Epsilon Archive for Student Projects

Abstract

Two haematological instruments intended for small animal clinic use were
evaluated at the Department of Clinical Chemistry, Swedish University of
Agricultural Sciences (SLU), Uppsala, Sweden. The QBC-V and Vet abc were
compared to a reference instrument, the Cell-Dyn 3500. Fresh blood samples were
obtained from canine and feline patients from the University Veterinary Hospital,
SLU. The QBC-V was used for 111 canine samples and 78 canine blood samples
were analysed with Vet abc. From cats, 23 samples were analysed with both
instruments.

For most of the erythrocyte parameters, correlation was excellent for both
instruments (r=0.97-0.99). However, some samples from animals with
regenerative anaemia had inconsistent results for both the erythrocyte and
leukocyte parameters.

The QBC-V had only a fair correlation (r=0.62) for total canine leukocyte count
and granulocyte count (0.63). When neutrophils and eosinophils were separated,
correlation was good for the neutrophils (0.92) but only fair for the eosinophils (0.66). Canine mononuclear cell counts (lymphocytes plus monocytes) had poor correlation (0.39). For feline samples, the QBC-V had fair correlation (0.75) for
the total leukocyte count, good correlation for the granulocytes (0.94) but poor for
the mononuclear cells (0.49).

The Vet abc had fair correlation (0.78) for total canine leukocyte counts, no to
poor correlation (0.01-0.54) for the differential leukocyte count, except for fair correlation (0.78) for the granulocyte count. Only ten results had no flags for potential errors in samples from the twenty-three cats but there was excellent correlation (0.97) for the leukocyte count and fair to excellent for the parameters in the differential count (0.63-0.95) for the samples without flags.

Both instruments had good correlation for canine thrombocyte counts. The Vet
abc seemed less accurate because it lacked a flag warning for thrombocyte clumps
or problems with differentiation of large thrombocytes from small erythrocytes.
Feline thrombocyte counts were not possible to evaluate because of their tendency
to form clumps.

In conclusion, both instruments determined most erythrocyte parameters well
and reasonably well for the total leukocyte count. Both instruments had problems
with the differential count in many patient samples. Granulocyte counts may be
accepted after careful evaluation of histograms and flags. Both instruments are
easy to operate but it is important that the operator is experienced with the
instrument and correctly can evaluate the flags, given results, and histograms. In
addition, manual review of patient's blood smears is recommended for diagnostic
information not provided by haematology instruments, such as erythrocyte
morphology, leukocyte morphology and the presence of platelet clumping.