Adsa Logo White Adsa Title White

Predictive models for early diagnosis of metritis and clinical cure of dairy cows using behavioral parameters.

V. R. Merenda


Predictive models for early diagnosis of metritis and clinical cure of dairy cows using behavioral parameters.
V. R. Merenda*1, R. C. Chebel1,2. 1Department of Large Animal Clinical Sciences, University of Florida Gainesville, FL, 2Department of Animal Sciences, Institute of Food and Agricultural Sciences Gainesville, FL.

Our objectives were to predict the occurrence of metritis and cure from metritis. Holstein animals (n = 564) were fitted with an automated device 21 d prepartum. Cows were examined for metritis (fetid, watery, red/brown uterine discharge) and, at diagnosis (d 0), were paired by parity and severity of metritis [metritis vs. puerperal (rectal temperature ≥39.5�C)] and were randomly selected to receive ampicillin trihydrate or ceftiofur crystalline free acid. Cows that were not diagnosed with any diseases (n = 374) were enrolled in the study. Cure from metritis was defined as the absence of metritic discharge and rectal temperature <39.5�C, 11 d after diagnosis. Rumination, resting, feeding and activity time from 10 d prepartum until 10 d after metritis diagnosis were used. In all models, farm, lactation, retained placenta, vaginal laceration score, stillbirth, male, twins, dystocia, rumination, resting, feeding, and activity time were included. In addition, BCS prepartum and at calving were included in the models to predict the occurrence of metritis, and severity of metritis and BCS at diagnosis were included in the models to predict the probability of cure. The area under the curve (AUC) according to the receiver operator characteristic for the model predicting metritis at 1 [AUC = 0.74, 95% confidence interval (CI) = 0.70, 0.78] and 2 (AUC = 0.78, 95% CI = 0.74, 0.81) DIM were (P < 0.01) acceptable. The models predicting puerperal metritis at 1 (AUC = 0.81, 95% CI = 0.77, 0.84) and 2 (0.82; 95% CI = 0.78, 0.85) DIM were (P < 0.01) excellent. The AUC of the models using behavioral changes peripartum to predict cure at d −1 (0.81; 95% CI = 0.74, 0.87) and 0 (0.81; 95% CI = 0.75, 0.87) relative to diagnosis were (P < 0.01) excellent, as was the AUC of the model using total behavior time on d 0 (AUC = 0.82, 95% CI = 0.75, 0.88; P < 0.01). This study demonstrates that behavioral changes peripartum allows for the early diagnosis of metritis and whether cows will be cured from metritis, allowing for the implementation of preventive strategies and earlier decision making regarding metritis treatment.

Keywords: predictive model, metritis, cure.

Biography: Victoria Rocha Merenda was born in Sao Paulo, Brazil. She was accepted in 2011 on State University of S�o Paulo, in Botucatu, where she got her degree in Veterinary Medicine. In 2015, she came to USA where she did an externship in North Florida Holsteins for one year. In 2016, she went to MN where she decided to become a graduate student. In the beginning of 2017, Victoria returned to USA to work with Dr. Chebel and pursue a Master of Science in University of Florida.