Pc. Bartlett et al., ENZYME-LINKED-IMMUNOSORBENT-ASSAY AND MICROBIOLOGIC CULTURE FOR DIAGNOSIS OF STAPHYLOCOCCUS-AUREUS INTRAMAMMARY INFECTION IN COWS, Journal of food protection, 59(1), 1996, pp. 6-10
Recent reports have indicated that the relative sensitivity and specif
icity of the ELISA test for detection of intramammary infection of cow
s with Staphylococcus aureus is not as high as originally reported. It
has been suggested that antibodies measured by enzyme-linked immunoso
rbent assay (ELISA) more closely reflect previous infection status rat
her than current infection status, and that the delay in antibody form
ation following infection and the persistence of antibodies after elim
ination of infection may be responsible for some of the discrepancy ob
served between ELISA and bacterial culture results conducted on the sa
me milk sample. This study (n = 209 cows) was undertaken to determine
if an ELISA for S. aureus intramammary infection more closely reflects
previous infection status than it does current infection status, and
to ascertain whether correction of this time-delay factor substantiall
y improves calculated values of ELISA relative sensitivity and specifi
city. Receiver-operator curves were constructed to compare different t
ime-related definitions of microbiologic culture results used for comp
arison with ELISA results. A greater degree of curvature in receiver-o
perator curves indicated that ELISA results did more closely reflect c
ulture results performed on milk samples taken 1 and 3 weeks previousl
y. Insignificant improvement in sensitivity and specificity occurred w
hen the database was limited to cows (n = 140) with milk production gr
eater than 13.6 kg/day. However, values of sensitivity were all less t
han or equal to 90%, and values of specificity were all less than 54%.