. Attack rate in Village C = 30/600 = 0.050, or 5.0%. Village A has the highest attack rate, followed by Village C, with Village B lowest. The ranking is consistent with Village A having the greatest exposure to contaminated water or the poorest sanitation, because it has the lowest latrine coverage and uses river water and shallow wells. Village B's protected spring and high latrine coverage are consistent with a lower local risk. However, attack rates alone do not prove the river is the source. The Village B data are particularly informative: all cases are commuters to Village A, which suggests they were exposed at the workplace rather than through their own village water supply. This points to a common exposure in Village A, which could be river water, food, or workplace sanitation.
2. The case-control study shows 80% of cases drank river water versus 30% of controls. Does this prove the river is the source? Consider confounding, selection bias, and the Village B data.
Read the worked answer
The association is strong, but it does not prove causation. Confounding is possible: people who drink river water may also have lower latrine coverage, lower income, or different occupations, and these factors could independently increase diarrhoea risk. Village of residence is a major confounder, because river water use is clustered in Villages A and C. A stratified analysis within each village is needed to see whether the association persists. Selection bias is also possible if controls are not representative of the source population, and recall bias may occur if cases remember water exposures differently from controls. The Village B data weaken the river hypothesis as a complete explanation: all 12 cases are commuters to Village A, and no local children or elderly people are affected. If the river were the main source, Village B residents who use the protected spring should have little risk, which is consistent, but the commuter pattern suggests workplace exposure in A. The team should investigate food handlers, shared water containers, and sanitation at the workplace, and should test water sources where possible.
3. Given limited chlorine tablets and one health education team, where should the district health office intervene first? Justify using the data and logical reasoning.
Read the worked answer
The first priority should be Village A. It has the highest attack rate (9.0%), the lowest latrine coverage (40%), and reliance on river water and shallow wells. It is also the likely amplification site because all Village B cases are in commuters to A, and no local transmission has occurred in B. Chlorine tablets should be used to treat household drinking water and shallow wells in A, with clear instructions on safe storage. The health education team should focus on A first, covering hand hygiene, safe water, food handling, and latrine use. Village C should be the second priority because it has an intermediate attack rate (5.0%), river water use, and only 55% latrine coverage. Village B needs a targeted investigation of the workplace in A rather than village-wide chlorination, because its local water supply appears protected and its attack rate is lowest. The team should also collect stool samples and water samples to confirm the pathogen and source, but intervention should not wait for laboratory confirmation. This approach uses the data to interrupt transmission at the most likely source while allocating limited resources efficiently.