Flow cytometric reticulocyte analysis. Multiinstitutional interlaboratory correlation study

Reticulocyte analysis by flow cytometry offers precision and sensitivity greater than those of conventional morphologic methods and permits derivation of a reticulocyte maturity index. However, interlaboratory variability has not yet been reported. The authors analyzed 310 samples at eight sites using 11 instruments over a 4-month period to examine intermethod and interlaboratory variabilities. Stains included thiazole orange, ethidium bromide, and auramine O. Instruments included models by Coulter, Becton Dickinson, TOA Medical Electronics, and Ortho Diagnostics. The coefficient of variation (CV) among all sites and methods on these samples varied as a function of the reticulocyte percentage, ranging from a mean CV of 69% for samples with < .5% reticulocytes to 24.1% for those with > 2.5% reticulocytes. The best performance was observed with the TOA R-1000 dedicated reticulocyte analyzers, with a mean CV of 18.4% for samples with < .5% reticulocytes and 4.6% for samples with > 2.5% reticulocytes. The reticulocyte maturity index showed comparable intersite precision, with a mean CV of 16% for samples with > 2.5% reticulocytes with multipurpose flow cytometers and a mean CV of 7.3% with the TOA R-1000 instruments. Interclass correlations among all sites ranged from .79 to .99 for the reticulocyte counts and .41 to .88 for the reticulocyte maturity index. The authors conclude that flow cytometric reticulocyte analysis is more precise than manual reticulocyte analysis. With greater automation of this methodology, further interlaboratory standardization of reticulocyte counts and the reticulocyte maturity index can be achieved.