Verification of Analytical Performance of LH750 and HmX Hematology Analyzers

Abstract

Laboratory diagnostics is a medical discipline playing an important role in patient management. Meaningful, accurate and precise routine measurements of complete blood count are essential for diagnosis, risk assessment, treatment and follow-up of patients. Automated instruments usually have a high level of precision, which, for cell counting and cell-sizing techniques, is greatly superior to that achievable with manual techniques. If instruments are carefully evaluated and their correct operation is ensured by quality control procedures, they produce test results that are generally accurate. In the light of this, the current work aims to verify the analytical performance of LH750 and HmX hematology analyzers regarding the complete blood picture parameters as a preparatory step for laboratory accreditation according to the ISO standard ISO 15189: 2012. The verification was conducted at Ain shams University Hospital hematology laboratory on the following hematology analyzers: - LH750 (serial number AS 21076). - HmX (serial number AS 45246). CBC parameters including hemoglobin, RBC count, MCV, WBC count and platelet count were verified for precision, accuracy, linearity, analytical measurement range, carryover, reference range and mode to mode comparability. In the current study, coefficient of variation was used for evaluation of precision and comparing it to the manufacurer specification for within-run precision and to the TEa for between-run precision. This evaluation revealed that both LH 750 and HmX were precise in 100% of analyzed sample regarding all parameters, which is referred to the tight control of IQC and reflect the validity of corrective actions that were done in case of appearance of any warning or rejection rules. For assessment of accuracy, percentage bias was calculated and compared to TEa. Regarding LH 750 % bias was <TEa in all parameters in all the 20 samples. Our accuracy, calculated as % Bias was acceptable. While for HmX accuracy regarding all parameters was accepted. All parameters passed the proficiency testing and were accepted in all the 20 samples. Our accuracy, calculated as % Bias was acceptable being less than total allowable error of each parameter. At least 80% of results are within the criteria for acceptable performance as specified in Subpart I of the CLIA regulations.