Bayesian Modeling of Within-Herd Transmission Dynamics of Swine Influenza Virus

Thumbnail Image

Persistent link to this item

View Statistics

Journal Title

Journal ISSN

Volume Title


Bayesian Modeling of Within-Herd Transmission Dynamics of Swine Influenza Virus

Published Date




Thesis or Dissertation


Influenza A virus (IAV), also known as swine influenza virus, commonly circulates in swine populations. IAV infection is a concern to swine producers, veterinarians and the general public, and is considered one of the top three respiratory diseases in terms of frequency of appearance in North American and the cause of an economically important respiratory disease in growing pig populations. On-farm assessment of health on an individual level for IAV infection involves the use of respiratory clinical signs (RCS) and behavioral observations by making inference from RCS to its causes using inductive reasoning as required for a Bayesian approach (BA). Therefore, the aims of this dissertation were to create Bayesian epidemiologic models using inductive reasoning with inverse probability, and to describe and better understand the within-herd transmission of IAV in wean-to-finish pig populations. A Bayesian approach is commonly used in veterinary medicine as it has been part of inductive reasoning regarding interventions, treatments and diagnoses. When veterinarians are managing patients or on-farm assessment of health, veterinarians start with their inferences from history and clinical signs to an underlying cause using inductive reasoning. The diagnostic test accuracies of RCS were 0.38 (95% Credible interval (CrI): 0.28-0.48 for Se) and 0.66 (95%CrI: 0.61-0.71 for Sp). The true IAV prevalence was estimated to be 0.24 (95%CrI: 0.16-0.30) and vaccination reduced such level of prevalence. By accounting for imperfect diagnostic test of RCS, the transmission rate was 1.40 day-1 (95% CrI: 0.42-5.52) and the reproductive number was 4.19 (95%CrI: 1.98-26.29). Waning rate of maternal derive antibodies (MDA) rate for IAV H1N1 was estimated to be 0.016 day-1 (95%CI: 0.013, 0.019) and time to lack of MDA maternal immunity was 64.09 days (95%CI 60.77- 77.40). An epidemic can occur at any point in time during a wean-to-finish period with more than one epidemic peak with low MDA population and the sufficiency of initially infected pigs. IAV transmissibility was elucidated as strongly periodic (p-value < 0.001) with peak timing in mid-June. The absolute IAV intensity was 0.18. The relative IAV intensity was 2.41, implying that IAV disease intensity at the periodic peak was 2.41 times higher compared to that at the periodic nadir. In conclusion, for a swine herd health management perspective, the use of RCS is able to potentially be used as on-farm assessment and measured for IAV transmissibility by inductive reasoning. Heterogeneity of MDA in wean-to-finish pig populations plays a crucial role in enhancing IAV transmission and waning MDA has interfered with vaccination to create more subclinical infections. Vaccination may be able to reduce the true IAV prevalence and has broader implications for the control and perhaps eradication of IAV. IAV transmissibility was elucidated as periodic. Better understanding waning of MDA, periodic IAV transmissibility, persistence and dynamics should result in a better design of the optimal control strategies and periodic IAV vaccination in growing pig populations.


University of Minnesota Ph.D. dissertation. January 2016. Major: Veterinary Medicine. Advisor: John Deen. 1 computer file (PDF); xii, 215 pages.

Related to




Series/Report Number

Funding information

Isbn identifier

Doi identifier

Previously Published Citation

Suggested citation

Homwong, Nitipong. (2016). Bayesian Modeling of Within-Herd Transmission Dynamics of Swine Influenza Virus. Retrieved from the University Digital Conservancy,

Content distributed via the University Digital Conservancy may be subject to additional license and use restrictions applied by the depositor. By using these files, users agree to the Terms of Use. Materials in the UDC may contain content that is disturbing and/or harmful. For more information, please see our statement on harmful content in digital repositories.