Step 3: Generate Hypotheses about Outbreak Sources

In a multistate outbreak investigation, a hypothesis is a guess about the source of the illnesses based on what information is known. Hypothesis generation is an ongoing process during an investigation. Early in multistate outbreak investigations, it may not even be clear whether the outbreak is caused by a contaminated food or something else, like contact with animals. Many sources of information help to quickly narrow down how people in the outbreak got sick.

Person, place, and time

Basic information about who is getting sick in a multistate outbreak and where and when the illnesses are happening can help investigators narrow the list of possible sources.

Investigators use a graph called an epidemic curve or epi curve to track the number of illnesses over time. The pattern of the epi curve helps investigators decide if sick people were exposed to the same source over a short period or if the exposure was over a longer time. Investigators also use maps to mark where sick people live so they can easily see if or how the outbreak is spreading. The timing and spread of illnesses can provide important clues about possible causes of the outbreak, including whether a food causing an outbreak has a short or long shelf life and whether a food is distributed regionally or nationally.

Routine enteric illness interviews

Before DNA fingerprinting information is available, state and local health departments already know that a possible foodborne illness has been reported in their state. State and local health departments interview these sick people with standard questionnaires. For many illnesses, like Salmonella, Listeria, and E. coli infections, states collect information on foods and other exposures that caused illnesses or outbreaks in the past. For some germs, the Council of State and Territorial Epidemiologists recommends that state and local health departments ask sick people about certain foods and other exposures.

Hypothesis-generating interviews

When investigators suspect that a contaminated food is causing a multistate outbreak, they consider many different foods as the possible source. State and local public health officials interview people to find out what and where they ate in the days or weeks before getting sick. These interviews are called hypothesis-generating interviews, and they usually are much more detailed than routine enteric illness interviews. The time period investigators ask about depends on how long it takes for a germ to cause illness, which varies.

CDC works closely with local and state public health officials to ask sick people in the outbreak a common set of questions. Depending on what is already known, interviews may focus on specific foods, but most hypothesis-generating interviews ask about many possible food exposures. View sample questionnaires and surveillance tools. Hypothesis-generating interviews typically include questions on:

  • A standardized list of many food items
  • Each meal a person ate before getting sick, including at restaurants
  • Food shopping habits
  • Travel
  • Dietary restrictions
  • Attendance at events where food was served

Prior experience with the outbreak strain

Information from prior experience with the strain causing an outbreak can also help investigators narrow the list of possible causes. Sometimes the germ has caused outbreaks in the past or has been found in foods, animals, the environment, or food facilities. These findings can provide clues about the source of an outbreak, but an outbreak source is only confirmed after other types of information support the link between a food and the illnesses.

Challenges of hypothesis generation

Interviews rely on memory, and it can take several weeks from the time someone gets sick to confirm they are a part of an outbreak. Someone may not remember what they ate many weeks ago. Remembering is harder when the contaminated food is an ingredient such as eggs, spices, or herbs, because people often do not know specific ingredients of the food they ate.

These challenges make it harder for investigators to quickly generate a hypothesis, and investigators may need to re-interview people when a new hypothesis is generated. Home visits can be helpful to look at the foods in a sick person’s pantry and refrigerator. With permission, investigators may also get information from a sick person’s shopper cards and other records of food purchases like receipts.