Type and frequency of contacts between Belgian pig herds

Knowledge of the frequency of direct and indirect contacts between pig herds is a requirement for understanding the potential between-herd transmission of pathogens. Our aim was to investigate the different contacts between Belgian pig herds. We obtained data by conducting a postal survey on 421 pig herds in August 2005 and by analysis of available information on livestock movements in the national identification-and-registration database (18-months period in 2004-2006). Direct contacts included transports of pigs by onto-farm, off-farm and between-farm movements. Indirect contacts included vehicles entering the herd and visitors entering the stables. The median number of direct contacts per herd made by onto-farm movements was 0.2/month (Q1: 0; Q3: 0.5). About 1.2% of herds had >=3 onto-farm movements/month. We used a zero-inflated negative-binomial regression model to describe differences in the number of onto-farm movements according to herd size and herd type. Piglet multipliers followed by finishing herds were predicted to have the most onto-farm movements. Farrow-to-finishing herds made less movements compared to breeding herds. A median of 3997 between-farm movements/month was made in Belgium; these mainly concerned piglets. The median number of origin herds during an 8-month period for between-farm movements was 4 (Q1: 2; Q3: 8). For a typical 1-month period, we constructed directed graphs of between-farm piglet and replacement stock movements, illustrating potential receivers and distributors of infection. Of these between-farm movements, many were made over a short distance (median straight-line distance 19km (Q1: 8; Q3: 36)). The median number of vehicles entering a herd and visitors entering the stables was 8/month (Q1: 6; Q3: 13) and 3/month (Q1: 2; Q3: 6) respectively. The number of indirect contacts by vehicles and persons were associated with herd size (Spearman's r: 0.7 and 0.2), herd type and other factors. Skewness of both direct and indirect contacts, illustrated that there was a wide variety in contact structure between pig herds in Belgium. Infection control might benefit by accounting for this variation in contacts and by targeting 'high-risk' herds in case of animal-disease emergencies.