To estimate the source of variance components for some hematological parameters and assess the utility of the conventional population-based reference interval, this study computed index of individuality for blood samples, which were from 13 dogs drawn...
To estimate the source of variance components for some hematological parameters and assess the utility of the conventional population-based reference interval, this study computed index of individuality for blood samples, which were from 13 dogs drawn once weekly for 4 consecutive weeks. Results were subjected to nested analysis of variance. For all parameters measured between-dog variations were greater than within-dog variation. Except for the parameters RBC and MCHC the index of individuality was <1.4. The low reliability coefficient and high index of individuality of ${\leq}0.8$ were found for the majority of hematological parameters. In practical term, the present study indicated that use of hemogram profiles alone in the evaluating clinical state of a single patient should be avoided because of their physiological or natural random variations, and that comparing a single measurement on the blood analytes from an individual dog to the conventional population-based reference range may be too insensitive to detect any significant changes in the blood components of that particular dog. A single measurement may not characterize an individual's average concentration of the parameters even shortterm period.