Error In Pregnancy Diagnosis By Per-Rectal Palpation In Beef Cows

Abstract

The objective of this study was to determine the error rate of pregnancy diagnosis while differentiating artificial insemination (AI) pregnancy from bull-bred pregnancy in a synchronized breeding program. Angus-cross beef cows (N=1941) were artificially inseminated using Ovsynch- CIDR or CO-Synch-CIDR synchronization protocols. Bulls were introduced two weeks after AI to breed cows that were not pregnant from AI. The pregnancy status was diagnosed and days in gestation of each cow were determined by per-rectal palpation 70 d after AI. Projected gestation length was determined based on days of gestation at pregnancy diagnosis and projected calving date. The gestation length (281.2 ± 4.71 d; Mean ± SD) for the study population was retrospectively calculated from breeding and calving dates. These projected and actual gestation lengths were matched to determine the error in diagnosing pregnancies resulting from AI or by natural service. The error in pregnancy diagnosis was classified as error due to over-estimation (diagnosed pregnant to AI but pregnant to bulls) or under-estimation (diagnosed pregnant to bulls but pregnant to AI) using mean ± 2 SD as cut-off. The error rate for pregnancy diagnosis was estimated as 7.0 % (3.4% due to under-estimation and 3.6% due to over-estimation) in differentiating AI pregnancy from bull-breeding pregnancy for the gestation length of 281 days. Similarly, the error rate for gestation lengths of 279 and 283 days were estimated as 5.8% and 7.1%, respectively. The error rate varied from 4% to 13% among clinicians. In conclusion, estimation of pregnancy rate from AI breeding programs should account for the error in pregnancy diagnosis accuracy.