The Automated Schistocyte Count Requires Microscopic Confirmation

Background: The automated fragmented red cell (FRC) is a promising diagnostic tool for the rapid diagnosis of patients with a suspected thrombotic micro-angiopathy (TMA).

Aim: To evaluate the FRC in comparison to the microscopic schistocyte percentage in order to determine its diagnostic utility.

Methods:  Schistocytes were evaluated in 136 samples by microscopy by two competent morphologists according to International Council for Standardisation in Haematology (ICSH) recommendations. This was compared with the FRC on the Sysmex XN-9000 (Sysmex Corporation, Kobe, Japan). The degree of agreement was measured using the Bland and Altman method and Demingregression statistical methods.

Results: Schistocytes were observed in patients with TMA (8.82%), malignancies (11.03%), haemodialysis (34.56%), haemoglobinopathies (8.82%), nutritional anaemias (13.24%) and in neonates (1.47%). The correlation co-efficient between the two morphologists was 0.97 (CI, 0.92 to 1.01) and the mean of the differences was -0.02 (CI, -0.53 to 0.50). The Sysmex overestimated the schistocyte count (2.61, CI, 2.15 to 3.07) and revealed a poor correlation with microscopy (0.21 CI, 0.05 to 0.38). There was no correlation between the percentage of microcytic (%Micro-R) and hypochromic red cells (%Hypo-He) and the FRC.

Conclusion: This study confirms that observer bias between morphologists can be reduced by implementing the ICSH guidelines. Measurement of the automated FRC by the Sysmex XN 9000 analyser overestimated the schistocyte percentage. The FRC requires confirmation by microscopy.