The Effect of Different Recall Periods on Estimates of Acute Gastroenteritis in the United States, FoodNet Population Survey 2006–2007

Abstract

A number of countries have estimated the prevalence of acute gastroenteritis by asking survey respondents to recall past episodes of diarrhea; however, the recall period used varies between studies. We conducted a survey to examine the effects of 7-day and 1-month recall periods on the estimated annual episodes of acute gastroenteritis. Further, we examine whether asking first about illness in the previous 7 days affects a person's response to a 1-month recall period. The Foodborne Diseases Active Surveillance Network (FoodNet) conducted a population-based telephone survey that included asking respondents about the occurrence of gastrointestinal symptoms. From February through April 2007, we randomly split respondents into two groups to examine effect of recall periods and question order. One group was first asked about symptoms in the 7 days before interview and then asked about symptoms in the month before interview. The other group was asked only about symptoms in the month before interview. Overall, the monthly prevalence of acute diarrheal illness (≥3 loose stools in 24-hours, lasting >1 day, or restricting daily activities) was 7.7%. This proportion was consistent among the respondents who were first asked about a 7-day recall period (n=1436) and those asked only about symptoms in the past month (n=2132). Extrapolation from the reported 7-day prevalence of 3.1% to an annual rate of 1.6 episodes per person, however, was almost twice the rate of episodes estimated when extrapolating from the month recall period. Similar findings were found with acute gastroenteritis (acute diarrheal illness or vomiting without respiratory symptoms). First asking respondents about a 7-day recall period did not affect the prevalence of acute gastroenteritis reported for a 1-month recall period. Recall period length did, however, have a major impact on estimates of acute gastroenteritis. Retrospective studies using different recall periods may not be comparable.