A Comparative Study of Biological and Analytical Variabilities in Automated Blood Cell Analysis

Baik, Sae Yun; Kwak, Yun Sik; Lee, Wee Gyo; Hyun, Bong Hak

Abstract

BACKGROUND: The National Committee for Clinical Laboratory Standards (NCCLS) recommends that the analytical variability must not exceed 25% of the biological variability in automated blood cell analysis. This study was conducted to determine whether routine automated blood cell analysis by Coulter STKS (Coulter Corp., Miami, FL, U.S.A) comforms with the NCCLS's recommendations.
METHODS
Routine CBC analysis with STKS was performed on 22 healthy volunteers. The tests included calculating WBC, RBC, hemoglobin, MCV, platelet, MPV, and percentages of the granulocytes, lymphocytes, and monocytes. Blood samples were collected twice in one week interval to study the total variability. For the analytical variability, blood samples from 12 subjects were tested twice immediately after venipuncture for within-run variability, and samples from 10 subjects were tested immediately and 6 hours after venipuncture for within-day variability. The analytical variability was calculated as the sum of within-run and within-day variabilities. The biological variability was calculated by subtracting the analytical variability from total variability. The ratios of analytical and biological variabilities were calculated by dividing the analytical variability by the biological variability.
RESULTS
Ratios of analytical and biological variabilities were as follows: 0.22 for WBC, 0.20 for RBC, 0.21 for hemoglobin, 0.39 for platelet, 1.98 for MPV, 0.07 for %granulocyte, 0.11 for %lymphocyte, and 1.81 for %monocyte. The ratio for MCV was not obtained because the analytical variability exceeded total variability.
CONCLUSIONS
The analytical variability did not exceed 25% of the biological variability in all test categories except platelet, MPV and the percentage of monocyte. Thus, it is recommended that the analytic variability of all test categories be reduced so as to be in conformity with the NCCLS' recommendations.