Objective: To compare automated (ADVIA 120) with manual differential leukocyte counts and to evaluate the possibility of identifying cases with reliable vs unreliable automated counts. Methods: EDTA-blood samples from 82 dogs, 52 cats, 30 cattle and 20 horses were used. ADVIA 120 results were allocated to two different groups, based on the evaluation of the perox cytograms: group A (reliable cytograms) and group B (unreliable cytograms). A 200-cell differential WBC count was performed by two observers (experienced and unskilled). Data comparison and methods agreement were assessed. Results: The t-test showed significant differences for many WBC classes. Differences in most cases only affected group B results and were always wider than in A. Differentials provided by the two observers showed larger differences between them than comparison between the skilled observer and the ADVIA 120. Agreement between methods was good for neutrophils and lymphocytes in group A, while it was poor or not reliable in group B. Monocyte counts did not agree or were unreliable in any species in both groups. In horses the confidence intervals were always too wide to be accepted. The ADVIA 120 tended to underestimate neutrophils and monocytes and overestimate lymphocytes in dogs, cats and cattle; the opposite happened in horses. Concerning monocytes, the bias was always linked to a proportional error. Conclusions: Statistical analysis revealed significant differences between automated and manual differentials; effects on clinical-hematologic interpretations appear unlikely. Evaluation of absolute counts in normal and pathologic leukograms is needed to confirm this supposition. Results suggest that automated counts are more reliable than manual counts provided by unskilled observers, particularly in cases with a normal leukogram. The different results reported in the two groups suggest that a skilled operator can reliably select the cases for which a manual differential is needed.
comparison between ADVIA120 and manual leukocyte differential counts
RIONDATO, Fulvio;MAGGI, ELISA;MINISCALCO, Barbara
2009-01-01
Abstract
Objective: To compare automated (ADVIA 120) with manual differential leukocyte counts and to evaluate the possibility of identifying cases with reliable vs unreliable automated counts. Methods: EDTA-blood samples from 82 dogs, 52 cats, 30 cattle and 20 horses were used. ADVIA 120 results were allocated to two different groups, based on the evaluation of the perox cytograms: group A (reliable cytograms) and group B (unreliable cytograms). A 200-cell differential WBC count was performed by two observers (experienced and unskilled). Data comparison and methods agreement were assessed. Results: The t-test showed significant differences for many WBC classes. Differences in most cases only affected group B results and were always wider than in A. Differentials provided by the two observers showed larger differences between them than comparison between the skilled observer and the ADVIA 120. Agreement between methods was good for neutrophils and lymphocytes in group A, while it was poor or not reliable in group B. Monocyte counts did not agree or were unreliable in any species in both groups. In horses the confidence intervals were always too wide to be accepted. The ADVIA 120 tended to underestimate neutrophils and monocytes and overestimate lymphocytes in dogs, cats and cattle; the opposite happened in horses. Concerning monocytes, the bias was always linked to a proportional error. Conclusions: Statistical analysis revealed significant differences between automated and manual differentials; effects on clinical-hematologic interpretations appear unlikely. Evaluation of absolute counts in normal and pathologic leukograms is needed to confirm this supposition. Results suggest that automated counts are more reliable than manual counts provided by unskilled observers, particularly in cases with a normal leukogram. The different results reported in the two groups suggest that a skilled operator can reliably select the cases for which a manual differential is needed.
| File | Dimensione | Formato | |
|---|---|---|---|
|
riondato ESVCP 2009.pdf
Accesso riservato
Tipo di file:
MATERIALE NON BIBLIOGRAFICO
Dimensione
72.16 kB
Formato
Adobe PDF
|
72.16 kB | Adobe PDF | Visualizza/Apri Richiedi una copia |
I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
