Analytical errors in nucleated red blood cell enumeration.

Area of Science:

• Veterinary Hematology
• Clinical Pathology
• Laboratory Automation

Background:

• Manual enumeration of nucleated red blood cells (NRBCs) in canine and feline peripheral blood is standard but labor-intensive.
• Automated analyzers offer increased efficiency, but accurate differentiation between NRBCs and leukocytes remains a challenge.
• The Sysmex XN-V Series analyzer has the potential to distinguish NRBCs, but analytical errors have been reported.

Purpose of the Study:

• To investigate discrepancies between automated and manual NRBC counts in canine and feline blood samples.
• To identify and evaluate the causes of analytical errors in automated NRBC enumeration.

Main Methods:

• Retrospective analysis of blood samples with elevated NRBC counts.
• Comparison of automated NRBC counts with manual blood film evaluation using correlation and agreement analyses (Spearman, Passing-Bablok, Bland-Altman).
• Assessment of the precision of the automated method and interobserver agreement of manual counts.

Main Results:

• Good agreement between automated and manual NRBC counts was observed at levels ≤1 NRBC ×10⁹/L in dogs and cats.
• Agreement was inadequate at levels ≥1 NRBC ×10⁹/L, with the automated method showing a negative proportional difference.
• High hematocrits, merged cell fragments/leukocyte clouds, and the presence of early erythroid precursors contributed to automated enumeration inaccuracies. Poor interobserver agreement was noted for manual counts (mean CV 27.3%).

Conclusions:

• The Sysmex XN-1000 V's WNR channel provides precise NRBC enumeration with good agreement to manual counts at ≤1 NRBC ×10⁹/L.
• Manual blood film review is essential for samples with ≥1 NRBC ×10⁹/L, earlier erythroid precursors, or potential gating errors.
• Specific patient populations (e.g., greyhounds, dehydrated animals) may require careful interpretation of automated NRBC results.