Sensitivity of cultivation of Salmonella enterica in pooled samples of pig faeces

We aimed investigate if the cost of bacteriological examination at herd level could be reduced by cutting down on the number of analyses without loss of sensitivity. Faeces samples sent to the lab for bacteriological examination were analysed by both the standard procedure (20 pen samples) and by mixing the same 20 samples into 10, 5 or 1 pooled sample(s), respectively. The relative sensitivity of the bacteriological analysis decreased to 94% following pooling from 20 into 10; to 92% by pooling from 20 into 5; and to 73% following pooling from 20 into 1. Percent agreement between the standard procedure and pooled samples was \u3e90% if only Salmonella-negative samples, or only Salmonella-positive samples, were pooled. Agreement was \u3e60% for pools of originally negative and positive samples. The suggested alternative pooling methods should be carried out in the laboratory