Agreement among evaluators of bovine embryos produced in vivo or in vitro

Six experienced individuals evaluated 40 embryos on videotape for stage of development and quality grade. These 40 observations comprised 15 embryos produced in vivo, 15 embryos produced in vitro, and 10 embryos that were repeated throughout the videotape. Embryos produced in vivo were recovered from uterine flushings of superovulated heifers 7 d after estrus, and embryos produced in vitro were harvested 7 d after insemination of in vitro-matured oocytes. Embryos of various stages (morulae, blastocysts, or degenerated) and quality grades (1 = excellent, 2 = good, 3 = fair, 4 = degenerated) were recorded on videotape for evaluation. After video microscopy, the embryos were stained and the number of nuclei per embryo was counted. Six evaluators reviewed the videotape and the percentage of agreement and kappa (k; agreement beyond chance) among evaluators were determined for classifications of stage and grade. Consistency of each evaluator's responses was estimated using the 10 repeated embryos. Agreement within evaluators was higher for stage of embryo development (89.2%) than quality grade (68.5%). Agreement among evaluators for stage was slightly higher with embryos produced in vivo (85.0%, k = 0.74) than in vitro (72.3%, k = 0.48). Agreement among evaluators for grade was similar with embryos from in vivo (61.0%, k = 0.46) and in vitro (57.7%, k = 0.42) production. For both sources of embryos, agreement was substantially better for Grades 1 and 4 than for Grades 2 and 3. The results of this study suggest that good to excellent agreement exists for classifying Day 7 bovine embryos by stage and by extremes of quality grade (Grades 1 and 4) but not by degree of abnormal morphology (Grades 2 and 3). Simple grading criteria of Grade 1 (highest quality), Grade 2 (morphologic defects), and Grade 3 (degenerated) maximized agreement among evaluators.