A national produce supply chain database for food safety risk analysis

During a foodborne crisis, risk assessors are often scrambling to assemble data needed to trace suspected foods along very complex supply chains. Although traceability systems ensure that stakeholders in the supply chain record lot-specific trace-back and trace-forward data, there are few databases available that describe in detail the flow of product in the complex web of supply chains. This paper presents the methodological approach used to design and assemble a relational database of nation-wide trade data for packaged ready-to-eat lettuce and leafy greens. The database was used in the development of an integrated simulation tool (Canadian GIS-based Risk Assessment, Simulation and Planning for food safety tool, i.e. CanGRASP) that can predict the spatial distribution and public health risk associated with contaminated food. The database includes the geographical coordinates of 5 domestic processors, 28 produce distribution centres and 2946 retail outlets from five of the top ten retail chains in Canada. It also includes other critical information to predict the fate of pathogens during distribution of contaminated product through the supply chain including: (a) product volumes handled by each stakeholder, (b) flow of product between stakeholders, (c) temperatures of product each season, and (d) times products spend in each step or during transit between steps, for each season. The database is used by both the simulation and mapping components of the integrated simulation tool during risk assessment exercises associated with emergency preparedness planning and training. Using the database, CanGRASP was able to assess the spread of the population at risk during a simulation of a hypothetical outbreak caused by fresh-cut leafy vegetables contaminated with Escherichia coli O157:H7 in the Canadian food distribution systems during both summer and winter seasons.