RESUMO
This study was designed to evaluate early pregnancy diagnosis using colour Doppler ultrasonography (US) for luteal vascularisation assessment. In Study 1, 28 ewes were artificially inseminated (Day 0), and luteal vascularisation was assessed from Day 12 to Day 19 by two evaluators using colour Doppler US, categorising the corpus luteum (CL) on a subjective scale ranging from 1 to 4. Females bearing a CL with score 2 or greater were presumably considered pregnant. Pregnancy was confirmed on Day 30 by B-Mode US. In Study 2, a predictive pregnancy diagnosis was performed on Day 17 in 197 ewes based on the criteria described in Study 1. Pregnancy was confirmed by B-mode US on Day 45. Agreement between evaluators was verified by an intraclass correlation coefficient (ICC) and Kappa index (κ). Performance of colour Doppler US for early pregnancy diagnosis was evaluated calculating sensitivity (Sens), specificity (Spec), negative predictive values (NPV), positive predictive values (PPV) and accuracy (Ac). In Study 1, luteal vascularisation assessment was unable to predict non-pregnant animals between 12 and 14 days after insemination, as all animals still had vascularised CL, and thus were considered pregnant. The colour Doppler US performance improved progressively until Day 17, when it reached maximum values (Sens = 100%, Spec = 76%, PPV = 73%, NPV = 100% and Ac = 86%). The subjective scale for luteal irrigation assessment showed medium to good agreement among evaluators on Day 12 and Day 13 (ICC = 0.66 and 0.68, respectively), and excellent agreement from Day 14 to Day 19 (ICC = 0.90, 0.80, 0.80, 0.84, 0.95 and 0.93, respectively). Agreement was almost perfect for score 1 CLs (κ = 0.87), and moderate for scores 2, 3 and 4 CLs (κ = 0.54, 0.48 and 0.41, respectively). In Study 2, performance of colour Doppler US as a tool to predict pregnancy status in ewes on Day 17 post-insemination was as follows: Sens = 93.5%, Spec = 80.8%, PPV = 85.6%, NPV = 91.1% and Ac = 87.8%. Subjective luteal vascularisation assessment using colour Doppler US to distinguish between pregnant and non-pregnant animals was considered a reliable tool which was highly efficient beginning 17 days after breeding.
