NURS 832 Week 2 Epi Info 7 Initial Survey Assignment Example

Initial Survey Built in Epi Info 7: A Practice Case Interview Form for Gastrointestinal Illness After a Church Potluck, With the Design Decisions Behind Each Field

Design Memo

Practice scenario. On a Sunday in March, 112 people attended a potluck lunch at a rural church in West Texas. By Tuesday, 41 had reported vomiting or diarrhea to the county health department. This practice survey was built in Epi Info 7 Form Designer to interview every attendee, ill or not, by telephone.

Questions the survey must answer. Who became ill, using a consistent case definition; when illness began, to build an epidemic curve and estimate the incubation period; and which food was associated with illness, by comparing attack rates among people who did and did not eat each dish. These are the core questions of a field investigation of this kind (Rasmussen & Goodman, 2019).

The Survey

Page 1: Identification and Demographics

• RespondentID: text, automatically numbered
• InterviewDate: date
• Age: number, legal range 0 to 105
• Sex: legal values male, female
• AttendedPotluck: yes or no; if no, the form ends

Page 2: Illness and Case Definition

• IllAfterPotluck: yes or no; if no, check code skips to page 3
• Vomiting, Diarrhea, Fever, Cramps, Nausea: yes, no or unknown, one field each
• OnsetDate: date; OnsetTime: time
• DurationHours: number
• SoughtCare: yes or no; Hospitalized: yes or no

Page 3: Food Exposure Grid

• For each of twelve dishes, ate it: yes, no or unknown (ham, chicken spaghetti, deviled eggs, potato salad, green bean casserole, cornbread, coleslaw, banana pudding, peach cobbler, sheet cake, iced tea, punch)
• TimeAte: time
• Leftovers taken home: yes or no

Page 4: Closing

• Others in household ill: yes or no; number ill
• Interviewer initials: text

Design Decisions

First, the form interviews every attendee, not only the ill. Attack rates by food require a denominator: the number who ate each dish and the number who did not. Macha and McDonough (2012) describe attack rates as the basic measure for comparing risk among exposed and unexposed groups in an outbreak, and that comparison is impossible without well respondents.

Second, onset is captured as a date and time in separate typed fields rather than as a description such as "Monday night." Exact times allow an epidemic curve in hours and an estimate of the incubation period from the meal, which helps point toward the type of pathogen.

Third, each symptom and each food is a separate yes, no or unknown field. Combining symptoms in one free-text box would require hand coding, and leaving out "unknown" would force respondents who cannot remember into a false yes or no. A question whose answer cannot be counted is a question the investigation will not be able to use.

Fourth, check code controls the flow. Respondents who did not attend end the interview after page 1, and those who were not ill skip page 2 entirely, which shortens calls and prevents blank symptom records from being misread as missing data. A second check flags any onset time earlier than the meal.

Fifth, Epi Info was chosen because it is free, built by CDC for exactly this kind of field questionnaire and links the form directly to its analysis tools, so the attack rate tables can be produced from the same data without re-entry (Centers for Disease Control and Prevention, n.d.).

Analysis Plan

Once interviews are complete, Epi Info's analysis tools will produce a frequency of symptoms among cases, an epidemic curve from onset date and time, and a two-by-two table for each dish setting illness among eaters against illness among those who skipped it, with a relative risk and confidence interval. The dish with the highest relative risk and the largest share of cases exposed becomes the leading hypothesis for laboratory testing of leftovers.

Testing

I entered three test records, an ill attendee, a well attendee and a respondent who could not remember two dishes, and confirmed that the skips worked, the unknown values were stored as distinct responses and the frequency output listed each dish separately. One problem emerged: the original age field accepted 999, which I corrected with a legal range. The same structure will be adapted for the measles outbreak questionnaire in Module/Week 7.

References

Centers for Disease Control and Prevention. (n.d.). Epi Info. Retrieved October 3, 2026, from https://www.cdc.gov/epiinfo/

Macha, K., & McDonough, J. P. (2012). Epidemiology for advanced nursing practice. Jones & Bartlett Learning.

Rasmussen, S. A., & Goodman, R. A. (Eds.). (2019). The CDC field epidemiology manual. Oxford University Press.