Epidemiology of bovine respiratory disease and mortality in commercial feedlots

Abstract

The objective of my research was to quantify epidemiologic parameters associated with feedlot mortality and bovine respiratory disease complex (BRDC); the most significant cause of morbidity in U.S. feeder cattle. We conducted four retrospective studies utilizing individual health and cumulative cohort-level feedlot data. We developed a database that represented 33 U.S. feedlots from six states over ten years. Each project used a subset of these data. We found that the timing of BRDC was associated with important performance and health indices. In the first study, we evaluated the effect of the timing of individual BRDC treatments on standardized net returns. We found important performance and health measures (e.g. hot carcass weight and number treatments) driving net return differences associated with the timing of BRDC. For the second cohort-level study, we classified temporal patterns of BRDC, and evaluated associations among temporal patterns and performance and health. Temporal patterns were significantly associated with mean daily weight gain, days on feed, carcass weight, yield grade, quality grade, cumulative mortality, and retreatment risk. We also evaluated combined mortality and culling risks and quantified the effects of risk factors using count models. All risk factors (arrival weight, gender, and arrival month) were significant and the effects were modified by one another; effects of these covariate patterns have been impossible to quantify in smaller studies. Finally, we assessed the ability of regression models to predict cumulative BRDC morbidity based on arrival risk factors; then assessed the additional value of incorporating daily BRDC morbidity and mortality information. The percent of correctly classified cohorts did increase across days, but the effect of day was modified by weight, month, and feedlot. Information on daily morbidity was beneficial in predicting cumulative morbidity, but daily mortality provided little benefit. Our database containing animal health and cohort-level data allowed us to generate novel information on the effects of the timing of BRDC in feedlot populations. We also demonstrated effects of covariate patterns on adverse health outcomes that heretofore had been difficult to quantify. Finally, we showed that a predictive model for BRDC may be useful for the feedlot industry; this model should be further developed with future research.