Network analyses using case-control data to describe and characterize the initial 2014 incursion of porcine epidemic diarrhea (PED) in Canadian swine herds

Abstract

The overall objective of this study was to describe the contact structure and animal movement patterns of porcine epidemic diarrhea (PED) case herds and matched control herds during the initial incursion of PEDV in Canada, and to evaluate possible mechanisms of spread during this period. Possible mechanisms of spread included transmission through a common-source, herd-to-herd transmission, and transmission due to low biosecurity. Three hypotheses were evaluated by assessing: 1) whether feed supplier, semen supplier and/or animal transportation company networks contained a higher proportion of case herds compared to randomly permuted networks, 2) whether the proportion of case herds in the giant weak component differed from randomly permuted networks, and 3) whether external herd biosecurity, defined as the number of mean contacts with other herds in a one-mode network, was different between case and control herds. The study period for recruiting case and control herds was from January 22, 2014 to March 1, 2014, and a 30-day history of each participating site was collected using a questionnaire. The study included swine herds located in central and eastern Canadian provinces. Multiple two-mode networks with swine herds and service suppliers were constructed. This included feed suppliers, animal movement, animal transportation companies, semen suppliers and a complete network with all service providers.The complete network consisted of 145 nodes. There were a total of 765 edges in the complete network and majority were between feed suppliers and primary herds 29.8% (228/765). The proportion of case herds in the largest feed supplier network was higher than what was expected using randomly permuted networks, suggesting that the likely mechanism of spread during this phase was a common-source through the feed network. A single feed supplier (FS1) had the highest out-degree and outgoing contact chain indicating its importance in disease spread throughout the feed and complete networks. Network descriptive measures, as well as the results of the hypotheses testing indicate little significance in the roles of animal movement, animal transportation companies, and semen suppliers during the initial phase of the 2014 Canadian PED outbreak.