上海市2011年电子图谱细胞形态现场调查情况分析

摘要

Objective In order to evaluate the applicability of virtual microscopy of cell morphology,Shanghai Center for Clinical Laboratory had investigated the quality of cell morphology by virtual microscopy in hospitals above Level 2 Grade A and some community hospitals from October to November in 2011.Methods A total of 15 cases from the established cell library were selected.The difficult degrees of these cases were similar.These cases were given by the World Health Organization haematology center.The double blind marshaling method was used,and the ability of blood cell morphological recognition of these hospitals were evaluated by scene investigation.Results The satisfaction of cell image quality investigation reached 97.5％.The survey satisfaction was 96.3％.For white blood cell count,the results of virtual microscopy and the result given by World Health Organization in counting average value,standard deviation and bias ％ with the paired t test statistics were P =0.436 0,P =0.150 5 and P =0.180 6.There was no statistical significance(P ＞0.05).However,according to Rümke statistical count ranges with the reference given by World Health Organization,there was bias,and should be corrected.For more than 80％ of the inspection items,the cell recognition rate was ＞ 60％,which was better than those of previously used blood smear method.In addition to strengthening the recognition of cell morphology,the establishment of a characteristic cell count classification standard should be considered.Conclusions Due to the virtual microscopy with image quality and clear cell distinguishment,the method has the advantages of standardization,quality and reliability,and can be accepted for most cell morphology and is worth popularizing.%目的 为了评估电子图谱显微镜法细胞形态学方法的适用性,上海市临床检验中心血液体液室于2011年10月至11月间对上海市二级甲等以上医院,以及部分社区卫生服务中心进行了一次电子图谱的细胞形态学调查.方法 从实验室建立的细胞库中选取15例难易程度相似的病例图片配合相关病史资料(图片均经过世界卫生组织权威机构诊断,并且确定参考范围),按双盲法编组后,现场调查各个医院的细胞形态学识别能力.结果 经调查细胞图片质量的满意度达到97.5％,此种调查形式的满意度达到96.3％.白细胞分类计数方面,电子图谱显微镜法与世界卫生组织(WHO)给出的预期结果两者间在计数平均值、标准差和偏倚％3个方面的经配对t检验统计分别为P=0.436 0,P=0.150 5,P=0.180 6,差异均无统计学意义(P＞0.05).但按照Rümke方法统计的计数范围与WHO给出的参考值存在偏倚,需要修正.细胞特征识别方面,80％以上的考察项,细胞识别的符合率＞60％,优于以前采用血涂片法调查的结果.此外应在加强形态识别学习之外,还应考虑建立特征性细胞计数的分级标准.结论 由于电子图谱显微镜法所观察的图片质量佳、细胞分辨清晰,该方法具有易于标准化、规范化、质量可靠的优点,且能够被大多数细胞形态检验工作者接受,值得推广.