J. Hoorfar et al., EVALUATION OF AN O-ANTIGEN ENZYME-LINKED-IMMUNOSORBENT-ASSAY FOR SCREENING OF MILK SAMPLES FOR SALMONELLA-DUBLIN INFECTION IN DAIRY HERDS, Canadian journal of veterinary research, 59(2), 1995, pp. 142-148
Levels of antibodies to the O antigens (0:1,9,12) of Salmonella dublin
were tested in 1355 serum, 1143 cow milk and 160 bulk milk samples fr
om dairy herds using an enzyme-linked immunosorbent assay (ELISA). In
order to define the background reaction, milk samples from all lactati
ng cows and serum samples from 9 animals were collected in each of 20
salmonellosis-free herds located on the island of Bornholm, where catt
le salmonellosis has not been reported. Similar samples were collected
from all stalled animals in 10 herds with recent (<6 months) outbreak
s of salmonellosis located in Jutland, where salmonella infection is e
nzootic. Using herd history of salmonellosis, herd location and clinic
al status of the herds as criteria, the optimal cutoff in the milk ELI
SA was determined as being at least 5% of the samples having optical d
ensity >0.5, resulting in herd sensitivity of 1.0 and herd specificity
of 0.95. While none of the sera in the herds from Bornholm was ELISA
positive, 2 herds had a few reactors in the milk ELISA. Using the same
cutoff, all but 1 bulk milk sample from 150 herds on Bornholm was ELI
SA-negative, and all 10 salmonellosis-positive herds from Jutland were
ELISA-positive. A significant correlation was found between ELISA rea
ctions in milk and in serum of cows (34% and 32% respectively, r(s) =
0.69, P < 0.001). The results indicate the possibility of applying cow
and bulk milk samples to screening and, eventually, regular certifica
tion of herds.