Evaluation of the erythrocyte and platelet morphological analysis given by the hematology analyzer Coulter STKS.

The morphological profiles of red blood cells (RBC) and platelets (Plt) derived from the distribution analysis given by the Coulter STKS were evaluated in two groups of patients suffering from non hematological (n = 293) and hematological (n = 257) conditions. The RBC and Plt flags were studied in terms of sensitivity of the morphological analysis and specificity and significance of each flag. When all RBC and Plt flags were considered, the percentages of false negatives and false positives in our subjects were found to be 4.7% and 13.4% respectively, with a global efficiency of 81.8%. The sensitivity of the alarm system was higher than 90% for all types of abnormality, except microcytosis (81%) and Howell-Jolly bodies (57% over a limited number of 7 cases). The specificity of the STKS response was found to be low except for anisocytosis (88.5%) and macrocytosis (86.1%). It was shown that the flags microcytosis and/or hypochromia and macrocytosis were poorly significant when they appeared in isolation (false positive rates of 86.6% and 84.2% respectively). Thus, these alarms could be eliminated from the review criteria. When considering only the flags anisocytosis, NRBCs, micro RBCs/RBC fragments, dimorphic RBC pop, PLT clumps and giant Plt, the percentage of false positives was 8.1%. However, it must be kept in mind that suspect leucocyte flags remained review criteria, resulting in a final false negative rate of 2.5% for RBC and Plt morphological abnormalities.