Performance of Automatic Blood Cell Analysis Parameters for Differential Diagnosis of Myelodysplastic Syndromes and Aplastic Anemia / 现代检验医学杂志

ABSTRACT

Objective To evaluate the performance of automated blood cell analysis parameters for differential diagnosis of myelodysplastic syndromes (MDS) and aplastic anemia (AA).Methods Data of automatic blood cell analysis parameters at diagnosis of confirmed patients with MDS and AA from December 2002 to February 2011 in Peking University Shenzhen Hospital were retrospectively reviewed.Results 33 cases of MDS and 36 cases of AA were recruited in this study.Based on the evaluable data,mean corpuscular hemoglobin concentration (MCHC) (328.58 ± 17.24 g/L vs 342.47±18.75 g/L,n=33/36) was significantly lower (P=0.002 1),while monocyte percentage (MONO％) (11.48±9.99 vs 6.94±2.50,n=32/34),platelet distribution width (PDW％) (13.51±4.24 vs 10.62±3.68,n=20/22) and platelet hematocrit (PCT％)(0.11 ±0.10 vs 0.04±0.07,n=11/15) were markedly higer (all P＜0.05) in patients with MDS than that of AA.No significantly differences for other blood cell analysis parameters were seen between patients with MDS and AA.Under the condition of best cut-off value,areas under the ROC curve of MCHC,MONO％,PDW and PCT were 0.706 (95％ confidence interval0.584～0.809),0.666 (0.540～0.778),0.668 (0.506～0.805) and 0.745 (0.538～0.894) respectively.MONO％ and MCHC had high specificities (97.06％ and 88.89％) and positive predictive values (93.3％ and 80.0％) for differential diagnosis of MDS from AA.Conclusion MONO％ and MCHC may be used as simple indicators for differential diagnosis of MDS and AA.