Avian influenza transmission risks: analysis of biosecurity measures and contact structure in Dutch poultry farming.

In the 2003 epidemic of highly pathogenic avian influenza in Dutch poultry, between-farm virus transmission continued for considerable time despite control measures. Gaining more insight into the mechanisms of this spread is necessary for the possible development of better control strategies. We carried out an in-depth interview study aiming to systematically explore all the poultry production activities to identify the activities that could potentially be related to virus introduction and transmission. One of the between-farm contact risks that were identified is the movement of birds between farms during thinning with violations of on-farm biosecurity protocols. In addition, several other risky management practices, risky visitor behaviours and biosecurity breaches were identified. They include human and fomite contacts that occurred without observing biosecurity protocols, poor waste management practices, presence of other animal species on poultry farms, and poor biosecurity against risks from farm neighbourhood activities. Among the detailed practices identified, taking cell phones and jewellery into poultry houses, not observing shower-in protocols and the exchange of unclean farm equipment were common. Also, sometimes certain protocols or biosecurity facilities were lacking. We also asked the interviewed farmers about their perception of transmission risks and found that they had divergent opinions about the visitor- and neighbourhood-associated risks. We performed a qualitative assessment of contact risks (as transmission pathways) based on contact type, corresponding biosecurity practices, and contact frequency. This assessment suggests that the most risky contact types are bird movements during thinning and restocking, most human movements accessing poultry houses and proximity to other poultry farms. The overall risk posed by persons and equipment accessing storage rooms and the premises-only contacts was considered to be medium. Most of the exposure risks are considered to be similar for layer and broiler farms. Our results, including those on farmer opinions, are relevant for the communication with farmers and poultry-related businesses about practices and risks. We conclude by providing recommendations for improvement of control strategies.