Date:
SWINE RESEARCH SNAPSHOT
BACKGROUND INFORMATION
INDUSTRY IMPLICATIONS:
IMPORTANT FINDINGS:
ARTICLE EXCERPT:
PRIMARY CONTACT:
For further information please contact:
or visit: https://www.uoguelph.ca/osrn/
Summer 2018
Pigs of Instagram? Using network analysis to understand swine disease outbreaks
PED emerged into Canada in
January 2014
Source of the infection was
thought to be feed
Network analysis was used to
describe the outbreak
Networks of case and control
herd contacts are illustrated
Contact patterns identified
between herds and feed
suppliers
The first case of PED was diagnosed in a swine herd in
south-western Ontario. Social network analysis (SNA) was
used to examine the contact patterns between feed suppliers
and herds that were positive and negative for PED during the
early months of the outbreak. SNA was also used to visualize
connections. This research has improved the overall
understanding of the PED outbreak that occurred in
Canadian swine herds in 2014.
NETWORK ANALYSIS FOR DISEASE OUTBREAKS:
A network consists of individuals or groups, referred to as
“nodes” that are connected by links, which are called “edges”
to portray a type of relationship (i.e. feed delivery). The
position of the nodes are important for determining how much
influence a certain node may have. The direct and indirect
connections between these nodes can be described and
illustrated.
A single feed supplier was: 1) linked with the spread of PED
2) directly connected to 8/9 case herds
3) had the greatest number of contacts with case herds
Amanda Perri
amperri@uoguelph.ca
Date:
RESEARCH METHODS:
RESEARCH TEAM & ACKNOWLEDGEMENTS:
Summer 2018
Figure 1: Visual representation of swine herd contacts with feed suppliers using network analysis. The
figure shows that a single feed supplier is connected to 8 of 9 case herds.
A questionnaire was used to gather information on herd contact patterns with feed suppliers between Dec
2013-Feb 2014. A total of 23 Canadian swine herds (n=9 case herds; n=14 control herds) were included
in the study. A case was defined as any swine herd with confirmed positive results for PED virus based
on RT-PCR and with typical clinical signs at the herd level within the study period. Control herds were
randomly selected and matched to cases based on herd size, herd type, and time of PED onset in case
herds. A full network of all herds and separate case and control networks were generated to represent
herd contact patterns with feed suppliers.
Amanda M Perri MSc1; Zvonimir Poljak DVM, MSc, PhD1; Cate Dewey DVM, MSc, PhD1; John CS
Harding DVM, MSc, Dipl. ABVP 2, Terri L O’Sullivan DVM, PhD1.
1Department of Population Medicine, University of Guelph1, Large Animal Clinical Sciences, University
of Saskatchewan2
This work was conducted with generous support by Ontario Ministry of Agriculture, Food, and Rural
Affairs, Ontario Pork, and NSERC- CRD.