Pre-analytical and analytical variability of reticulocyte counts in dogs.

Abstract

Automated fluorescence-based haematology analysers are now available for reticulocyte enumeration in veterinary medicine, but manual counting is still largely used. This study aimed to evaluate potential sources of analytical and pre-analytical errors when performing automatedand manual counts. Automated and two-operator double-blind manual reticulocyte counts were performed on 15 blood samples. The intra-assay variation of the automated and manual counts and the interoperator variationin the manual counts were then calculated. In addition, the effects of storage were evaluated using samples refrigerated at 4°C or stored at room temperature for 2, 4, 12, 24, 48 or 72 hours after sampling. Intra-assay coefficients of variation were lower for automated counts than for manual counts. Comparison between automated and mean total manual reticulocyte count showed no significant differences. In both refrigerated samples and those stored at room temperature, an increase in reticulocyte count was recorded only after 72 hours. Staining artefacts occurred only in one stored sample counted manually. The presence of cytoplasmic particles other than RNA can cause misinterpretation of cells, leading to an erroneous reticulocyte count. The use of an automated analyser is preferable for reticulocyte enumeration in dogs. Common storage conditions seem to minimally affect reticulocyte evaluation; however, it is recommended to perform the analysis as soon as possible after sampling.