Observer reliability in counting erythrocyte ghost cells and impact of short-term storage of canine and feline blood samples

Abstract

The presence of erythrocyte ghost cells (EG) in blood smears indicates intravascular haemolysis or in-vitro haemolysis. However, observer reliability in detection of EG has not been documented. Immediate blood smear preparation is advised but may not always be practical. This study investigated the reliability of EG detection and the impact of delayed blood smear preparation. Forty blood smears from dogs and cats were evaluated twice by four observers. EG development was studied in 59 dogs and nine cats, with smears prepared at 0, 2, and 24 h post-collection. Intra- and inter-observer reliability, as assessed using Krippendorff’s Alpha, ranged from 0.567 and 0.949, and 0.522 and 0.848, respectively, indicating moderate to substantial agreement. No significant increase in EG occurrence was noted between 0 and 2 h, but a significant increase was observed at 24 h. The findings suggest that EG can be reliably detected and that blood smears for the evaluation of EG can be prepared up to 2 h after blood collection without compromising detection of EG.