Wedding reception, 210 guests, 76 ill: identifying the vehicle when the laboratory may be silent
You are the MOH called on Monday morning about a wedding reception held on Saturday evening in a rural GN division. About 210 guests attended; 76 have since reported vomiting, and 14 were treated at the district hospital, two of whom were admitted overnight. None is critically ill. The reported incubation period is 2–7 hours, median 4 hours, with vomiting in about 90% and diarrhoea in about 40%; only a few had fever. The caterer prepared food in a home kitchen from mid-afternoon and transported it without refrigeration. A quick line-list shows that 108 guests ate the chicken curry, of whom 68 became ill; 102 did not eat the chicken curry, of whom 8 became ill. Almost everyone ate rice: 74 of the 200 rice eaters were ill, compared with 2 of the 10 who did not eat rice.
1. Calculate attack rates and relative risks for the chicken curry and for rice, and state whether the epidemiological evidence is sufficient to implicate the chicken.
Read the worked answer
Chicken curry.
- Attack rate, exposed: 68/108 = 63.0%
- Attack rate, unexposed: 8/102 = 7.8%
- Relative risk: 63.0/7.8 = 8.0
- Risk difference: 55.1 percentage points
- Attributable fraction in the exposed: (63.0 − 7.8)/63.0 = 87.6% - about seven of every eight illnesses among chicken eaters are attributable to the chicken.
Rice.
- Attack rate, exposed: 74/200 = 37.0%; unexposed: 2/10 = 20.0%; RR = 1.85.
- But only ten guests did not eat rice, and two of them were ill. The confidence interval around this estimate is very wide, and near-universal exposure means rice cannot discriminate between guests. Rice is not implicated.
Is the evidence sufficient? Yes, on epidemiological grounds alone. The association is strong (RR 8), dose-consistent, biologically coherent with a 2–7-hour incubation and vomiting-predominant illness without fever, and consistent with a preformed toxin in a protein-rich food left unrefrigerated. The short incubation and the absence of fever point to Staphylococcus aureus or Bacillus cereus (emetic type) rather than a salmonella-type infection. A negative laboratory result would not overturn this: preformed toxin survives reheating, so the organism may be dead while the toxin remains, and cultures taken on Monday may be sterile. Epidemiology should drive the control measures; the laboratory confirms, it does not veto.
2. Set out your specimen collection and investigation plan, with the limits of each test.
Read the worked answer
Human specimens. Stool or rectal swabs from at least 10–15 currently ill guests, including the two admitted patients, transported in Cary-Blair; vomitus where available. Blood cultures only for the admitted patients with fever or systemic signs. Collect early - yield falls sharply after 48–72 hours.
Food and environment. Retrieve any remaining chicken curry, other leftovers, and the original containers; refrigerate immediately and send for quantitative culture - S. aureus at ≥10⁵ CFU/g is supportive. Test for staphylococcal enterotoxins (ELISA or PCR for sea–see) and for B. cereus emetic toxin. Sample the kitchen surfaces, utensils and water supply.
Food handlers. Nasal swabs and hand/finger swabs, plus inspection for skin lesions on hands and forearms. A carrier is only meaningful if the phage type or molecular profile matches the food and patient isolates - otherwise the finding is a red herring that can unfairly end someone's livelihood.
Limits to state in the report. Toxin is heat-stable but the organism is not, so a negative culture does not exclude toxin-mediated illness. Late collection reduces yield for all pathogens. A single food item may have been cross-contaminated, so 'the chicken' may mean 'the chicken plus the knife'. Results should be interpreted alongside the line-list, not in place of it.
3. What control measures and communication would you put in place, and what will you say about the likely organism?
Read the worked answer
Immediate control. No further food from the same kitchen until it has been inspected and cleaned; exclude symptomatic food handlers from food work for 48 hours after symptoms resolve; reinforce hand hygiene, separate raw and cooked items, and - the decisive failure here - keep protein-rich food either above 60 °C or below 5 °C, and cool it rapidly if it must be stored. Home catering for large events needs a basic food-safety briefing and a checklist at the time of booking, because this is a recurrent pattern in this setting.
Case finding and care. Active case finding among guests, a simple case definition, oral rehydration advice, and clear criteria for attending hospital (blood in stool, high fever, dehydration, pregnancy, young children, elderly). Two of the guests were admitted; confirm their clinical course and ensure no secondary household spread.
Communication. Tell the community, factually, that the pattern - vomiting within a few hours of a meal, no fever - fits a toxin produced by bacteria in food that was kept warm for several hours, that it is not a personal failing of the family, and that the illness is usually short-lived with rehydration. Avoid naming the caterer publicly or speculating on blame before the investigation concludes. Advise attendees to seek care if symptoms persist beyond 24–48 hours or if there is blood or high fever, which would suggest a different pathogen. Report to the food safety authority and the regional epidemiologist, and write the outbreak report with the line-list, the analysis and the recommendations, so the next event can be prevented.