The predictive value, sensitivity and specificity of palpation per rectum and transrectal ultrasonography for the determination of bovine luteal status

Luteal status (classifications 1 and 3 = high progesterone levels, 2 and 4 = low progesterone levels) was diagnosed in randomly cyclic Holstein cows, using palpation per rectum and B-mode transrectal ultrasonography, respectively. The results were compared in 2×2 contingency tables with radioimmunoassay (RIA) classifications. Cows with milk progesterone levels of 5.0 ng/ml or higher were classified high and cows with levels below 5.0 ng/ml of progesterone were low. The predictive value, sensitivity and specificity results were 1) 88.2, 68.2 and 88.9% for Classification 1; 69.6, 88.9 and 68.2 for Classification 2; 90.9, 83.3 and 87.5 for Classification 3; and 77.8, 87.5 and 83.3 for Classification 4. Within each test, the diagnosis of high progesterone status was more accurate than of low status. The types of errors are discussed. A correlation coefficient of 0.68 resulted between RIA progesterone concentration and ultrasonographic luteal diameter. Ultrasonography was a better diagnostic test than palpation per rectum.