The effect of different fecal sample weights on the detection of Salmonella
enterica in swine feces was examined. Sample weights evaluated were rectal
swabs and fecal samples weighing 1 g, 10 g, and 25 g. Comparisons were mad
e on matched fecal samples obtained from individual pigs housed on 2 commer
cial swine farms in North Carolina. Relative sensitivity (number of positiv
e pigs per fecal weight category/number positive in all weight categories)
increased (P < 0.001) with fecal sample weight, and ranged from 9% for rect
al swabs to 78% for 25-g samples. Stomaching of fecal samples did not affec
t detection of S. enterica. These observations demonstrate that fecal sampl
e weight can markedly influence estimates of prevalence of S. enterica in e
pidemiologic studies. Failure to consider the imperfect sensitivity of bact
erial culture in the design and interpretation of epidemiologic studies wil
l lead to underestimation of prevalence and reduced power to detect the pre
sence of S. enterica-infected herds.