Cross evaluation of three flow cytometric F cell counting methods performed by different laboratories

Three flow cytometric methods of counting F cells were evaluated in the settings of an external laboratory assessment scheme. The laboratories to participate with a different method were located in Oxford (method O), Athens (method A) and Jerusalem, (method J). Two monoclonal anti-gamma chain antibodies were used: monoclonal antibody produced by P. Beverley (Oxford) (BEV) and an antibody provided by Bioatlantic S.A.R.L. (France) (BIO). The specimens tested were mixtures in five predefined ratios of a sample with homozygous deltabeta-thalassemia with 100% F cells with a sample with no F cells. A central independent laboratory prepared and distributed the aliquots (at room temperature) to the participating centers within 2 (O), 3 (A), and 6 (J) days. The performance of the three methods was evaluated by: 1) deviation indices, 2) relative accuracy, as percent difference of the counts from the target values, and 3) bias and linearity by linear regression of the counts versus the target values [parameters: slope (s), y-intercept (y), R squared (Rs), and F ratio]. The highest score of performance was obtained by method A with both monoclonal antibodies.