Line lists

Examine the PulseNet Canada line list for the six cases (HTML version: Module 1 – Laboratory line list, Excel version: Module 1 – Laboratory line list).

Note that the cases range in age from 7 to 82 years, two are male and four are female. The PulseNet upload dates (i.e., the dates the PFGE patterns were shared with PulseNet Canada) range from July 14 to 21, 2014; the laboratory received dates (i.e., the dates the samples were received at the local or provincial laboratory for initial testing) range from July 7 to July 15, 2014.

In Canada, laboratory-confirmed E. coli cases are investigated by local public health authorities. This investigation typically involves interviewing cases over the telephone and filling out a questionnaire. These questionnaires usually include questions on the case’s illness, potential risk factors, food history, and travel history. The interview also provides an opportunity to identify people in occupations that should be excluded from work until their illness clears (e.g., food handlers, daycare workers), provide education on risk factors (e.g., hand washing, safe food handling) and to answer any questions the case may have. Key variables from the case interview, such as age, sex, date of illness onset and sometimes risk factors/exposures, are typically shared with provincial/territorial health officials.

Provincial/territorial health officials may also request that local health authorities share the questionnaire if it is suspected that the case is part of an outbreak spanning beyond the local level.

You share the available laboratory information with the affected provinces by email and ask them to share their available exposure information. Here are the responses that you receive:

  • Manitoba (n=2) reports that they do not have the questionnaires for either case from local public health yet.
  • Alberta (n=1) reports that their case is a 35-year old female, with no travel outside the province during her incubation period. The case reports eating hamburgers (well done), grilled chicken, and various salads. Case has a dog. Date of illness onset: July 4, 2014.
  • Saskatchewan (n=3) reports that they began investigating their cluster of cases last week (on Thursday July 17, 2014).
    • Two of their cases attended the same BBQ on Canada Day (July 1), ate hamburgers and various salads at the event, and are not from the same household; their dates of illness onset are July 3, 2014 (52 year-old female) and July 4, 2014 (22 year-old female).
    • The third case (82 year-old male) was hospitalized, has no connection to the BBQ, but also reports eating hamburgers and salads prior to his illness onset (July 1, 2014).

Question 1-3: Based on the information provided, is it reasonable to suspect that a common food item is the source of the illnesses? What additional information would you like to obtain in order to determine whether there is a common food source?

  • 4/4 cases with available food history reported eating hamburgers and salad, making them plausible hypotheses worth investigating further. Both ground beef and leafy greens have been identified, through literature searches, as the source of many E. coli O157:H7 outbreaks in the past.
  • In order to determine if there is a common food source, further details on the food items of interest should be collected (e.g., brand, size, lot numbers, packaging, store it was purchased from). If any of the cases have leftover ground beef/hamburgers or salads, samples should be collected for testing.
  • The Canada Day BBQ in SK may also be worth investigating further. Clusters of illness following events are helpful for narrowing down the source because there is usually a limited number of food items available and multiple illnesses. By comparing what those who became ill ate to those who did not become ill, it is possible to identify which food items may be the source. Questions for SK could include:
    • Are other people sick? Can they do an analytic study (e.g., cohort study, case-control study) to determine the source of infection?
    • Are hamburgers, salads, or any other food items from the BBQ available for laboratory testing?
    • Can additional information about the food served at the event be gathered (e.g., purchase receipts, brand/size/lot codes)
  • Remember that hamburger/ground beef and salads are just two hypotheses of many. There are many other food items that are often eaten at BBQs that could also be plausible sources for the outbreak. At this stage in the investigation there is not enough information available, hence alternative hypotheses should continue to be explored.

Question 1-4: How would you organize this information coming from the provinces? What key data elements would you want to capture?


The information collected during the interview should be entered into a line list.

line list contains the information needed to describe an outbreak in terms of person, place, and time. It is a table that contains key information about each case in an outbreak, with each row representing a case and each column representing a descriptive variable such as demographic, clinical and epidemiologic information.

For this outbreak, variables you should consider adding include:

  • Unique ID number/case identifier
  • Case status (e.g., not a case, suspect, secondary, probable, or confirmed)
  • Demographic information and outcome:
    • Age
    • Sex
    • Geolocator(s) (e.g., province, region, city)
    • Symptoms (one column for each symptom)
    • Date of illness onset
    • Death
    • Hospitalized
  • Laboratory information:
    • Pathogen
    • Serovar/serotype
    • Typing results (e.g., PFGE, genotype)
    • Date of specimen collection
    • Date reported (by laboratory)
  • Exposures: This text field can be used to capture key food exposures of interest as well as information such as lot codes, purchase dates and purchase locations, which are important for guiding food safety investigations. If there are common exposures identified, it is best to capture each different exposure in a separate column.
  • Notes: This text field can be used to capture information which is believed to be relevant but is not captured in any of the above variables.