Chemiluminescent immunoassay as a microtiter system for the detection of Salmonella antibodies in the meat juice of slaughter pigs

Chemiluminescent immunoassay (CLIA) was applied in the screening of swine meat juice samples obtained from different laboratories in Germany, using the indirect enzyme linked immunosorbent assay (ELISA) as test for comparison. Out of the 1350 samples tested, 987 were found acceptable for validation of results. A good level of agreement between the two tests was obtained with a kappa value of 0.824 at 20% cut-off and 0.798 at 40% cut-off. At 20% and 40% cut-off levels, a sensitivity of 96.2% and 97.3%, respectively, and a specificity of 94.6% and 95.1%, respectively, were demonstrated between CLIA and ELISA. The detecting lipopolysaccharide (LPS) antigen was tested for specificity and a cross-reaction with two Escherichia coli and Yersinia strains was found when tested with ELISA. This reaction was not observed in CLIA, possibly because of the broader measurement spectrum of this test, which allows a more distinctive definition of immunologic reactions. The same explanation can be given for the increased number of meat juice samples which were positively detected only in ELISA but not in CLIA. Because of the wide detection range in CLIA, a normalization scheme was necessary to obtain reproducible results in this test system. The samples positively classified in screening were further tested for reciprocal titres in both test systems, and a higher correlation between screening and titration results was obtained for CLIA. Based on the results of this study, CLIA can be used as a reference method in detecting Salmonella antibodies in the meat juice of slaughter pigs.     

Chemiluminescent immunoassay in comparison with the indirect ELISA as reference method for detecting Salmonella antibodies in swine meat juice

The efficiency of chemiluminescent immunoassay (CLIA) in detecting Salmonella antibodies in the meat juice of slaughter swine was compared with the indirect ELISA (BgVV method). Based on the screening test results of 987 meat juice samples obtained from different laboratories in Germany, a good level of agreement between the two systems was obtained with a kappa value of 0.824 at 20% cut-off and 0.798 at 40% cut-off. At 20% and 40% cut-off levels, a sensitivity of 96.2% and 97.3%, respectively, and a specificity of 94.6% and 95.1%, respectively, were demonstrated between CLIA and ELISA. The detecting LPS antigen was tested for specificity and a cross-reaction with two E. coli and Yersinia strains was found when tested with ELISA. This reaction was not observed in CLIA, possibly because of the broader measurement spectrum of this test which allows a more distinctive definition of immunologic reactions. The same explanation can be given for the increased number of meat juice samples which were positively detected only in ELISA but not in CLIA. The positively classified samples in screening were further tested for reciprocal titers in both test systems, and a higher correlation between screening and titration results was obtained for CLIA. Towards the end of the study, a preliminary comparison of CLIA with two available commercial ELISA test kits was conducted and the same tendency was observed, namely, wider detection range of CLIA compared to the other tests. Based on the results of this study, CLIA can be used as a reference method in detecting Salmonella antibodies in the meat juice of slaughter pigs.     

Control panels of meat juice samples for a Salmonella enzyme-linked immunosorbent assay.

In the Danish pig production system, an indirect enzyme-linked immunosorbent assay (ELISA) for detection of antibodies in meat juice is used for Salmonella surveillance. Quality control (QC) of this ELISA was previously based on repeated testing of control serum samples. The purpose of the study reported here was to collect, characterize, and implement a panel of meat juice pools for supplemental internal QC. Muscle samples for extraction of meat juice were collected from slaughter pigs of 5 herds infected with Salmonella spp. and from 4 herds without Salmonella infection. A QC panel with 39 pools of meat juice, yielding ELISA optical density (OD) values covering the full range of expected OD values, was prepared and tested repeatedly to determine mean and SD OD values. Each pool was tested twice on each microtitration plate, and the results were used to determine limits for validity of future tests. This QC panel was included as an internal QC to be tested every month. Besides the QC panel, 2 panels containing 100 samples of meat juice with OD above the positive cut-off value and 100 samples with OD below that value were prepared for quarterly control of the diagnostic sensitivity (DSe) and the diagnostic specificity (DSp) of the ELISA. The inclusion of these panels in the QC system will provide information about drifts in DSe and DSp of the test. The procedures described here can be applied to other tests where meat juice samples are used for testing.     

Dual-label time-resolved fluoroimmunoassay for simultaneous quantification of haptoglobin and C-reactive protein in meat juice from pigs

A new method was developed to simultaneously measure 2 acute-phase proteins (APPs) by time-resolved immunofluorometry. The assay, based on double-label quantification of haptoglobin (Hp) and C-reactive protein (CRP) in meat juice samples from pigs, was constructed by use of a combination of europium and samarium chelate lanthanides as labels. Meat juice samples from 154 pigs were used for analytic and clinical validation of the assay through determination of precision, accuracy, limit of detection, and quantification. The analytic performance of the assay was satisfactory, with good intra-assay and interassay precision and accuracy. The levels of Hp and CRP were increased in the meat juice samples of diseased animals compared with healthy ones. According to the results, higher sensitivity could be achieved if the cut-off values of both proteins were taken into account for clinical relevance rather than used individually. Since the dual assay saved both time and sample, it could be used as a rapid and sensitive screening test in porcine production.     

Development and Bayesian evaluation of an ELISA to detect specific antibodies to Sarcoptes scabiei var suis in the meat juice of pigs

Samples of ear scrapings, serum and diaphragmatic muscle were collected from 271 fattening pigs at the slaughterhouse. The scrapings were examined for the presence of mites, and tests for specific antibodies to Sarcoptes scabiei var suis in the serum and meat juice were made with an experimental ELISA. The cut-off value for the meat-juice ELISA was estimated at an optical density of 0.5 by receiver operating characteristic curve analysis, on the basis of the cut-off value for the serum ELISA of 0.4. The results of the three tests were used in a Bayesian model to estimate the characteristics of each test. The specificity of the tests of the ear scrapings was considered to be 1 and their sensitivity was estimated by Bayesian analysis to be 0.86, with a 95 per cent confidence interval (CI) of 0.73 to 0.99. The sensitivity of the meat juice ELISA (0.71, 95 per cent CI 0.6 to 0.8) and its specificity (0.77, 95 per cent CI 0.66 to 0.89) were comparable with the sensitivity (0.73, 95 per cent CI 0.6 to 0.8) and specificity (0.81, 95 per cent CI 0.69 to 0.95) of the serum ELISA.     

Experimental studies in pigs on Trichinella detection in different diagnostic matrices

A total of 72 specific pathogen-free (SPF) and Iberian pigs (three animals per group) were inoculated with 200, 1000 or 20,000 muscle larvae of T. spiralis, T. nativa, T. britovi and T. pseudospiralis. For each animal, the muscle larva burden was evaluated in nine muscle samples by digestion. The anti-Trichinella IgG kinetics in blood samples, taken twice prior and at days 5, 10, 15, 20, 25, 30, 40, 50 and 60 post-inoculation, and in muscle juice, obtained at necropsy, was evaluated by an ELISA using an excretory/secretory antigen. The mean larval recovery rate in SPF/Iberian pigs corresponded with the level of inoculum dose, and tongue, diaphragm and masseter were identified as predilection muscles. In SPF and Iberian pigs receiving 20,000 larvae of T. spiralis, an earlier seroconversion was detected from day 25 post-inoculation. At a 10-fold dilution, the muscle juice showed a good test agreement with blood serum.     

Comparison of muscle fluid and serum for detection of antibodies against hepatitis E virus in slaughter pigs

Muscle fluid was investigated as an alternative to serum for detecting anti-hepatitis E virus (HEV) antibodies in slaughter pigs. Samples of serum and diaphragmatic muscle juice from 67 pigs were analysed by anti-HEV IgG ELISA. Compared to the serum ELISA, the ELISA on diaphragmatic muscle fluid had a sensitivity of 93%, a specificity of 91.7%, a κ value of 0.84 and a determination coefficient (R(2)) of 0.77; both tests had global agreement of results. Muscle fluid can be used as an alternative to serum for serological detection of HEV antibodies in slaughter pigs.     

Prevalence of Neospora caninum infection in Australian (NSW) dairy cattle estimated by a newly validated ELISA for milk

AIM: To determine the performance characteristics of an Institut Pourquier (IP) enzyme-linked immunosorbent assay (ELISA) for the detection of antibodies against Neospora caninum in bovine milk and subsequent determination of the prevalence of N. caninum infection in New South Wales (NSW) dairy cattle. METHODS: Matching serum and milk samples from 93 cattle were assayed in two commercially available ELISAs for the detection of anti-N. caninum antibodies. Serum test results of one ELISA (IDEXX) were used to determine the N. caninum infection status of the cattle. Optimised cut-off values for the IP ELISA using milk samples were determined by two-graph receiver operating characteristic (TG-ROC) analysis and then applied to a representative sample of 398 milk samples from dairy herds around NSW. RESULTS: When this ELISA was applied to a representative collection of 398 milk samples from dairy cattle across NSW it demonstrated a 21.1% prevalence of N. caninum infection in those cattle. From the TG-ROC analysis an IP ELISA protocol was derived which suggested a cut-off threshold that would allow milk testing with 97% sensitivity and specificity, respectively, relative to serum testing. CONCLUSIONS: The prevalence of N. caninum in NSW dairy cattle was higher than previously believed. When used on individual milk samples this ELISA demonstrated high sensitivity and specificity and so could be used to accurately identify N. caninum infection. TG-ROC analysis of the IP ELISA optimised the protocol and prescribed cut-off values enabling the ELISA to be used for the screening of N. caninum antibodies in the milk of dairy cattle.     

Evaluation of an enzyme-linked immunosorbent assay for diagnosis of trichinellosis in swine.

Experimental and field trials were conducted to evaluate an ELISA for its ability to detect Trichinella-infected domestic swine and to compare ELISA results with muscle-digestion test results. The ELISA used was a commercial double-antibody kit, containing an excretory-secretory antigen, and was evaluated principally for epidemiologic use. Experimentally induced infection in swine (4 groups of 3 pigs each; inoculated with 0, 50, 500 or 5,000 larvae) was detected as early as postinoculation week 4, with seroconversion of all inoculated swine by postinoculation week 8. The rate of seroconversion appeared to be affected by initial larval dose, time after inoculation, and immunocompetence of the individual host. Determination of antibody kinetics generally revealed rapidly increasing antibody titer, followed by its steady decrease in most pigs. Once seropositive, however, all pigs remained seropositive for the duration of the 10-week study. Presence of muscle larvae was confirmed in all infected pigs at termination of the study. We recognize that the experimental conditions may not be truly representative of those under which natural infection develops in pigs; however, the ELISA detected an infected pig with muscle larval density of 0.87 larvae/g of tissue. Results of a field trial (n = 310) indicated no muscle digestion test-positive pigs (35 g of diaphragm muscle digested/pig), but 3 samples tested positive by ELISA for specificity of 99.0%.     

Comparing the results of the serological detection of Salmonella antibodies in blood serum and meat juice from different muscles from slaughter pigs

The German Salmonella Monitoring Programme started by the QS-System in 2002 (Blaha, 2004) is mandatory due to the so-called "Salmonella Regulation for Pigs" since 2007 (Anonym, 2007). The Regulation does not clearly prescribe the specific muscle which is to be taken as source of the meat juice. Thus, at different slaughter plants meat samples are also taken from different muscles and several scientific papers describe various muscles as source of the meat juice, too. The objective of this study was to compare the serological results of meat juices from three different locations (diaphragm pillar, neck, belly muscle) to each other and to those of the blood serum from exactly the same animals. All samples were simultaneously tested for Salmonella antibodies by two serological tests (Salmo-type Pig Screen, LDL, Germany; HerdChek Swine Salmonella, IDEXX, Germany). Comparisons were carried out between the various sample kinds per animal and between the two test systems. The analysis of all results of the meat juices revealed in both test systems a clear decline of the OD% values from the diaphragm pillar to the neck to the belly muscle. The average OD% values of all samples were higher when measured by the HerdChek ELISA (IDEXX, Germany) than by the Salmotype ELISA (LDL, Germany), especially in blood serum. Since the results of the meat juice samples gained from the diaphragm pillar were in both test systems by far the closest to the results of the corresponding serum blood samples, it is recommended to amend the "Salmonella Regulation for Pigs" by prescribing meat from the diaphragm pillar as the only muscle for gaining meat juice.     

Field evaluation of the enzyme-linked immunosorbent assay for swine trichinosis: efficacy of the excretory-secretory antigen

A field evaluation of an enzyme-linked immunosorbent assay (ELISA) for swine trichinosis was done with sera obtained from 5 herds experiencing ongoing transmission of Trichinella spiralis. Epizootiologic studies conducted on these herds offered an opportunity to evaluate the accuracy of an ELISA, using larval T spiralis excretory-secretory antigens. Sera from 162 infected pigs and 143 serum samples from noninfected pigs originating from the same farms were tested. The infection status of the pigs was determined by digestion of diaphragm or tongue muscle samples. Two criteria were established to classify the ELISA optical density (OD) readings: Criterion I stated that an OD greater than or equal to 5 times the mean OD of several normal swine sera pools was positive; criterion II stated that a OD greater than or equal to 4 times the normal sera values was positive. The results obtained did not reveal obvious serologic variations among infected herds located in the 4 states involved. Overall, the test detected 93% (criterion I) and 96% (criterion II) of infected pigs. The majority of false-negative sera was from hogs that had less than 5 larvae/g of muscle; 1 hog had 73.8 larvae/g of diaphragm muscle. The false-positive rates were 8% for criterion I and 9% for criterion II. The actual rate for these false-positive samples may have been overestimated, because generally, only small tissue samples (0.4 to 10 g) were digested; larger sample sizes might have altered the results. The relevance of this qualification is that these pigs originated from herds with prevalence rates greater than 50%. Other factors that may account for occasional false-positive sera or false-negative sera in the swine trichinosis ELISA are discussed.(ABSTRACT TRUNCATED AT 250 WORDS)     

Evaluation of enzyme-linked immunosorbent assay (ELISA) and immunofluorescent antibody (IFA) test for the detection of porcine reproductive and respiratory syndrome virus (PRRSV) antibody in pigs from conventional farms

To evaluate the immunofluorescent antibody (IFA) test and enzyme-linked immunosorbent assay (ELISA) for detecting the porcine reproductive and respiratory syndrome virus (PRRSV) antibody, conventional pigs in PRRSV-positive and -negative commercial farms were examined. Antibody development patterns in ELISA and IFA tests were compared in 3 week old piglets experimentally infected with the PRRSV. The virus was detected from 2 days post infection (PI) and then the antibody titers and S/P ratios rose by both methods. A total of 208 serum samples were collected from 4 PRRSV-negative farms and 210 samples from PRRSV-positive farms, and were tested for the PRRSV antibody by IFA and ELISA. The titer of 64 should be set as the cut-off point in IFA for field sera. Similarly, the cut-off S/P ratio should be set at 0.4 in ELISA. A high degree of correlation was observed between antibody titers by the two methods in these 418 samples, with a correlation coefficient of 0.84. The coincidence rate between the two tests was 84.7% (354/418). In non-coincident cases, ELISA was able to detect the antibody with a low titer in the serum samples which were negative in IFA but from PRRSV positive farms. ELISA was more sensitive than IFA to detect PRRSV infected animals or farms.     

C-reactive protein measurements in meat juice of pigs

A time-resolved immunofluorometric assay was evaluated for measurement of C-reactive protein in meat juice from diaphragmatic muscle collected from slaughtered pigs. Analytical and clinical validation of the method was performed by using meat juice samples, obtained by freezing and thawing muscle pieces. The intra- and inter-assay coefficients of variation ranged from 2.2-5.8% to 7.9-14.3%, respectively. The limit of detection was 0.00038 microg/ml. The method measured the CRP concentrations in a linear manner with a good accuracy (r=0.99). CRP concentrations in serum were highly correlated with those in diaphragmatic meat juice (r=0.90; p<0.001). CRP concentrations were significantly higher in clinically affected pigs compared to non-diseased pigs. The assay described here provides a sensitive method for measuring CRP concentrations in meat juice, which can represent a suitable alternative to serum or blood samples and simplifies the process of sampling collection at slaughter.     

Evaluation of three commercial enzyme-linked immunosorbent assays for the detection of antibodies against Salmonella spp. in meat juice from finishing pigs in Spain

The control of animal salmonellosis is considered as a major objective in Europe and indirect ELISAs will be important tools for the implementation of control programs for this infection in pigs. We analyse the results yielded by three commercial ELISAs (Herdcheck Swine Salmonella, SALMOTYPE Pig Screen, and PrioCHECK Salmonella) on meat juice samples from a population of slaughter pigs of Aragon, NW Spain, to assess their efficacy using traditional and latent-class approaches. Overall, the Herdcheck Swine Salmonella detected more Salmonella-infected pigs than the other two tests, but its relative sensitivity was low (65.9%). A similar result was observed when only serotypes detectable by this test were considered (69.1%). When a Bayesian approach was used the Herdcheck Swine Salmonella showed also the highest overall accuracy (sensitivity = 88% and specificity = 74%). Our results suggest that a relatively small proportion of the observed prevalence in herds would be explained by using these ELISAs. Also, this study points out that when different ELISA tests are used within the same herd, results may differ substantially. Thus, caution is advised if it is decided to use these assays for herd health classification in Spanish Salmonella control programs.     

Development and validation of an enzyme-linked immunosorbent assay for the diagnosis of porcine proliferative enteropathy.

The objective of this study was to develop an indirect enzyme-linked immunosorbent assay (ELISA) using a sonicated pure culture of Lawsonia intracellularis as the antigen (So-ELISA). A total of 332 serum samples, consisting of 232 experimentally infected animals and 100 animals naturally infected with L. intracellularis, were used to assess the diagnostic sensitivity. Three hundred and fifty-five sera from uninfected animals were used to determine the diagnostic specificity. The receiver operating characteristic and mean +3 standard deviation of optical density (OD) values from uninfected animals were used for selecting cut-off points. The diagnostic accuracy of So-ELISA was considered to be high as the area under the curve index was 0.991 with 0.0029 standard error. The optimal cut-off for So-ELISA was set at 0.45 OD with 89.8% sensitivity and 99.4% specificity based on a combination of good sensitivity and high specificity. No cross-reactivity was found in sera from pigs exposed to Brachyspira pilosicoli, B. hyodysenteriae, Campylobacter mucosalis, C. jejuni, or C. coli. Inter- and intracoefficient of variation of all control sera tested with So-ELISA was less than 10%. The observed agreements between So-ELISA and the immunoperoxidase monolayer assay tested with experimental challenge animals and field samples were 95.08% with 0.88 kappa and 90.65% with 0.74 kappa value, respectively. So-ELISA was able to detect the seroconversion of infected animals at 2 to 4 weeks after exposure to L. intracellularis. Based on the validation results, So-ELISA could be used as an alternative serology for proliferative enteropathy diagnosis.     

Evaluation of recombinant Bhlp29.7 as an ELISA antigen for detecting pig herds with swine dysentery

Swine dysentery (SD) results from infection of the porcine large intestine with the anaerobic intestinal spirochaete Brachyspira hyodysenteriae. Diagnosis of SD traditionally has relied on detecting the spirochaete in the faeces of acutely affected pigs. To date simple and reliable serological assays that can be applied as a diagnostic tool at the herd level have not been available. In the current study a recombinant histidine tagged 29.7 kDa lipoprotein of B. hyodysenteriae (His6-Bhlp29.7) was used as an ELISA plate-coating antigen. Sera (n=1121) from slaughter-aged pigs on 19 farms were tested in this ELISA. Following optimization of the ELISA conditions using hyperimmune control sera, a set of 464 sera from slaughter-aged pigs from five herds where SD did not occur was tested. From these results a suitable cut-off value for herd negativity was defined as the mean optical density reading plus three standard deviations. Testing of 337 pig sera from six farms with SD then showed that the sensitivity of the test at the herd level was 100%, with all six farms having one or more serum samples exceeding the cut-off value for negativity. Finally, 320 sera from eight herds suspected of having SD were examined. Four of these herds were shown to have pigs with titres consistent with SD. The true health status of the other four herds that were serologically negative could not be confirmed. In conclusion, when used on sets of 40 sera from slaughter-aged pigs the His6-Bhlp29.7 ELISA as established proved to be a useful adjunct to the diagnosis of SD at the herd level.     

Comparing validation of four ELISA-systems for detection of Salmonella derby- and Salmonella infantis-infected pigs

The objective of this study was the comparative evaluation of four indirect Salmonella ELISA tests at study time approved in Germany to detect Salmonella infection in pigs.Three tests are based on a LPS-antigen mix and directed against specific IgG antibodies. The fourth test is based on a purified S. Typhimurium whole-cell lysate antigen and discriminates between Salmonella-specific IgM-, IgA-, and IgG- antibodies. In a longitudinal study, two groups of six weeks old hybrid piglets were orally infected with a porcine S. Infantis or S. Derby strain. Clinical and bacteriological parameters were monitored weekly during an observation period of 130 days after infection and serum samples were investigated in parallel with the respective ELISAs. Apparently, the LPS-based ELISA systems used in this study failed to recognize S. Infantis-infected pigs although those animals shed the pathogen in high amounts throughout the study until day 81 post infection (p. i.). In contrast, the isotype-specific Salmonella Typhimurium whole-cell-lysate based ELISA was capable of detecting Salmonella-infected pigs from day ten p. i. at all tested serotypes and revealed the highest sensitivity in detection of S. Infantis-infected pigs. Furthermore, it became apparent that the often used surveillance cut-off value of 40 OD% is not appropriate for intra-vitam detection of S. Infantis- and S. Derby-infected pigs. In contrast, the cut-off values of the ELISAs given by the suppliers result in considerable higher detection rates.     

Serodiagnosis of pleuropneumonia using enzyme-linked immunosorbent assay with capsular polysaccharide antigens of Actinobacillus pleuropneumoniae serotypes 1, 2, 5 and 7.

Capsular polysaccharide antigens of serotypes 1, 2, 5 and 7 of Actinobacillus pleuropneumoniae were used in enzyme-linked immunosorbent assays (ELISAs) to test sera from experimentally infected and field pigs. Specific reactions were found in sera of experimental pigs with antigens of serotypes 1, 5 and 7 whereas the serotype 2 antigen was cross-reactive. A 1:200 serum dilution was used for testing of 300 sera from 21 swine herds in southern Ontario. Cases of pleuropneumonia had occurred in 11 of these herds, but not in the others. The negative cut-off value was the mean optical density at 405 nm (OD405) + three standard deviations (SD) for 16 negative reference sera. Sera from four pigs naturally infected with Actinobacillus suis were tested and found to react to varying degrees with each of the antigens. Therefore a second cut-off value was determined as the mean OD405 + 2 SD for the A. suis sera. Sera which, in the ELISA produced OD readings above the latter cut-off were considered positive for antibodies to A. pleuropneumoniae; those which were lower than the former cut-off were considered negative. Readings between the two cut-off values may have been due to low positive titers or cross-reactivity, possibly with A. suis, and could not be used to predict pleuropneumonia. Of the pleuropneumonia-free herds, none had positive reactors to serotypes 5 or 7, whereas one and two herds had positive reactors to serotypes 1 and 2, respectively. Of the pleuropneumonia positive herds, six had positive reactors to serotype 1, one to serotype 2, four to serotype 5, and eight to serotype 7.     

Latex agglutination test for detecting Trichinella spiralis infections in pigs using muscle extract

The latex agglutination (LA) test, using muscle-juice samples of pigs experimentally infected with Trichinella spiralis and slaughtered 95 days post-infection (p.i.), gave visible results in 3 min; even in a pig receiving an infection dose as low as 10 larvae. The test appeared reliable and easy to perform without the need for special equipment or sample treatments which are necessary for ante-mortem diagnostic methods. The muscle-juice sample could be obtained by compressing the muscle pieces with the fingers at any time post-mortem and was used undiluted. The results of the LA test using serum or muscle-juice samples correlated with those of the enzyme-linked immunosorbent assay (ELISA). Positive results in the LA test and ELISA appeared 27 days p.i. with the use of sera from the pigs infected with greater than or equal to 600 larvae and 56 days p.i. with the serum of a pig infected with 10 larvae. The complement-fixing antibodies were detected in the sera using complement ELISA 86 days p.i. This assay was negative when muscle-juice samples were used.     

Evaluation of serum and milk ELISAs for paratuberculosis in Danish dairy cattle

A milk and a serum ELISA for detection of antibodies against Mycobacterium avium ssp. paratuberculosis (MAP) were evaluated against the complement-fixation test (CFT) and culture of faecal samples from 580 cows collected between August 1996 and December 1996. Milk and serum were obtained concurrently from six dairy herds infected with MAP and from two dairy herds without history of infection with MAP.

A cut-off value of 7 OD% was used in the ELISAs. At this cut-off value, all six culture-positive herds were positive in the serum ELISA but one was negative in the milk ELISA. All six culture-positive herds were positive in the CFT. In the two culture-negative herds, the serum and the milk ELISA deemed all serum samples negative at this cut-off value, whereas four serum samples from one of these herds were positive in the CFT. The highest cut-off value enabling the milk ELISA to record all six culture-positive herds as positive was 4 OD%. The highest cut-off value enabling the serum ELISA to record all six culture-positive herds as positive was 17 OD%. Individual-sample relative sensitivities of the ELISAs ranged from 49 to 64% and relative specificities were 80–96% at the cut-off values of 4, 7 and 17 OD%.