2015 ACVIM Forum Research Abstract Program

Apart from parametric and robust methods on untransformed data, the choice of the statistical method has minimal clinicallyrelevant impact on reference interval accuracy. Parametric and robust methods on untransformed data are only appropriate when an analyte population distribution is Gaussian, which cannot always be ascertained from the sample distribution. Consequently, their use should be discouraged. During reference interval construction, accuracy likely relies more on non-statistical steps such as choice of which animals to include, rather than on the statistical analysis. There is increasing awareness that myocardial injury occurs in primary non-cardiac diseases and often goes unrecognized. Myocardial injury is associated with increased morbidity and death in both human and canine patients with non-cardiac diseases. This prospective study assessed the presence of myocardial damage and characterized the acute phase response in a population of cats with systemic disease.