Estimates of Food-borne Illness in Canada
The Public Health Agency of Canada estimates that each year roughly one in eight Canadians (or four million people) get sick due to domestically acquired food-borne diseases. This estimate provides the most accurate picture yet of which food-borne bacteria, viruses, and parasites (“pathogens”) are causing the most illnesses in Canada, as well as estimating the number of food-borne illnesses without a known cause.
In general, Canada has a very safe food supply; however, this estimate shows that there is still work to be done to prevent and control food-borne illness in Canada, to focus efforts on pathogens which cause the greatest burden and to better understand food-borne illness without a known cause.
About the estimates of food-borne illness in Canada
The Public Health Agency of Canada estimates that each year roughly one in eight Canadians (or four million people) get sick with a domestically acquired food-borne illness.
The Agency has estimates for two major groups of food-borne illnesses:
- Known food-borne pathogens: There are 30 pathogens known to cause food-borne illness. Many of these pathogens are tracked by public health systems that monitor cases of illness.
- Unspecified agents: Because you can’t “monitor” what is not yet identified, estimates for this group of agents were developed by first looking at the health effects or symptoms that they are most likely to cause—acute gastrointestinal illness (AGI) (i.e. vomiting and diarrhea). Unspecified agents were defined as: agents with insufficient data to estimate agent-specific burden; known agents not yet identified as causing food-borne illness; microbes, chemicals, or other substances known to be in food whose ability to cause illness is unproven; and agents not yet identified.
To estimate the total number of food-borne illnesses, the Agency estimated the number of illnesses caused by both known food-borne pathogens and unspecified agents.
Table 1. Estimated annual number of domestically acquired food-borne illnesses due to 30 known pathogens and unspecified agents transmitted through food in Canada, circa 2006Table 1 – Footnote 1
||Estimated annual number of illnesses
(90% credible interval)
- Table 1 – Footnote 1
- The data used were based on the 2000-2010 time period, and the 2006 Canadian Census was used as a referent population thus the estimates are based circa the year 2006.
|30 known pathogens
||1.6 million (1.2–2.0 million)
||2.4 million (1.8–3.0 million)
||4.0 million (3.1–5.0 million)
Table 2. Top four pathogens causing domestically acquired food-borne illnesses in Canada, circa 2006
||Estimated annual number of illnesses
(90% credible interval)
||1,047,733 (679,576 – 1,434,048)
||176, 963 (95,225 – 270,160)
||145,350 (95,686 – 212,971)
||87,510 (58,832 – 125,525)
The full article entitled “Estimates of the Burden of Food-borne Illness in Canada for 30 Specified Pathogens and Unspecified Agents, circa 2006” is published in the journal Foodborne Pathogens and Diseases.
If you would like a copy of the article in either English or French please contact: enteric.surveillance.entérique@phac-aspc.gc.ca.
Methods and Data Sources
Surveillance and Data Systems
Many surveillance systems are used in Canada to provide information about the occurrence of food-borne illness. Most of the Agency’s surveillance systems rely on data from provincial/territorial and local public health ministries/units/authorities. Systems focus on specific pathogens likely to be transmitted through food to detect outbreaks, monitor trends and risk factors.
Each surveillance system plays a role in detecting and preventing food-borne illness and outbreaks.
Table 3. Surveillance systems used in developing Canadian Estimates for Food-borne IllnessTable 3 – Footnote 1
||Timeframe of data used
- Table 3 – Footnote 1
- Though PulseNet Canada and the Canadian Integrated Program for Antimicrobial Resistance Surveillance (CIPARS) were not used in developing these estimates they are other important surveillance systems for food-borne illness in Canada.
|Canadian Notifiable Disease Surveillance System (CNDSS)
||Collects the number of laboratory confirmed illnesses reported to local public health units/regions to provincial public health authorities and to the national level on an annual basis.
|National Enteric Surveillance Program (NESP)
||Collects detailed case level data from invasive listeriosis cases in participating provinces.
|Enhanced National Listeriosis Surveillance
||Collects aggregate counts of laboratory isolates of select enteric pathogens (species and subtype) reported through the provincial laboratories on a weekly basis.
|Provincial Reportable Disease Surveillance System
||Collects the number of laboratory confirmed illnesses reported from local public health units/regions to provincial public health authorities only.
|National Studies on Acute Gastrointestinal Illness (NSAGI) Population Surveys
||Population surveys asking Canadians about vomiting and diarrhea.
||Ontario, British Columbia
||2001-2002, 2002-2003, and 2005-2006
||An integrated enteric pathogen surveillance system based on a sentinel site surveillance model that collects information on both cases of infectious gastrointestinal illness and sources of exposure within defined communities. This system provides detailed case information on clinical illness and risk factors.
||Waterloo Region, Ontario
Under-reporting and under-diagnosis
In general, to be captured in a Canadian surveillance system a sick individual must: seek care; have a sample (stool, urine or blood) requested; and submit a sample for testing. In addition, the sample must be tested with a test capable of identifying the causative agent; and finally the positive test result must be reported to the surveillance system (Figure 1). Surveillance systems only capture a small portion of total illnesses given all these necessary steps (i.e. there is under-diagnosis and under-reporting taking place).
Figure 1 Burden of illness pyramid
Text Equivalent – Figure 1
Estimating Canadian food-borne illnesses for 30 known food-borne pathogens
Two main methods were used to estimate the number of Canadian food-borne illnesses for the Listing of 30 known pathogens.
The first approach:
For each pathogen with surveillance data, we used data from various surveillance systems and corrected for under-reporting and under-diagnosis. We then multiplied the adjusted number by the proportion of illnesses acquired in Canada (that is, not acquired during international travel) and the proportion transmitted by food, to estimate the number of illnesses that are domestically acquired and food-borne (Figure 2).
The second approach:
For common pathogens that are not part of standard surveillance, we estimated the number of Canadians who would experience symptoms (e.g. diarrheal illness) and the proportion of those symptoms that is related to the particular pathogen. We then multiplied this number by the proportion of illnesses acquired in Canada and the proportion transmitted by food, to yield an estimated number of illnesses that are domestically acquired and food-borne (Figure 3).
Alternative approaches were used to estimate illnesses where suitable data from surveillance or data on proportion of symptoms attributed to the pathogen were not available.
Then, the estimates for each of the pathogens were added together to arrive at an overall pathogen specific total. An uncertainty model to generate a point estimate and 90% credible interval (i.e. upper and lower limits that account for variability and uncertainty of the data) was used.
Figure 2Figures 2 and 3 – Footnote * First approach: For pathogens where laboratory-confirmed cases were scaled up
Text Equivalent – Figure 2
Figure 3Figures 2 and 3 – Footnote * Second approach: For pathogens where Canadian population was scaled down
Text Equivalent – Figure 3
- Footnote *
- Probability distributions were used to model uncertainty in each data inputs. Point estimates were bounded by a 90% credible interval.
Listing of 30 known pathogens by estimation method
Pathogens for which laboratory-confirmed illnesses were scaled up
National reportable disease data
- Brucella spp.
- Campylobacter spp.
- Clostridium botulinum
- Cryptosporidium spp.
- Cyclospora cayetanensis
- VTEC O157
- Giardia sp.
- Hepatitis A
- Salmonella spp., nontyphoidal
- Salmonella Typhi
- Shigella spp
- Vibrio cholera
- Vibrio spp., other
- Vibrio vulnificus
Provincial reportable disease data
- Trichinella spp.
- Listeria monocytogenes
- Vibrio parahaemolyticus
- Yersinia enterocolitica
Pathogens for which Canadian population scaled down
- Toxoplasma gondii
- Clostridium perfringens
- E. coli, other diarrheagenic
- VTEC non-O157
- Bacillus cereus
- Staphylococcus aureus
Estimating Canadian food-borne illnesses for unspecified agents
Unspecified agents that cause acute gastrointestinal illness fall into four general categories:
- Agents with insufficient data to estimate agent-specific burden
- Known agents not yet recognized as causing food-borne illness
- Microbes, chemicals, or other substances known to be in food that could at some time be shown to cause illness Agents not yet described
- Agents not yet described
To estimate food-borne illnesses from unspecified agents, we used symptom-based data from surveys to estimate the total number of episodes of acute gastrointestinal illnesses (AGI) and then subtracted the number of illnesses accounted for by known AGI pathogens. We then multiplied this number by the proportion of domestically acquired illnesses and of illnesses attributable to food, just as we did for the known agents. Finally, as with the known-pathogens estimate, we used an uncertainty model to generate a point estimate and 90% credible interval (upper and lower limits) (Figure 4).
Food-borne illnesses due to chemicals that cause acute gastrointestinal illness are included in the estimate of illnesses due to unspecified agents. However, chemicals or unspecified agents that do not cause acute gastrointestinal illness are not included in the estimates.
Figure 4Figures 4 – Footnote * Approach for unspecified agents
Text Equivalent – Figure 4
- Footnote *
- Probability distributions were used to model uncertainty in each data input. Point estimates were bounded by a 90% credible interval.
- Footnote **
- Estimated proportions were based on 25 known pathogens that cause acute gastrointestinal illness. Five pathogens were not included because their primary symptoms are not acute gastrointestinal illness.
Improvements to previous estimates
The Agency’s 2013 estimates of illnesses from food-borne diseases in Canada are more accurate than the estimates published in 2008 of 11 million episodes of food-borne illness each year based on better data and methodologies. The 2008 estimates used values from earlier United States Centers for Disease Control and Prevention estimates applied to a Canadian estimate of the average number of episodes of acute gastrointestinal illness per person occurring each year. In addition, the methodology used for the 2013 estimates is different from that used in 2008. As a result of these differences, no strict side-by-side comparison can be made between the two sets of estimates. The 2013 estimates do not mean that there is less food-borne illness occurring, but rather, that more accurate estimates are now possible.
The 2013 estimates of illnesses from food-borne disease in Canada reflect improvements in methodology since 2008. Perhaps most importantly, these new estimates identify and rank the most important bacteria, viruses and parasites (“pathogens”) responsible for causing food-borne illness. These more specific estimates can further inform policy and regulatory priorities to prevent future illnesses.
The following list highlights the major differences in data and methodology between the new estimates and those published in 2008, and how they affect the estimates of illnesses from food-borne diseases in Canada.
Differences between 2008 and 2013 methodology
- Included international travel-related illnesses.
- Did not estimate illness for individual pathogens.
- Utilised 1.3 episodes per person/year based on the following AGI case definition: any diarrhea or vomiting in the past 28 days excluding those with chronic conditions.
- 36% = Proportion applied to rate of acute gastrointestinal illness (known pathogens and the unspecified agents included) estimated to be food-borne (based on the US 1999 estimates).
- Uncertainty and variability of each input was not calculated.
- Excluded international travel-related illnesses.
- Estimates of illness for 30 known pathogens.
- Pathogen-specific multipliers used to adjust for under-reporting and under-diagnosis.
- Pathogen specific proportion domestically acquired and food-borne applied.
- Utilised 0.63 episodes per person/year based on the following AGI case definition: 3 or more loose stools in 24 hours or any vomiting in the past 28 days excluding those with chronic conditions, or concurrent symptoms of coughing, sneezing, sore throat or runny nose.
- 20% = Proportion of the unspecified agents estimated to be food-borne (based on pathogen specific information on proportion food-borne).
- The Agency used many data sources, with varying degrees of reliability, to determine the estimates of food-borne illnesses. For each estimate, a formula was used to account for the cumulative effect of uncertainty and variability of the data inputs.
Effects of Difference
- 2013 estimate focused on food-borne illnesses acquired in Canada: 2013 estimates were limited to food-borne illnesses that were acquired in Canada, which reduced the number of food-borne illnesses in 2013 vs. 2008.
- 2013 estimated number of illnesses caused by known pathogens: more accurate: Utilising specific multipliers and proportion domestically acquired and food-borne for the 30 known pathogens yielded more accurate estimates for each known pathogen and, ultimately, greater accuracy in the overall estimate of food-borne illness.
- 2013 estimate of acute gastrointestinal illnesses (AGI): more precise: A more specific case definition for AGI was used to be more precise in the estimate and to minimize the chance of estimating illness that was not infectious (i.e. related to chronic illnesses such as Crohn’s disease) and not truly gastrointestinal (i.e. symptoms related to a respiratory infection). The impact of this is a lower overall estimate in 2013 vs. 2008.
- 2013 estimate used a smaller proportion of unspecified AGI determined to be food-borne: Reduced the number of food-borne illnesses in 2013 vs. 2008.
- 2013 estimate accounted for uncertainty: The results were upper and lower 90% credible limits, (i.e. a 90% credible interval). This means that 90% of the time the true value of the estimate falls within the upper and lower values.
Although we cannot compare these estimates to determine trends, we can turn to other data sources for information about trends in some important infections that are transmitted commonly through food.
Data from the Canadian Notifiable Disease Surveillance System (CNDSS) and National Enteric Surveillance Program (NESP) provide the best measures of disease trends. Although these systems include only a portion of the pathogens that make up the estimates, it does allow us to see changes over time for these important food-borne pathogens.
According to these systems some food-borne illnesses have dropped substantially over the past decade, but infections caused by one of the most common pathogens – Salmonella have not declined.
Trends in food-borne illness for 2011 compared to the 1998-2000 baseline period:
- No significant change in the rate of Salmonella infection (NESP).
- 35% decrease (95%CI 33-36%) in the rate of campylobacteriosis (CNDSS)
- 68% decrease (95%CI 65-71%) in the number of O157 Verotoxigenic Escherichia coli (VTEC) infections (NESP)
- 27% decrease (95%CI 22-32%) in the rate of shigellosis (CNDSS)
Other important pathogens commonly transmitted through food (e.g. norovirus, Clostridium perfringens, Toxoplasma gondii) are not tracked in part because they cause mild symptoms of short duration and because of current limitations in laboratory capacity and techniques. Common prevention measures (e.g. safe food handling) that would decrease illness caused by tracked pathogens would also decrease illness caused by pathogens not currently being tracked.
Figure 5 Relative rates of laboratory-confirmed infections with Campylobacter, VTEC O157, Salmonella, and Shigella compared with 1998–2000 rates, by year, 2001–2011, CNDSS and NESP
Text Equivalent – Figure 5
Comparison to US methodology and results
Canada used similar methodologies as the United States Centre for Disease Control and Prevention (US-CDC) for estimating the burden of food-borne illness in their country.
The overall total estimate (specified and unspecified agents) for Canada is slightly less than the US-CDC estimate with approximately one in eight Canadians compared to the US-CDC estimate of one in six Americans, experiencing food-borne illness per year.
Top 4 Pathogens contributing to total domestic food-borne illness in:
- Clostridium perfringens
- Campylobacter spp
- Salmonella spp., non-typhoidal
The United States
- Salmonella spp., non-typhoidal
- Clostridium perfringens
- Campylobacter spp.
- Inclusion of adenovirus and exclusion of Mycobacterium bovis and Streptococcus Group A (pathogens excluded are not relevant to Canada’s domestic food supply).
- Incorporated duration of illness and bloody diarrhea to define severe cases for some pathogens.
- Ratio of Bacillus cereus and Staphylococcus aureus to Clostridium perfringens using reported provincial data was applied to the estimate obtained through the population incidence of Clostridium perfringens from UK study.
- Estimated rotavirus, astrovirus and sapovirus for the total population.
- Included illnesses due to viruses for the total population as part of specified pathogen estimate.
The United States
- Did not include adenovirus but did include Mycobacterium bovis and Streptococcus Group A.
- Included only bloody diarrhea in definition of severe cases for some pathogens.
- Used data from outbreak reporting system and then applied an outbreak to sporadic ratio for Bacillus cereus, Staphylococcus aureus and Clostridium perfringens.
- Estimated rotavirus, astrovirus and sapovirus for < 5 years of age only.
- Illnesses related to viruses for those greater than 5 years of age are included in the unspecified agents’ estimate.
Effects of Methods
- Pathogens causing food-borne illness included in the estimate: Minimal impact to differences in overall estimates.
- Canadian definition of severity of illness included duration: Results in a higher proportion of cases being considered severe, and ultimately in a lower estimate of under-diagnosis in Canada compared to the United States.
- Approach to estimating bacterial food-borne toxins: Different approaches therefore difficult to compare the effect.
- Viruses calculated for total population: Reduced the number of cases in the unspecified portion as these were now part of the number of viruses estimated within the specified pathogens portion for Canada compared to the United States. Results in a lower total food-borne estimate for Canada compared to the US as these viruses have a low proportion food-borne.
- Unspecified agents: Reduced the number of cases estimated in the unspecified portion (60% of total) compared to the US (80% of total). Also reduces proportion food-borne (20%) compared to US (25%) applied to AGI caused by unspecified agents