![[ Visit Client Website ]](images/banner.jpg)
BACKGROUND: Through routine follow-up of a laboratory positive Campylobacter jejunitest result, Maricopa County Department of Public Health (MCDPH) became aware of a potential foodborne outbreak related to a catered event held by a local orthodontist’s office. The event was determined to be an annual party in which the orthodontist practice hosted multiple dentists’ offices. An estimated 89 persons from 19 dental offices attended the event. Of those, an estimated 36 persons became ill, representing an attack rate of 40.5%. The caterer for the event was non-responsive, and MCDPH was only able to acquire rudimentary information on the food provided and attendees present.
METHODS: Due to the lack of information on individual attendees and MCDPH staff constraints, it was unreasonable to identify and contact each person who was at the event. As such an electronic survey which asked questions about potential illness and food items eaten at the event was developed. The electronic survey was distributed to the office manager for each office which attended the event. The office manager was then asked to forward the survey to all of their staff members.
RESULTS: A total of 60 persons (36 confirmed/probable cases and 24 controls) completed the survey, representing a 68.5% response rate. Cross tabulations based on illness status were created for each food item. Odds ratios, 95% confidence interval, and p-values were calculated using MedCalc® software. Analysis of likelihood of illness based on timeframe for eating was conducted using the logistic regression function in SPSS®. Significant food items were lettuce (OR 5.7, 95% CI 1.7-18.5, p-value = .004), chicken enchiladas (OR 3.9, 95% CI 1.3-11.7, p-value = .013), and chicken tacos (OR 3.1, 95% CI 1.03–9.3, p-value = .004). Eating between 11 am- 12 pm was also significantly associated with illness (p-value = .046).
CONCLUSIONS: Electronic surveys can be a good use of limited resources, since they can gather a lot of information while requiring relatively little personnel time. The electronic survey had a 68.5% response rate, compared to the 43.5% response rate for a previous foodborne outbreak investigation utilizing phone calls and mailed letters. However, the user completed surveys had issues with completeness. Additionally, depending on how the survey fields were constructed, it can be cumbersome to extract information for analysis. As such a balance must be struck between constructing a survey that is intuitive for the user to complete while still allowing for reasonable data extraction.