Cytometry Part B (Clinical Cytometry) 78B:201–210 (2010)

A Model for Continuous Quality Control Incorporating Sample-to-Sample Assessment of Optical Alignment, Fluorescence Sensitivity, and Volumetric Operation of Flow Cytometers Denise Lawrie, Lindi M. Coetzee, and Deborah K. Glencross* National Health Laboratory Services and Department of Molecular Medicine and Haematology, Faculty of Health Sciences, University of the Witwatersrand, Johannesburg, South Africa

Background: Bead count rate (BCR) monitoring successfully identifies pipetting error during single platform CD4 enumeration. Despite rigorous prescribed quality control performed, preliminary data suggested that BCR outliers could also be attributed to occasional failure of flow cytometric volumetric operation. The aim of this report was to use counting beads in a model of continuous quality control (CQC) to monitor overall flow cytometric performance (laser alignment, fluorescence stability and volumetric operation). Methods: The proposed CQC model used FlowCheckTM and IMMUNOTROLTM blood controls daily. Extended monitoring of fluidics (FPV; beads and sheath only) and sample preparation (SPV; blood, IMMUNOPREPTM and beads) was done daily on five flow cytometers over five consecutive days prior to testing patient samples. Sample-to-sample CQC included monitoring BCR, selected time/fluorescence histograms (Time vs. Count; Time vs. Fluorescence and Forward Scatter vs. Fluorescence) and full peak coefficient of variation (FPCV) for 2000 samples tested. Results: Prescribed quality controls showed Half Peak CV values of 0.05, Spearman correlation) was found between BCR and absolute cell numbers (red cells R2 ¼ 0.0002, white cells R2 ¼ 0.002, lymphocytes R2 ¼ 0.004, and platelets R2 ¼ 0.0005), size of red blood cells (i.e. Mean Cell Volume, R2 ¼ 0.029), concentration of red blood cells (i.e. Hematocrit, R2 ¼ 0.014) or markers of nonspecific inflammation (ESR, R2 ¼ 0.05, n ¼ 25). Pipetting method (retrospective analyses). Although all laboratories in the SA-NHLS network use the standardised PLG CD4 protocol, automated pipetting is only available to higher volume testing facilities (>300 samples per instrument per day). To assess the influence of different pipetting techniques on accuracy of CD4 counts across the network, data collected from a previous national exercise conducted with 30 SA-NHLS laboratories during 2004 was analyzed and showed an overall agreement of 97.5  5.3% (%similarity model used) (31) between normal (forward) and reverse pipetting for absolute CD4 counts (n ¼ 380, CD4 counts of 2–2,766 cells/ll with a bias of 5.6  16 at the clinically significant CD4 count level of < 220 cells/ll). Subsequent introduction of automated pipetting (PrepPlus2, Beckman Coulter) in the Johannesburg CD4 laboratory consistently showed an agreement of 97.5–102.5% (4 PrepPlus 2 instruments) compared with manual forward pipetting, with batch CVs of