Source attribution of foodborne pathogens in the Netherlands using structured expert elicitation

Abstract

Foodborne pathogens represent a significant public health burden. Quantifying the relative importance of various potential sources of foodborne infection is challenging due to data scarcity and uncertainty in empirical studies. Structured expert judgment (SEJ) provides a valuable methodological alternative to gain insights into source attribution of foodborne pathogens. We conducted a SEJ study to attribute human cases of 26 foodborne pathogens in the Netherlands to seven major transmission pathways, 20 food groups, and two animal groups, in a typical post-COVID19 year, using Cooke's classical model. The elicitation process involved snowball recruitment, expertise self-assessment, and a workshop where experts answered calibration questions to capture their uncertainty as input for the model. Subsequently, experts completed the target questions to obtain attributable proportions at the ‘kitchen door’ level. Results indicated that transmission was predominantly (>50%) foodborne for Staphylococcus aureus , Listeria monocytogenes , Yersinia spp., Bacillus cereus , Clostridium perfringens , certain non-typhoidal Salmonella serotypes, Campylobacter spp., hepatitis E virus and Toxoplasma gondii , whereas person-to-person transmission was the primary pathway for astrovirus, rotavirus, norovirus, and sapovirus. Brucella spp. and typhoidal Salmonella were attributed primarily (>85%) to international travel. All other pathogens showed attributions of <50% to any individual pathway. Substantial differences were observed when dividing foodborne transmission into food groups. Key contributors included food handlers and vermin, various meats (e.g., pork, beef, chicken), and shellfish. These SEJ-derived estimates complement existing data by providing pathogen-specific insights in the Dutch context.