Population Survey

FoodNet determines the incidence of laboratory-confirmed infections for bacterial pathogens transmitted commonly through food. However, these reports represent only a subset of the true number of cases of diarrheal illness that occur in the community. Most diarrheal illnesses are undiagnosed and, therefore, not reported.

The FoodNet Population Survey is a periodic survey of randomly selected residents in the FoodNet surveillance area. FoodNet conducts this survey to collect information that can be used to estimate how often people in the surveillance area get acute diarrheal illness, how often they seek care for these illnesses, and how often they’re exposed to things linked to diarrheal illness.

Methods

The most recent FoodNet Population Survey was conducted during 2018–2019. This survey used modernized and expanded methods, including multiple modes of administration and multiple sampling frames. It asked participants questions about basic demographics, episodes of diarrhea or vomiting, and various exposures associated with diarrheal illness in the following categories: food consumption, drinking water consumption, raw milk consumption, dietary practices, food handling practices, contact with animals, contact with recreational water, and recent travel history.

FoodNet used two probability sampling approaches for the 2018–2019 FoodNet Population Survey:

  • The first approach used random digit dialing to conduct computer-assisted telephone interviews through landline phones and cell phones.
  • The second approach used address-based sampling to administer surveys online. This involved sampling from residential addresses that are listed in U.S. Postal Service delivery files.
  • FoodNet allocated 30% of the surveys to the random digit dialing and 70% to address-based sampling.

FoodNet interviewed 38,743 adults and children. FoodNet attempted to recruit a child from 80% of selected households and an adult from 20% of selected households. FoodNet chose the child or adult with the next birthday to interview. If a child aged 12 years or younger was selected, a parent or guardian provided information about the child. FoodNet used weights to adjust for disproportionate sampling probabilities and the multi-mode survey design, and to better represent demographic characteristics in the surveillance area.

FoodNet conducted the first five population surveys in 12-month cycles in 1996–1997, 1998–1999, 2000–2001, 2002–2003, and 2006–2007. FoodNet conducted all interviews via landline phones, using sample designs and protocols similar to those used for CDC’s nationwide Behavioral Risk Factor Surveillance Surveys (BRFSS).

FoodNet used random digit dialing and a disproportionate stratified sample method. FoodNet used post-stratification weights to adjust the distribution of the sample by age and sex to match the U.S. Census population.

Access the Data