Canine complete blood counts: a comparison of four in-office instruments with the ADVIA 120 and manual differential counts

Background: With more use of bench-top in-office hematology analyzers, the accuracy of reported values is increasingly important. Instruments use varied methods for cell counting and differentiation, and blood smears may not always be examined. Objective: The purpose of this study was to compare canine CBC results using 4 bench-top instruments (Hemavet 950, Heska CBC-Diff, IDEXX LaserCyte, and IDEXX VetAutoread) with ADVIA 120 and manual leukocyte counts. Methods: EDTA-anticoagulated canine blood samples (n = 100) were analyzed on each instrument. Manual differentials were based on 100-cell counts. Linear regression, difference plots, paired t-tests, and estimation of diagnostic equivalence were used to analyze results. Results: Correlations of HCT, WBC, and platelet counts were very good to excellent between all in-office instruments and the ADVIA 120, but results varied in accuracy (comparability). Hemavet 950 and Heska CBC-Diff results compared best with ADVIA results and manual leukocyte differentials. HCT and platelet counts on the IDEXX VetAutoread compared well with those from the ADVIA. Except for neutrophil counts, leukocyte differentials from all instruments compared poorly with ADVIA and manual counts. Reticulocyte counts on the LaserCyte and VetAutoread compared poorly with those from the ADVIA. Conclusions: The Hemavet 950 and Heska CBC-Diff performed best of the 4 analyzers we compared. HCT, WBC, and platelet counts on the LaserCyte had minimally sufficient comparability for diagnostic use. Except for neutrophils (granulocytes), leukocyte differential counts were unreliable on all in-office analyzers. Instruments with a 5-part leukocyte differential provided no added benefit over a 3-part differential. Assessment of erythrocyte regeneration on the LaserCyte and VetAutoread was unreliable compared with the ADVIA 120.