Untangling the complex inter-relationships between horse managers’ perceptions of effectiveness of biosecurity practices using Bayesian graphical modelling.

This study explores the perceptions of horse managers in Australia regarding the effectiveness of various on-farm biosecurity practices, specifically using a Bayesian network modelling approach. The researchers found that different biosecurity measures, such as movement restrictions and hygiene practices, are perceived differently by horse managers, suggesting that education strategies may need to be tailored accordingly.

Study Design and Methodology

• The researchers utilized a dataset gathered from interviews with 200 horse managers in Australia post the 2007 equine influenza outbreak. The data collected focused on the horse managers’ perception of the effectiveness of various on-farm biosecurity practices.
• To analyze the multiple variables and potential interrelationships, the researchers employed a Bayesian network modelling approach. This approach does not require prior knowledge of relationships which makes it suitable for analysing such multivariate data.
• Computationally intensive methods were leveraged to identify an optimal graphical statistical model directly from the observed data, aligning this methodology to multivariate regression.
• The identified optimal model structure was first established via an exact (exhaustive) search algorithm, and then pruned for overfitting using a parametric bootstrapping approach.

Major Findings

• Perceptions about the effectiveness of movement restrictions and access control were generally linked but were separate from perceptions about hygiene practices for personal equipment.
• Horse managers who perceived the effectiveness of movement restrictions and reducing personal contact with horses also saw value in curtailing professional visits.
• Measures such as disinfecting vehicles, using a disinfectant footbath, changing into clean clothes upon arrival at premises, and washing hands before contact with horses were perceived as effective grouped together.
• Interestingly, those managers who believed in the effectiveness of disinfecting vehicles (a hygiene measure) were less inclined to believe in the effectiveness of controlling access to horses (an access control measure).

Implications

• The findings provide valuable insights into the various perceptions surrounding the effectiveness of different biosecurity practices in relation to the horse industry.
• The differential perceptions of hygiene measures and access control measures may call for tailored extension education strategies aimed at horse managers who hold stronger beliefs in the effectiveness of one approach over the other.