PRECISION AND ACCURACY OF MONOCYTE COUNTING - COMPARISON OF 2 HEMATOLOGY ANALYZERS, THE MANUAL DIFFERENTIAL AND FLOW-CYTOMETRY

The Coulter STKS (Coulter, Hialeah, FL), the Abbott CD3500 (Abbott Dia gnostics, Abbott Park, IL), a 400-cell manual differential, and flow c ytometry using double-staining with fluorescence-labelled monoclonal a ntibodies (CD45-FITC and CD14-PE) on a Coulter Epics Profile II were e valuated for precision and accuracy in relative monocyte counting. STK S, CD3500, and Profile II achieved a precision analogous to a 3,542-, 1,835-, and 11,998-cell differential, respectively, demonstrating the superiority of automated methods, Analysis of 156 normal and abnormal samples revealed that the mean relative monocyte counts of the manual differential, CD3500 and Profile II were not significantly different. Only the STKS results showed a positive bias (0.79% +/- 1.65), which w as increased in lymphocytic samples. Linear regression between the Pro file II as independent viable, and the other techniques yielded accept able correlation coefficients (STKS: 0.861, CD3500: 0.844, manual diff erential: 0.833). Profile II results were also compared to those of a Becton Dickinson FACScan (Becton Dickinson, Mountain View, CA), which yielded an excellent correlation (r = 0.991) but a slightly smaller re lative monocyte count (bias - 0.39% +/- 0.60) of the latter. On the ba sis of these data, the authors recommend the use of monoclonal antibod ies as a new reference method, but also indicate the need for further methodological investigations.