Risk factors for metritis in Michigan dairy cattle using herd- and cow-based modelling approaches

Abstract

A prospective study, using the Michigan portion of the National Animal Health Monitoring System, was conducted to determine disease and management factors that affect the incidence of metritis in Michigan dairy herds. A stratified random sample representing 1% of all Michigan dairy herds was observed. Data related to herd, feeding, and reproduction management, disease and environmental risk factors, and cow-based disease and reproductive data were collected monthly for a 12 month period. Cows with complete management records that were observed for at least 3 months postpartum were included in the study (85 herds, 5278 cows). A multivariate linear regression model was developed for lactational incidence of metritis by herd, and a multivariate random-effects logistic model was developed for the occurrence of metritis by cows. Significant factors associated with increased occurrence of metritis included larger herds, problem calvings, overconditioning, underconditioning, and more active reproductive management. Factors associated with reductions in metritis involved good nutritional management, and working relationships with veterinarians for disease management. Implications for the significance of these factors are discussed, with emphasis on differences in interpretation and use of results from herd and cow-based analyses.