Abstract
The Coulter LH750® is a last generation automated hematology analyzer. In 1999 we evaluated the optimization of immature granulocytes/bands (IG/B) detection in the Coulter GEN*S (Blood Vol.94 n° 10 sup.1, 1999 page 45 b), in which we observed that the Mean Volume of the Neutrophil population (M NE V) and the standard deviation of the neutrophil volume (SD NE V) were the most useful parameters to improve the detection of the presence of IG/B. Recently the utility of SD NE V has been proposed in the diagnosis of septic patients.(ISLH 05). The purpose of this study is to confirm that the NE morphometric parameters in the LH750 have the same values and cut-off as the GEN*S to discriminate the cases with lefts shift and if it is possible to create rules/flags that use these parameters to allow the automated detection of IG/B with cut-off levels specific to the origin (emergency room, outpatients, intensive care etc.) of the patient sample.
In addition we investigated if the parameters are useful when taking into account not only the % of IG/B, but also their absolute number as recommended by the NCCLS. The overall intent being to optimize the efficiency of our laboratory in the detection of IG/B by decreasing the number of smears, but without increasing the number of false negatives. We have selected from the routine samples of a 2000 beds hospital, 340 consecutive cases on the LH750® that triggered the bands /Immature granulocytes flags, Immature NE 1 (NE1) or Immature NE 2 (NE2). We considered the samples with more than 0,9 IG/B (103/mL) as positive. Blood smears were performed on all the positive samples (340) and a WBC differential was performed according the NCCLS protocol (H 20A). The percentage and absolute number of IG/B were collected. The statistical analysis was performed using Medcalc® version 8.1.0.0. A ROC Curve and the cut-off for the detection of positive samples with immature IG/B using the SD NE V and M NE V was calculated.
RESULTS: For the flag NE1 the best parameter to select the postive cases with left shift was the SD NE V with a cut-off of >24,2. (although this flag is actually defined for a moderate left shift). For the flag NE2 the best parameter to select the postive cases with left shift was the M NE V with a cut-off of > 161.
CONCLUSION: Before the study performed in 1999 we validated the Imm NE1 and Imm NE2 flags according to the NE% and NE# or the WBC values, that statistically demonstrated limited utility. Parameters which permitted to optimize the detection of IG/B, were the Neutrophile morphological parameters; M NE V and SD NE V. The values of the LH 750 are slightly different to those of the GEN*S. We will subsequently study their utility and advantages compared to the IG/B counting in the detection of bacterial infections and sepsis due to better reproducibility.
Author notes
Corresponding author
