J Clin Pathol Qual Control.  1999 Dec;21(2):343-348.

Method to Harmonize Laboratory Results Using Patient Samples in Automated Blood Cell Analyzers

• Kim HK
• Song KS
• Park Q
• Cho DH

Abstract

BACKGROUND: Although automated blood cell analyzers is well controlled. Discrepancies often occur in patient results obtained from different analyzers in a laboratory which use two or more analyzers, The aim of this study is to harmonize laboratory results among three automated blood cell analyzers using patient samples according to Streichman et al's method.
METHODS
Two random patient samples per day were processed on three blood cell analyzers : Technicon H*3 (Miles Corp., Tarrytown, NY, USA), Sysmex SE9000 (Baxter Healthcare Corp., McGaw Park IL, USA), Coulter Gen-S (Coulter Electronics INc., Hialeah, FL. USA). We chose Coulter Gen-S as the reference instrument for WBC, RBC, hemoglobin, and MCV, and Technicon H*3 for platelet. The normalized ratios were calculated, which is divided the results obtained from the other two by those obtained from the reference analyzer.
RESULTS
The bias of all parameters except Coulter Gen-S platelet is less than 1 SD. The consistent bias in the Coulter Gen-S platelet was detected (mean normalized ratio is 0.90+/-0.05 which is a bias of more than 1 SD). After change of aspiration volume and calibration, the mean normalized ratio of Coulter Gen-S platelet was improved up to 0.99+/-0.08.
CONCLUSION
The harmonization method using patient samples, proposed by Streichman et al., was considered to be simple and inexpensive for continuous monitoring, as well as effective in detecting discrepancy among different analyzer.

MeSH Terms

Bias (Epidemiology)
Blood Cells*
Blood Platelets
Calibration
Delivery of Health Care
Humans