Wedding-reception diarrhoeal outbreak in a rural district
You are the district medical officer in a rural district of 210,000 people. On Monday morning the local clinic reports that 41 people who attended a wedding reception on Saturday evening have presented with watery diarrhoea and abdominal cramps; two are dehydrated and one has been referred to the regional hospital. There were 78 guests. You obtain the menu and interview all 78 guests by telephone: 40 ate the chicken curry and 34 of them became ill; 38 did not eat the chicken curry and 7 became ill. Of 52 who ate the rice salad, 26 became ill; of the 26 who did not, 15 became ill. The median incubation period is 16 hours (range 10–22 hours) and the median duration of illness is 22 hours. Nobody has fever or bloody stool. The food was cooked in a home kitchen by three family members on Saturday afternoon and left at room temperature until the 8 pm meal.
1. Calculate the attack rates and relative risk for each menu item and state which is the most likely vehicle, with the caveats you would attach.
Read the worked answer
Chicken curry: attack rate in those who ate it 34/40 = 85%; in those who did not 7/38 = 18.4%. Relative risk = 85/18.4 = 4.6, so people who ate chicken were about four to five times more likely to become ill.
Rice salad: attack rate in those who ate it 26/52 = 50%; in those who did not 15/26 = 57.7%. Relative risk = 0.87 - no association, if anything slightly protective.
The chicken curry is therefore the implicated vehicle. Caveats: this is a small cohort with wide confidence intervals; guests who ate chicken may also have eaten other items, so confounding by co-consumption is possible; recall of the menu by telephone is imperfect; and a single dish cannot be declared the cause without food and stool testing. The 85% attack rate among exposed is high but plausible for a heavily contaminated dish.
Also note the overall attack rate of 41/78 = 53%, which is very high for a point-source food-borne outbreak and supports heavy contamination of a single item served to most guests. Ask whether any guests who ate no chicken but ate a sauce or garnish made with the same utensils became ill.
2. Which organism best fits the clinical and epidemiological picture, and how does it differ from the main alternatives?
Read the worked answer
Clostridium perfringens is the best fit: a long, slow-cooked meat dish cooled slowly and held at room temperature allows spores to germinate and multiply; the incubation of 10–22 hours with a median of 16 is typical (usually 8–16 hours); the illness is watery diarrhoea with cramps, little or no vomiting, no fever, and resolves within about 24 hours. It is a toxin-mediated illness, so antibiotics are not indicated.
Alternatives: Staphylococcus aureus and Bacillus cereus (emetic type) cause vomiting with a short incubation of 1–6 hours - not seen here. Bacillus cereus diarrhoeal type resembles C. perfringens but is classically linked to cooked rice, which here shows no association. Salmonella and Campylobacter typically give fever, sometimes bloody stool, and a longer incubation of 12–72 hours. Norovirus gives prominent vomiting, a 12–48 hour incubation and clear person-to-person spread, which would show secondary cases among non-guests. Enterotoxigenic E. coli is possible but less classically linked to reheated stewed meat.
The absence of fever and blood and the short, self-limiting course argue against invasive bacterial pathogens and against Shigella. Laboratory confirmation of C. perfringens requires quantitative culture of stool or food (typically more than 10⁵ organisms per gram) or detection of the enterotoxin, which is often not available.
3. What public health and clinical actions would you take next?
Read the worked answer
Clinical: ensure oral rehydration is available at all health posts, use ORS and zinc where appropriate, and reserve intravenous fluids for severe dehydration. No routine antibiotics. Advise that ill food handlers stay away from food work until 48 hours after symptoms resolve.
Surveillance: notify the regional public health unit, keep a line list with onset times, and plot an epidemic curve to confirm a point source and check for secondary cases that would suggest a transmissible agent.
Investigation: interview the three family members who cooked - ask about recent diarrhoea, skin lesions on hands, and how the chicken was cooled and reheated. Collect stool from current cases (including the hospitalised patient) and any leftover food, with proper containers and cold chain. Inspect the kitchen: holding temperature, time at room temperature, refrigerator space, hand-washing, water source. Take water samples if the water supply is unverified.
Control: advise rapid cooling, refrigeration below 5°C, reheating to at least 74°C, separating raw and cooked food, and avoiding cooking large volumes of meat hours in advance for large gatherings. Run community education through the clinic and the wedding organisers, and write a short outbreak report with your attack-rate table for the district health office.