► Click here to download the original poster (PDF)
Joao Mariano Pego, Anália Carmo, Angela Silva, Gilberto Marques, Rui Bartolo, Fernando Rodrigues
Clinical Pathology Department, Centro Hospitalar e Universitário de Coimbra, Coimbra, PORTUGAL
The analysis and quantification of cells in body fluids play a critical role in the diagnosis and prognosis of a disease as well as in the establishment of the therapeutic plan. Traditionally, the cell count was performed by manual microscopy using a counting chamber. This method is time-consuming, labor-intensive and requires experienced laboratory staff around the clock. In order to turn this analysis easier several hematology analyzers were validated for automated cell count in body fluids. The aim was to evaluate the analytical performance of the automatic analyzer, the HORIBA Yumizen H2500, to compare the results with those obtained with the UniCel DxH 800 Coulter analyzer, and with the results from slide review, in different body fluids.
123 body fluids were analyzed including cerebrospinal fluid (CSF), pleural, peritoneal, and synovial fluids. All fluids were analyzed up to two hour after their collection. The Yumizen H2500 analyzer offers a channel to quantify the total cell count of body fluids using impedance and another channel to assess leucocytes differential based on the Double Hydrodynamic Sequential System "DHSS" flow cytometry. The results obtained with the Yumizen H2500 analyzer were compared to those obtained with the DxH 800 and to those obtained with slide review. For the Yumizen H2500 analyzer the measure intervals were: WBC > 15.106cells/L and WBC < 10.000.106cells/L; RBC > 10.000.106cells/L and RBC < 5.000.000.106cells/L. For the DxH the measure intervals were: WBC > 20.106cells/L and WBC < 89.000.106cells/L; RBC > 10.000.106cells/L and RBC < 6.200.000.106cells/L.
Distribution of the 123 samples involved in correlation study
| Parameter | Level 1 | Level 2 | Level 3 | |||
| CV limit (%) | Range | CV limit (%) | Range | CV limit (%) | Range | |
| BFRBC (106/L) | 10 | 10.000 –100 000 | 5 | 100 000 –5 000 000 | -- | -- |
| BFWBC (106/L) | 20 | 15-100 | 15 | 101-600 | 5 | 601-10 000 |
| BFMN# (106/L) | 35 | 15-100 | 20 | 101-600 | 10 | 601-10 000 |
| BFMN (%) | 35 | 5-30 | 25 | 30-80 | 15 | 80-100 |
| BFPN# (106/L) | 35 | 15-100 | 20 | 101-600 | 10 | 601-10 000 |
| BFPN (%) | 35 | 5-30 | 25 | 30-80 | 15 | 80-100 |
Precision study: Repeatability performances
| Parameter | Bftrol level 2 | Bftrol level 3 | ||
| CV limit (%) | CV limit (%) within-run | CV limit (%) | CV limit (%) within-run | |
| BFWBC (106/L) | 10% | 5% | 5% | 3% |
| BFRBC (106/L) | 6% | 5% | 4% | 3% |
Precision study: Reproducibility performances
A. Evaluation of the correlation of RBC according the expression: YH2500=f(DxH800). B. The bias was calculated according the expression: DxH800=f(YH2500-DxH800)
A. Evaluation of the correlation of WBC according the expression: YH2500=f(DxH800). B. The bias was calculated according the expression: DxH800=f(YH2500-DxH800)
Evaluation of the correlation between differential count determined by YH2500 and slide review. A. Polymorphonuclear cells B. Mononuclear cells
In spite of the cytomorphological characteristics of the different fluids, the Yumizen H2500 analyzer had the ability to evaluate all type of fluids and to report total cell, RBCs and WBCs with an adequate linearity, a good limit of detection and a good correlation coefficient as compared to the results obtained with DxH 800.
References: Hussong JW, et al. Chicago, IL: ASCP Press; 2015; Sandhaus LM. Clin Lab Med. 2015; 35(1): 93-103
