Efficacy of an automated technology at detecting early postpartum estrus events: Can we detect resumption of cyclicity?

S Borchardt; T A Burnett; W Heuwieser; J L Plenio; R S Conceição; R L A Cerri; A M L Madureira

doi:10.3168/jdsc.2023-0463

Efficacy of an automated technology at detecting early postpartum estrus events: Can we detect resumption of cyclicity?

JDS Commun. 2023 Nov 17;5(3):225-229. doi: 10.3168/jdsc.2023-0463. eCollection 2024 May.

Authors

S Borchardt¹, T A Burnett², W Heuwieser¹, J L Plenio³, R S Conceição⁴, R L A Cerri⁴, A M L Madureira²

Affiliations

¹ Clinic of Animal Reproduction, Freie Universität Berlin, 14163, Berlin, Germany.
² University of Guelph, Ridgetown Campus, Ridgetown, ON, Canada N0P 2C0.
³ Institute for Veterinary Epidemiology and Biostatistics, Freie Universität Berlin, 14163, Berlin, Germany.
⁴ Faculty of Land and Food Systems, University of British Columbia, Canada V6T 1Z4.

Abstract

The objective of this observational study was to evaluate the efficacy of a neck-mounted automated activity monitor (AAM) at detecting early postpartum resumed ovarian cyclicity. A total of 192 lactating cows (primiparous = 73 and multiparous = 119) were enrolled in this study. Cows were continuously monitored by a neck-mounted AAM early postpartum (7 to 30 d in milk; DIM). Calving was classified as assisted (forced extraction of a calf) or unassisted (normal calving). Retained fetal membrane, metritis, hyperketonemia, clinical mastitis, and milk production were recorded. Cows were classified as healthy (i.e., no disease events) or sick (i.e., any disease event). Estrus events were alerted by the AAM using a proprietary algorithm set by the AAM company. Blood samples, from the coccygeal vein, were collected at 15, 18, 21, 24, 28, and 30 DIM for progesterone (P4) analysis. Resumption of cyclicity was considered when P4 concentration was ≥1 ng/mL on any collection day. Cows were considered anovular when P4 concentration was <1 ng/mL on all collection days. Cows were classified as true positive: P4 ≥ 1 ng/mL and at least one estrus alert; false positive: P4 < 1 ng/mL and at least one estrus alert; true negative: P4 < 1 ng/mL and no estrus alerts; and false negative: P4 ≥ 1 ng/mL and no estrus alerts. Statistical analyses were performed by frequency distribution and mixed effects logistic regression procedures on SAS (SAS Institute Inc.). The specificity, sensitivity, accuracy, and positive predictive value of the sensor to detect cows that had resumed cyclicity were 84.0%, 34.1%, 52.1%, and 79.2%, respectively. Out of the 192 cows, 35.9% (69/192) were anovulatory and 37.5% (72/192) had no estrus events between 7 to 30 DIM. Healthy cows were more likely to resume cyclicity in early lactation compared with cows that were sick (78.3 ± 1.9 vs. 32.8 ± 3.1%, respectively) independent of parity. In conclusion, the sensor had a high specificity for detecting anovular cows, but it had lower sensitivity, and thus was not effective at detecting cyclic cows, perhaps due to silent ovulation early postpartum.