Assessment of biosecurity practices and development of a scoring system in swine farms using item response theory

Brazil is the fourth largest producer and exporter of pork in the world, but has never reported yet the occurrence of some economically important diseases such as porcine reproductive and respiratory syndrome (PRRS) and porcine epidemic diarrhea (PED). Most of the swine farms in Brazil are characterized by intensive production being part of large integrated companies, where biosecurity practices help to prevent the introduction and spread of disease-causing infectious agents. The assessment of biosecurity in farms is not straightforward because of the large number of practices that constitute an on-farm biosecurity program. It is therefore necessary to combine the measurement of several parameters in order to characterize the level of biosecurity on a given farm. Thus, the objective of the study was to develop a biosecurity score to estimate the biosecurity level (theta or θ) in swine farms using the item response theory (IRT) and explore the relationship between the scores and independent variables. The IRT is a latent trait method extensively used in other fields, and offers the advantage to quantify the latent trait, here the biosecurity level, and to identify the practices that discriminate the farms avoiding the use of extensive questionnaires and redundant questions. In this study, 604 farms were evaluated in the main swine production regions of the state of Rio Grande do Sul, Brazil. Thirty-five practices were considered in order to quantify the biosecurity level on a given farm. After a recursive process 14 practices were selected to compose the biosecurity score to estimate the biosecurity level (θ). The variables identified with greater capacity of distinguishing the farms as to their biosecurity level were if the farm has 'feed bin outside of the barn limit (external feed loading)', has 'perimetral fence around the farm or barn', and if 'transit of trucks inside the farm is prohibited'. The biosecurity level was associated with some independent variables, e.g. the farm operation type, the integrated company and some owner characteristics. In addition, the results demonstrated that biosecurity practices related to management (internal biosecurity) are adopted with higher frequency compared to segregation and sanitation practices (external biosecurity). The IRT model proved useful and valid to estimate the biosecurity level in swine farms. Moreover, the biosecurity score described here has a relatively low number of items, which makes the application of this tool easier and faster compared to other previously described biosecurity assessment.