Development and External Validation of a Predictive Model of Severe Neonatal Calf Diarrhea in Hanwoo Calves Using Animal, Environmental, and Management Risk Factors

Abstract

Background: Neonatal calf diarrhea accounts for most pre-weaned calf losses in Hanwoo cattle. A novel predictive model of severe neonatal calf diarrhea will help veterinarians and farmers prevent disease in calves. Hypothesis/Objectives: Development and external validation of a simple predictive model for severe neonatal calf diarrhea in Hanwoo cattle. Animals: Hanwoo calves were used to develop the model (n = 3179) and for its external validation (n = 1383). Methods: Retrospective, observational study. The predictive model was developed using logistic regression analysis with data from Hanwoo calves from 2019 to 2022. The model was externally validated using data from Hanwoo calves in 2018 and 2023. Results: After univariable and multivariable logistic analyses, the month of birth, rainy weather, duration of pregnancy, dam parity, retained fetal membranes, prevalence of neonatal calf diarrhea, induction of parturition, bedding type, and management of failure transfer of passive immunity were selected as predictors, with a sensitivity of 74.1% (95% confidence interval [CI]: 68.9%–78.7%) and specificity of 72.2% (95% CI: 70.6%–73.8%; area under the curve [AUC]: 0.79, 95% CI: 0.766–0.814). In external validation, the accuracy was 83.3% (95% CI: 81.2%–85.2%). Sensitivity and specificity were 60% (95% CI: 50.0%–69.3%) and 85% (95% CI: 82.9%–86.9%), respectively. Conclusions and Clinical Importance: We have identified predictors for severe neonatal calf diarrhea in Hanwoo calves and have developed a simple, easily calculated scoring prediction model based on these predictors.