Evaluation of a cow-side milk progesterone assay and assessment of the positive predictive value of oestrus diagnosis by dairy farmers in New South Wales.

OBJECTIVES The three objectives of this study were to determine the positive predictive value (PPV) of oestrus diagnosis (heat detection accuracy) by dairy farmers, calculate the diagnostic sensitivity and specificity of the P4 Rapid milk progesterone assay for detecting a corpus luteum and evaluate the economics of using a cow-side milk progesterone assay designed to aid oestrus diagnosis. METHODS Milk samples were collected from 752 cows diagnosed in oestrus by farm personnel on 14 dairy farms. Samples were tested using the P4 Rapid milk progesterone assay to estimate the PPV of oestrus diagnosis at each farm and a crude pooled mean of PPV of oestrus diagnosis across all farms. A further 156 milk samples were collected from cows with luteal tissue status determined by transrectal ultrasound and tested by the P4 Rapid assay to enable calculation of the sensitivity and specificity of the P4 Rapid assay. RESULTS For pooled farm samples, the PPV was 97.0%, with a range between farms of 88.9-100%. Sensitivity of the P4 Rapid milk progesterone assay for detecting a corpus luteum was 90.1% and specificity was 98.7%. Misclassification of oestrus in cows previously identified as pregnant was the most common cause of false-positive oestrus diagnoses by farm personnel. CONCLUSION Routine testing of milk progesterone in all cows diagnosed in oestrus is not economically justified and may even slightly reduce submission rates; conversely, strategic use of cow-side milk progesterone assays can improve herd reproductive performance by facilitating decisions on whether to rebreed cows previously diagnosed as pregnant.