Premium
Evaluation of Sysmex XT‐2000 i V analyzer performance across a network of five veterinary laboratories using a commercially available quality control material
Author(s) -
Daly Susan,
Freeman Kathleen P.,
Graham Peter A.
Publication year - 2021
Publication title -
veterinary clinical pathology
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.537
H-Index - 51
eISSN - 1939-165X
pISSN - 0275-6382
DOI - 10.1111/vcp.13016
Subject(s) - hematology analyzer , harmonization , external quality assessment , quality (philosophy) , medical physics , medicine , computer science , statistics , veterinary medicine , mathematics , pathology , philosophy , physics , epistemology , acoustics
Background Laboratory and instrument harmonization is seldom reported in the veterinary literature despite its advantages to clinical interpretation, including the use of interchangeable results and common reference intervals within a system of laboratories. Objectives A three‐step process was employed to evaluate and optimize performance and then assess the appropriateness of common reference intervals across a network of six Sysmex XT‐2000 i V hematology analyzers at 5 commercial veterinary laboratory sites. The aims were to discover if harmonization was feasible in veterinary hematology and which quality parameters would best identify performance deviations to ensure a harmonized status could be maintained. Methods The performance of 10 measurands of a commercially available quality control material (Level 2—Normal e‐CHECK (XE)‐Hematology Control) was evaluated during three 1‐month time periods. Precision and bias were assessed with Six Sigma, American Society of Veterinary Clinical Pathology (ASVCP) total error quality goals and biologic variation (BV)‐based quality goal approaches to performance measurement. Results Instrument adjustments were made to 1 analyzer twice and 3 analyzers once between evaluations to improve performance and achieve harmonization. Sigma metrics improved from 37/50 > 6 to 58/60 > 6 and to all >5 over the course of the harmonization project. BV‐based quality goals for desirable bias and for laboratory systems of 0.33 × CV I (within‐subject biologic variation) were more sensitive and useful for assessing performance than the ASVCP total error goals. Conclusions Optimization and harmonization were achieved, and because BV‐derived bias goals were achieved, common reference intervals could be implemented across the network of analyzers.