Hematology analyzer comparison: Ortho ELT-8/ds vs. Baker 9000 for healthy dogs, mice,and rats

A Baker 9000 hematology analyzer (electronic impedance) was purchased to replace an Ortho ELT-8/ds analyzer (laser optics) due to discontinued technical support. An analytical comparison of hemograms from healthy dogs, rats, and mice was made from paired disodium ethylenediamine tetra-acetate anticoagulated blood samples. Both instruments were calibrated with human blood products, and the ELT-8/ds hematocrit (HCT) was calibrated to a spun packed cell volume (PCV) for each species. For Beagle dogs (n = 49), Baker 9000 mean platelet (PCV) counts had a negative bias of -89 X 10(3)/microliter when compared to ELT-8/ds values. Mean +/- standard error manual PLT counts compared well with Baker 9OOO values for dogs (n = 10): 369 +/- 28 vs. 367 +/- 27 X 10(3)/microliter; r = 0.93. For CD-1 mice (n = 44), Baker 9000 mean white blood cell (WBC) counts had positive biases of 1. 1 X 10(3)/microliter when compared to ELT-8/ds and 0.5 X 10(3)/microliter when compared to hemacytometer counts. Diluted microsamples using the predilution mode on the Baker 9000 compared well with undiluted samples for mice. For Sprague-Dawley rats (n = 70), Baker 9000 mean WBC, red blood cell (RBC), and PLT counts had absolute biases of 0.8 X 10(3)/microliter, -1.09 X 10(6)/microliter, and -357 X 10(3)/microliter, respectively, when compared to ELT-8/ds values. Baker 9000 RBC, WBC, and PLT counts from rats compared well with reference hemacytometer counts. The Baker 9000 HCT determination for rats had an absolute negative bias of 6% when compared to the ELT-8/ds values or spun PCV. The Baker 9000 required whole blood calibration to PCV for accurate determination of HCT for rats. The biases between analyzers may be due to inherent physical differences between the analytical methods and/or the calibration techniques.