ABSTRACT
In this study the ability of the Coulter MAXM analyzer to quantify reticulocytes was evaluated. The results were compared with those obtained by a microscopic method according to NCCLS H44-P recommendations and with the results from the automated analyzer Sysmex R-1000. Duplicate samples from 330 patients were analyzed. The reference intervals obtained with the three methods were: MAXM 0.37-1.80%, median 0.83%; manual 0.40-2.30%, median 1.00%; R-1000 0.60-1.95%, median 1.06%. The imprecision (CV) at all concentrations is lower than the microscopic method (low 16.1% vs 67%; normal 16.9% vs 28.9%; high 9.5% vs 13.0%). The MAXM shows a good overall correlation with the microscopic method (intercept 0.01, slope 0.89, r2 = 0.87) despite evidence of a significant overestimation at low concentrations [difference (D) 0.30] and a moderate underestimation at normal (D -0.15) and high (D -1.04) concentrations; the same behavior is shown in comparison with results from the R-1000, with an overall underestimation by MAXM (D -0.26). When compared with the microscopic method, MAXM shows a modest sensitivity at low reticulocyte counts (54.8%) and satisfactory sensitivity for high counts (87.3%), with an overall agreement of 81.3%.
