Finn Laura S, Hall Janice, Xu Min, Rutledge Joe C
The Department of Pathology, The University of Washington, Children's Hospital and Regional Medical Center, Seattle, Washington 98105, USA.
Lab Hematol. 2004;10(2):112-8. doi: 10.1532/LH96.04027.
Manual differential white blood cell (WBC) counting has been considered the gold standard and is routinely used to validate differentials obtained with other methodologies. To validate the accuracy of automated lymphocyte counts, we compared 2-part differentials using a Coulter HmX hematology analyzer and a Coulter XL flow cytometer to analyze 57 pediatric samples with WBC counts ranging from 0.7 k to 33.4 k. These data were compared with manual differential counts. We found excellent correlation between the two automated lymphocyte and monocyte counting methods that surpassed the manual correlations, indicating manual lymphocyte or monocyte counts are unnecessary in the setting of quality scatterplots. To evaluate the use of automated differentials for our most labor-intensive cases (low WBCs, which frequently require manual differentials) we then compared 3-part differentials using the HmX hematology analyzer and flow cytometer for 51 samples with total WBC < or = 1.1 x 10(9)/L. Manual differentials (< or = 100-cell counts) were available on 27 samples. Although the correlations for manual versus automated or flow differentials were good for all cell types, the correlation between the two automated methods was better, irrespective of the hematology analyzer scatterplot quality. Preliminary data provide additional evidence that automated differentials in samples with WBC of < or = 1.1 x 10(9)/L are acceptable for reporting, thus saving technologist time and improving patient care by decreasing the resulting turnaround time. These studies suggest that comparison with a standardized procedure like flow cytometry would be a better method for validation of automated differentials than comparison with the less precise, more laborious manual differential.
手工白细胞分类计数一直被视为金标准，并且经常用于验证通过其他方法获得的分类结果。为了验证自动淋巴细胞计数的准确性，我们使用库尔特HmX血液分析仪和库尔特XL流式细胞仪对57份白细胞计数范围为0.7×10³至33.4×10³的儿科样本进行了两分类分析，并将这些数据与手工分类计数结果进行比较。我们发现两种自动淋巴细胞和单核细胞计数方法之间具有出色的相关性，超过了与手工方法的相关性，这表明在质量散点图良好的情况下，无需进行手工淋巴细胞或单核细胞计数。为了评估自动分类在我们最耗费人力的病例（白细胞计数低，通常需要手工分类）中的应用，我们随后使用HmX血液分析仪和流式细胞仪对51份白细胞总数≤1.1×10⁹/L的样本进行了三分类分析。27份样本有手工分类结果（≤100个细胞计数）。尽管手工分类与自动分类或流式分类在所有细胞类型上的相关性都很好，但两种自动方法之间的相关性更好，无论血液分析仪散点图质量如何。初步数据提供了更多证据，表明白细胞计数≤1.1×10⁹/L的样本中的自动分类结果可用于报告，从而节省技术人员时间，并通过缩短周转时间改善患者护理。这些研究表明，与像流式细胞术这样的标准化程序进行比较，将是验证自动分类结果比与不太精确、更耗费人力的手工分类进行比较更好的方法。
