# Table 3 Results of the mixed-effects Poisson regression models comparing incidence rates of outbreaks

Mixed effect model Covariate Exposure level IRR 95%CI AIC
(fixed effects model AIC)
1 Categorization of the number of farms by province that had bought the vaccine 1    133.8
(based on modelling of crude incidence rates) (0-4 farms) Referent   (140.1)
2 1.4 0.70, 2.7
(5-10 farms)
3 4.0 1.9, 8.6
(>10 farms)
2 Categorization of the number of farms by province that had bought the vaccine 1
(based on modelling of crude incidence rates) (0-4 farms) Referent   132.9
2 1.3 0.66, 2.6 (136.1)
(5-10 farms)
3 3.6 1.7 , 7.5
(>10 farms)
Categorization of the number of vets or retailers by province who had bought the vaccine 1
(0-1 vet or retailer) Referent
2 1.9 0 .91, 4.1
(>1 vet or retailer)
3 Categorization of the number of farms by province that had bought the vaccine 1    116.3
(based on modelling of standardized incidence rates) (0-4 farms) Referent   (114.3)
2 1.3 0.56, 2.9
(5-10 farms)
3 2.7 1.1, 6.5
(>10 farms)
4 Categorization of the number of farms by province that had bought the vaccine 1    116.0
(based on modelling of standardized incidence rates) (0-4 farms) Referent   (114.0)
2 1.2 0.51, 2.6
(5-10 farms)
3 2.6 1.1, 6.4
(>10 farms)
Categorization of the number of vets or retailers by province who had bought the vaccine 1
(0-1 vet or retailer) Referent
2 1.8 0.71, 4.7
(>1 vet or retailer)
1. AIC = Akaike Information Criterion. IRR = incidence rate ratios (IRR) indicating the increase in the risk of the disease with, respectively, the increasing number of vaccine purchaser farms or the increasing of veterinarians or retailers who had purchased the vaccine, per province. The AIC values in brackets refer to fixed-effect models based on the same covariates.