Early pregnancy diagnosis by transrectal ultrasonography in dairy cattle

The objective of the present study was to determine differences in time of detection of pregnancy between heifers and cows and the interval after insemination at which the maximum sensitivity and negative predictive value of transrectal ultrasonography were obtained. One-thousand-four-hundred transrectal ultrasonographies (TRUS-1; 1079 in cows and 321 in heifers) were performed using a 5-MHz linear-array transducer. The cattle were randomly assigned to have TRUS performed once between days 24 and 30 (estrus = day 0) in cows or between days 21 and 27 in heifers. Every TRUS diagnosis was subsequently compared with a second TRUS diagnosis (TRUS-2), performed 3-8 days later, after day 30 (range 31-38) for cows and after day 27 (range 28-35) for heifers. The sensitivity and specificity between cows and heifers for the common days of TRUS (from 24 to 27) were compared. In cows, sensitivity increased gradually from 74.5% at day 24 to 100% at day 29 (P < 0.01). Specificity increased from days 24-25 and reached a plateau of 96.6% on day 26 (P < 0.01). In heifers, sensitivity increased from 50% at day 21 to 100% at day 26 (P < 0.01). Specificity increased from 87.5% at day 21 and remained steady at 94% starting on day 23 (P > 0.05). The sensitivity for cows and heifers was 89.2 and 96.8%, respectively (P < 0.05) and the specificity was 93.0 and 93.4% (P > 0.05). In this study, heifers were diagnosed pregnant earlier than cows, and the maximum sensitivity and negative predictive value were obtained 3 days earlier in heifers than cows (days 26 and 29, respectively).