Evaluation of the specificity of three enzyme-linked immunosorbent assays for detection of antibodies against Salmonella in bovine bulk milk

BACKGROUND The Swedish Salmonella control program has been running for decades and has resulted in a low prevalence of Salmonella in Swedish food producing animals. Routine bacteriology is used to detect Salmonella, however, bacteriology is time consuming, costly and has a low sensitivity. Different enzyme-linked immunosorbent assays (ELISAs) have been developed for detection of antibodies against Salmonella Dublin and S. Typhimurium in bovine bulk milk, individual milk samples as well as in sera. Screening bulk milk for antibodies against Salmonella spp. could improve the cost-effectiveness of the surveillance in Swedish dairy cattle, but as characteristics of tests may vary in different populations, tests should always be evaluated in the specific population where they will be used. Hence, the aim of this study was to evaluate the specificities of three bovine ELISAs when used to analyse bulk milk samples from Swedish dairy cattle. A second aim was to compare the performance of the two Dublin ELISAs tested. METHODS Bulk milk samples for analysis were randomly selected from samples collected within the Swedish bulk milk sampling scheme and analyzed with the three ELISAs; a Danish in-house Dublin ELISA, PrioCHECK(®) Salmonella Ab bovine Dublin ELISA and PrioCHECK(®) Salmonella Ab bovine ELISA (hereafter named mixed ELISA). The specificities of the ELISAs were calculated assuming a disease-free status in Sweden i.e. that all test positive samples were assumed to be false positive results. This assumption can be used when a disease is known to be infrequent. RESULTS The calculated specificities of the two Dublin ELISAs and the mixed ELISA, when using the producer's recommended cut-off value of the corrected optic-density percent (ODC%) were 99.4% (95% Confidence Interval (CI): 98.8% -99.8%), 99.4% (95% CI: 98.8% -99.8%) and 97.9% (95% CI: 96.8% -98.7%), respectively. The correlation between the ODC% values of the two Dublin ELISAs was 0.83. CONCLUSIONS We conclude that the evaluated ELISAs have sufficiently high specificities to be used as supplement to bacteriological examinations in the Swedish Salmonella control program in cattle as well as a primary screening test in routine surveillance for S. Dublin.