Use of animal data in public health surveillance for emerging zoonotic diseases

Infectious agents transmitted between animals and humans (zoonoses) are important causes of emerging infectious diseases with major societal, economic, and public health implications. In order to prevent and control emerging zoonotic diseases (EZDs), they should ideally be identified in animals before they affect the human population. The utility of animal data for public health EZD surveillance was investigated in this thesis in four studies: a systematic literature review of current EZD surveillance systems and three critical examinations of pilot agricultural animal health surveillance systems. The first critical examination used expert-elicited criteria of EZD surveillance needs to evaluate a sentinel clinical pre-diagnostic system. The other two studies used statistical modeling to assess the ability of a laboratory-based system and an integrated system with both human and animal data to detect known patterns and outbreaks. The systematic review identified few evaluated surveillance systems, hence an evidence base for successful systems could not be obtained. Experts identified diagnostic data from laboratories and information on potential human exposures as important for public health action. While the sentinel animal surveillance system was not deemed useful on its own, identified gaps and biases in laboratory submissions suggest that sentinel veterinarians could inform animal laboratory surveillance. Seasonal trends and expected events of public health importance were identified in animal diagnostic laboratory data, however, statistical surveillance in either pre-diagnostic or diagnostic data streams did not provide adequate early warning signals for action. While the integrated surveillance for Salmonella bacteria allowed for the examination of the relationship between human and animal data, statistical alerts did not correlate with expert-identified investigations. Laboratory surveillance is likely the best candidate for EZD surveillance in animals, however, this information needs to be supplemented with potential human exposure information, as well as knowledge of data gaps and biases inherent in the data. Without this additional risk information to convert the animal data into risk for humans, the best use of animal laboratory data at this time is to help generate hypotheses in epidemiological investigations and in helping evaluate programs by examining longer-term trends.Medicine, Faculty ofPopulation and Public Health (SPPH), School ofGraduat