Predictive value of multiple drag-swab sampling for the detection of Salmonella from occupied or vacant poultry houses

The results of four independent complete drag-swab samplings of a 31-farm broiler-production complex are described. Samplings of vacant poultry farms (i.e., between flocks) and occupied farms (i.e., containing chickens) were completed in four 3-day intervals beginning in July 1992 and ending in May 1993. During the first two sampling periods, two drag-swab assemblies were dragged through each individual house and pooled into one sample to evaluate overall incidence of Salmonella detection among vacant and occupied houses. During the second two sampling periods, four drag-swab assemblies were dragged through houses to evaluate frequency of Salmonella detection on individual assemblies within each house. Over the four sampling periods, mean isolation frequencies were 42.2% for vacant poultry farms and 76.2% for occupied poultry farms. When four assemblies were dragged in vacant houses and cultured separately, the probability of detecting salmonellae contamination increased with increasing numbers of assemblies, indicating that multiple drag-swab assemblies should be used for sampling either vacant or occupied broiler houses for the presence of Salmonella. On vacant farms, 89.5% of the total isolations occurred in only one house of two-house farms; the remaining 10.5% of farms were positive in both houses. Salmonella isolation from occupied farms occurred in 54.2% of the samplings in only one house, whereas 45.8% of occupied farms were positive in both houses. These data show a higher frequency of Salmonella isolation on multiple swabs in occupied poultry houses, possibly indicating a quantitatively greater presence of Salmonella in these houses.