# Comparative analysis of platelet counts using Beckman Coulter DxH-900, Mindray BC-6800 Plus, CellaVision DM9600 and the flow cytometry reference method: addressing the challenge of giant platelets

## TL;DR

This study compares different methods for counting platelets, finding that optical methods perform better than impedance methods when giant platelets are present.

## Contribution

The study introduces a comparative evaluation of platelet counting methods in the presence of giant platelets, proposing correction factors and digital morphology integration for improved accuracy.

## Key findings

- Impedance methods showed high imprecision and underestimated platelet counts in samples with giant platelets.
- Optical methods correlated well with the reference method but overestimated platelet counts.
- Digital morphology showed strong agreement with the reference method in challenging samples.

## Abstract

To evaluate the analytical performance of platelet counts obtained by impedance- and optical-based methods on the DxH-900 and BC-6800 Plus analyzers, and by digital morphology using the CellaVision DM9600, in comparison with the international reference method (IRM), in samples with thrombocytopenia and both normal-sized and giant platelets.

Platelet counts were analyzed in peripheral blood samples from patients with thrombocytopenia and normal platelet counts, including cases with normal-sized and giant platelets. Each sample was tested using two automated analyzers: the DxH-900 (impedance) and the BC-6800 Plus (impedance and optical). Results were compared with the IRM based on flow cytometry, performed on the DxFlex analyzer. Precision was assessed. The methods were compared by Passing-Bablok regression and bias calculation. Additionally, platelet morphology was reviewed digitally using CellaVision DM9600.

All methods demonstrated high precision and strong correlation with IRM in samples with normal-sized platelets. However, significant discrepancies were observed in the presence of giant platelets. Impedance methods showed higher imprecision (CV>10 %) and underestimated platelet counts with a negative bias exceeding −25 %. Conversely, the optical method showed better correlation (r=0.9888) and precision (CV=1.6 %), although it overestimated counts with a positive bias of 19.4 %. Digital morphology also demonstrated strong agreement with IRM.

These findings highlight the analytical limitations of routine technologies in challenging samples and emphasize the importance of method selection for accurate platelet reporting. A correction factor for optical counts and the integration of digital morphology could enhance diagnostic reliability in cases involving giant platelets.