Laboratory and clinical evaluation of white blood cell differential counts. Comparison of the Coulter VCS, Technicon H-1, and 800-cell manual method.

Abstract

Eight hundred-cell manual white blood cell differential count evaluations of Coulter VCS and Technicon H-1 included estimates of accuracy, imprecision, random and systematic errors, clinical relevance, and detection of immature neutrophils. Accuracy was acceptable, except for H-1 testing of monocytes. Instrument imprecision was similar, except for tighter monocyte values as measured by the VCS. Deming regression determined random errors were greater for H-1 monocytes and proportional errors were similar. Constant errors for the VCS were less than the H-1 for neutrophils and monocytes and greater than the H-1 for eosinophils and lymphocytes. H-1 morphologic false-negative rates were twice those for VCS. True-positive and false-negative rates in cases with immature neutrophils were 70.6% and 20.3% for the VCS and 35% and 55.9% for the H-1. The reference false-positive rate was 4.4%. Clinically appropriate VCS flags were generated in distributionally abnormal cases. Predictive values were higher for the VCS.