Laboratory evaluation of the Miles H.3 automated reticulocyte counter. A comparative study with manual reference...

To evaluate the performance of the new commercial Miles H.3 RTX analyzer in counting reticulocytes.

The results from the counter were compared to those obtained from microscopic methods, following the National Committee for Clinical Laboratory Standards H44-P guidelines, and to the results from the Sysmex R-1000 counter. In total, 279 samples were analyzed in duplicate with each of the three methods. One hundred thirty-three samples were from healthy subjects, while 146 were from patients with various pathologies, 10 of whom presented with posttherapeutic aplasia of the bone marrow and 9 with iron-deficiency anemia.

The reference intervals for the normal controls are different for each of the three methods (manual: 0.35-2.35%, 16 to 116 x 10(9)/L; Miles H.3: 0.65-2.30%, 35.1 to 112.0 x 10(9)/L; Sysmex R-1000: 0.6-1.85%, 28.0 to 85.0 x 10(9)/L). The overall imprecision was lower for the instruments than for the microscopic method (Miles H.3: coefficient of variation, 11.6%; R-1000: coefficient of variation, 4.2%; microscopic method: coefficient of variation, 24.2%). The Miles H.3 shows a good correlation with the other methods, yet it overestimated the low values with respect to both the microscopic method (intercept, 0.55; slope, 0.70) and the R-1000 (intercept, 0.44; slope, 0.78). This became particularly pronounced in patients with marrow aplasia.

Miles H.3 can produce results with an acceptable degree of accuracy. The agreement with the dedicated fluorescence-based flow cytometer R-1000 at normal and high concentrations is also good. The possibility of providing reticulocyte indices as well as erythrocyte indices (mean volume, mean hemoglobin content, mean hemoglobin concentration) and the relative dispersion indices could be useful in understanding red cell pathophysiology in normal and iron deficient patients.