Borchardt S, Burnett T A, Bruinjé T C, Madureira A M L
Farm Animal Clinic, Division for Ruminants and Camelids, Unit for Reproduction Medicine and Udder Health, School of Veterinary Medicine, Freie Universitaet Berlin, 14163 Berlin, Germany.
University of Guelph, Ridgetown Campus, Ridgetown, ON, Canada N0P 2C0.
JDS Commun. 2025 Jul 23;6(5):705-709. doi: 10.3168/jdsc.2025-0785. eCollection 2025 Sep.
This meta-analysis aimed to evaluate the efficacy of automated activity monitoring (AAM) in detecting estrous expression and ovulatory status in cows during the voluntary waiting period (VWP). A comprehensive literature search was conducted in PubMed, ScienceDirect, and Google Scholar using specific search terms. Inclusion criteria focused on studies that assessed estrous expression within the VWP using modern AAM systems alongside blood progesterone (P4) measurements. Four manuscripts involving 2,198 cows were included. Data extraction was performed by a single investigator and validated by a coauthor. The analysis considered estrous expression and ovulatory status determined through serial blood P4 measurements. Cows were classified based on P4 concentrations and estrus alerts into true positive, false positive, true negative, and false negative categories. Statistical analyses were conducted using MedCalc, incorporating 7 experimental groups from the selected manuscripts. The pooled proportion of ovulatory cows with P4 ≥1 ng/mL by 49 DIM (in most studies) was 79.8% (95% CI: 74.9%-84.3%), with significant heterogeneity (I = 86.1%). The proportion of cows with estrous expression detected by AAM by 60 DIM (in most studies) was 64.0% (95% CI: 48.3%-78.3%), also showing significant heterogeneity (I = 98.1%). Sensitivity and specificity of AAM systems to identify ovulatory cows were 70.3% (95% CI: 55.1%-83.4%) and 60.0% (95% CI: 42.5%-76.3%), respectively, both with significant heterogeneity. Positive predictive value was 88.1% (95% CI: 84.9%-91.0%), and negative predictive value was 35.3% (95% CI: 26.3%-44.7%), indicating variability among experimental groups. The findings suggest that although AAM systems show promise in confirming ovulation in early lactation, the lack of estrus detection by AAM does not confirm anovulatory status. The heterogeneity in the data suggests there might be inconsistencies in the precision or configuration of the AAM systems across farms and studies.
本荟萃分析旨在评估自动活动监测（AAM）在检测奶牛自愿等待期（VWP）发情表现和排卵状态方面的有效性。使用特定检索词在PubMed、ScienceDirect和谷歌学术上进行了全面的文献检索。纳入标准侧重于使用现代AAM系统并结合血液孕酮（P4）测量来评估VWP内发情表现的研究。纳入了4篇涉及2198头奶牛的手稿。数据提取由一名研究人员进行，并由一名共同作者进行验证。分析考虑了通过连续血液P4测量确定的发情表现和排卵状态。根据P4浓度和发情警报将奶牛分为真阳性、假阳性、真阴性和假阴性类别。使用MedCalc进行统计分析，纳入了所选手稿中的7个实验组。到49天产犊间隔（DIM）（在大多数研究中）时，P4≥1 ng/mL的排卵奶牛合并比例为79.8%（95%置信区间：74.9%-84.3%），存在显著异质性（I² = 86.1%）。到60 DIM（在大多数研究中）时，通过AAM检测到发情表现的奶牛比例为64.0%（95%置信区间：48.3%-78.3%），也显示出显著异质性（I² = 98.1%）。AAM系统识别排卵奶牛的敏感性和特异性分别为70.3%（95%置信区间：55.1%-83.4%）和60.0%（95%置信区间：42.5%-76.3%），两者均存在显著异质性。阳性预测值为88.1%（95%置信区间：84.9%-91.0%），阴性预测值为35.3%（95%置信区间：26.3%-44.7%），表明各实验组之间存在差异。研究结果表明，尽管AAM系统在确认早期泌乳期排卵方面显示出前景，但AAM未检测到发情并不能确认无排卵状态。数据中的异质性表明，不同农场和研究中AAM系统的精度或配置可能存在不一致。
