Diagnostic performance of two serologic tests and fecal culture for subclinical paratuberculosis, and associations with production

A sample of 3 dairy herds served by the same private veterinary practice was selected on the basis of similarity of size, calf-rearing practices and owners cooperation. All animals were tested for paratuberculosis 3 or 4 times at 6-month intervals by faecal culture, (ELISA), agar-gel immunodiffusion (AGID) serology. Faecal-culture results were reported to herd owners through their veterinary practitioner. Production data and somatic-cell counts were collected from 2 of the herds using DHI (Dairy Herd Improvement, Powell, OH) records. Sensitivity and specificity relative to faecal culture of single test results were 71% and 83% for ELISA and 38% and 100% for AGID, respectively. Repeated tests gave sensitivity and specificity of 73% and 61% for ELISA and 33% and 100% for AGID when tests were interpreted in parallel testing. A maximum-likelihood procedure was used to generate estimates of sensitivity and specificity of single (91% and 82%) and repeated (73% and 66%) ELISA testing. There were no significant differences by sign test between faecal culture and ELISA in the time period in which infection was first identified. Faecal culture detected infection earlier than AGID. Mean milk production (controlling for differences in somatic-cell counts) was 3275 lb per lactation or 18.8% higher in faecal culture-negative cows than in faecal culture-positive cows. No associations were found between milk production and ELISA results..