Abstract
The association between herd size and sero-prevalence of Salmonella was assessed in a random-effects model with herd size, county and date of slaughter as fixed effects. A total of 510915 meat-juice samples from 14593 herds located in 13 counties in Denmark was included in the study. A random-effects model was developed from separate models for smaller strata of data from herds with approximately equal sizes. The combined model was analysed and the results reported. Herd size was positively associated with the sere-prevalence of Salmonella enterica, but the size of the association was biologically of little importance, because the within-herd and the between-herd variations were relatively large in comparison. The relative magnitudes of the variance components indicated that factors associated with both the herd level and the pig level could be important in the prediction of seroprevalence of S. enterica.
Original language | English |
---|---|
Journal | Preventive Veterinary Medicine |
Volume | 34 |
Issue number | 2-3 |
Pages (from-to) | 191-203 |
ISSN | 0167-5877 |
Publication status | Published - 1998 |
Keywords
- pig microbiological diseases
- herd size
- Salmonella enterica
- random effects
- modelling