Source Attribution

Definition and Purpose

Human illness attribution (source attribution) of infectious enteric diseases is one of the key scientific objectives of the FoodNet Canada system (formerly known as C-EnterNet).

Infectious enteric disease can be caused by any of a wide variety of germs: viral, bacterial, or parasitic. People get sick from these germs through contaminated food or water or by contact with other sick people or animals.

Food-borne source attribution is the partitioning of the human disease burden of one or more food-borne infections to specific sources, where the term “source” includes animal reservoirs and vehicles (e.g., foods) (Pires et al., Food-borne Pathogens and Disease, 2009, 6(4): 417-424).

FoodNet Canada uses a broader and more specific definition: The partitioning of the human infectious gastrointestinal illness to specific sources, where the term “source” includes animal reservoirs and vehicles (i.e., encompassing all food-borne, water-borne, animal-to-person, and person-to-person transmission routes relevant to the illness epidemiology of gastrointestinal illness).

In source attribution, the “source” includes both the reservoir of the pathogen or the vehicle. The reservoirs are hosts from which the pathogens originate, which can include diseased and asymptomatic carriers/shedders among human beings, and ill and carrier animals (food animals, pets, wildlife; the former being often referred to as the point-of-production source). The vehicles facilitate contact between pathogen reservoirs and people. These vehicles can include water, food, animals, soil, manure, compost, and other persons.

Source attribution for gastrointestinal illness helps us look at each of the many combinations of factors that can cause illness and, for each, estimate:

• the number of human cases,
• hospitalizations,
• deaths, and
• the related costs for each of the above.

As a result, source attribution has become an important tool to better inform food and water safety policy development and evaluation. Ultimately, source attribution helps to decrease the burden of enteric disease in Canada. It also helps us evaluate whether new policies and interventions on the farm, in the environment, or in food production are effectively decreasing disease in Canada.

Methodological approaches

Several methods are available that can be used to estimate the source and burden for the human illnesses caused by enteric diseases (Pires et al., 2009). Each method has specific strengths and limitations. Experts on source attribution have concluded that none of the currently available methods are able to give an accurate estimate for source attribution on their own. These approaches are quite different in concept and in scope and address slightly different questions; thus their results are considered more complementary than comparable.

• Microbial subtyping comparison: This approach involves describing each potential source with phenotypic or genotypic microbial characteristics, while comparing them to the characteristics of the pathogens that infected the human cases. The more source-specific the microbial characteristics used are, the easier it is to point out the source of contamination with the highest confidence.
• Comparative Exposure Assessment: Comparative exposure assessment determines the relative importance of the known transmission routes by estimating the human exposure to that pathogen via each route. The approach requires, for each known transmission route, information on the prevalence and dose of the pathogen(s) in the various sources, the changes in prevalence and levels of the pathogen(s) through the transmission chain, and the frequency by which humans are exposed to each route.
• Outbreak data analysis: Once an outbreak is detected, an investigation is initiated to identify the causative agent and source pathogen and the reason of the outbreak (what the vehicle implicated is and why) so that further cases can be prevented. The implicated vehicle can be detected through microbial tests or epidemiological association. Outbreak investigations are summarized in reports and databases. Outbreak summaries can be used to quantify the association between the pathogens and the sources.
• Case control study: This approach can be defined as an analytical epidemiologic study design in which individuals affected by the disease under study (also called cases) are compared to asymptomatic individuals (also called controls) regarding past exposures. The relative role of exposures is estimated by comparing the frequency of exposures among cases and controls. When infections are associated with an exposure, the proportion of cases attributed to that exposure can be calculated. Epidemiologists define this number as the population attributable fraction (PAF).
• Intervention study: This approach compares the incidence rates before and after an intervention has taken place. Interventions are actions that modify the exposure of humans to a given pathogen through a specific reservoir or vehicle. The intervention can be designed or it can be the result of natural occurrence (driven by a food safety issue) such as withdrawal of all poultry production following a dioxin exposure.
• Expert elicitation: Experts are asked to provide their estimates (opinions) on the proportion of cases attributed to each main transmission routes or to each food category. Their opinions are synthesized to provide a unique average estimate or a range of estimates on most likely sources and routes (mean value and its confidence or credible interval).

Evidence Generated by FoodNet Canada

Since 2005,FoodNet Canada researched and developed several products using various source attribution methods in Canada. These include:

• An outbreak data analysis from Canada and from all over the world;
• A cryptosporidiosis QMRA (2007-2008);
• A case-case comparison of cryptosporidiosis cases;
• An expert elicitation survey among Canadian food safety experts in 2008-2010;
• An analysis of the human case data from its sentinel site surveillance to estimate the proportion of international travel-related cases of gastrointestinal illness; and
• A healthy control survey (n=1200) implemented to better understand disease by comparing to the human enteric cases in Sentinel Site 1 to the healthy population.

Finally, because all approaches yield results that are more complementary than comparable, efforts are being made to synthesize the results from the various studies undertaken in Canada.

Improved source attribution will help determine, with greater certainty, the most important sources and pathways of each pathogen that cause infectious enteric disease. This information will help focus resources allocated to food and water safety and communicable disease prevention in the most cost-effective manner, in order to reduce the burden of enteric disease in Canada. It also will provide Canada with the ability to evaluate the effectiveness of interventions at the farm, processing, and environment levels to improve food and water safety and public health.