血细胞分析全国质评参加实验室的室内质量控制数据分析

Analysis of internal quality control data of complete blood count from laboratories participating in national external quality assessment

目的了解我国临床实验室血细胞分析室内质量控制开展情况和存在问题，促进室内质量控制的规范开展。方法收集2012至2017年参加全国血细胞分析室间质量评价实验室5项参数的室内质量控制数据，包括白细胞计数（WBC）、红细胞计数（RBC）、血红蛋白（Hb）、红细胞比容（Hct）和血小板计数（PLT），共计12次。经有效性确认后，对不同批次回报的实验室数据进行分析，包括不同浓度水平质控物的使用比例；分析12次实验室回报变异系数（CV）的第25百分位数（P25）、第50百分位数（P50）、第75百分位数（P75）、第90百分位数（P90）随时间的变化趋势；比较同时使用3个浓度水平质控物的实验室检测CV的差异；将2017年实验室回报CV与基于生物学变异、卫生行业标准和德国医学会的精密度要求进行比较；统计质控规则的使用情况。 结果2012至2017年的12次质评活动收集到的实验室（最多2 402家、最少1 449家）室内质量控制数据，剔除无效数据后，将剩余实验室（最多2 332家、最少1 431家，占96.0%～99.2%）的信息用于数据分析。分别有61.9%～66.1%、18.2%～ 23.6%和14.3%～17.3%的实验室使用1个、2个和3个浓度水平的质控物，使用2个以上浓度水平质控物的实验室由33.9%增加至38.1%。WBC、RBC、Hb和Hct低、中、高值与PLT中值室内质控P75和P90的CV以及PLT低值室内质控P90的CV随时间变化的下降趋势差异均有统计学意义（均P〈0.05）；PLT低值室内质控P75的CV、PLT高值室内质控P75和P90的CV下降趋势差异无统计学意义。WBC和PLT低值与中值、低值与高值室内质控CV差异有统计学意义，但中值与高值室内质控CV差异无统计学意义，RBC、Hb和Hct不同浓度水平室内质控CV差异均无统计学意义。除Hct低值和中值、PLT低值室内质控CV外，其他参数85%实验室的CV可满足基于生物学变异的最低要求；超过85%实验室各参数的CV可达到行业标准的要求；除PLT低值室内质控CV外，80%实验室其他参数的CV可达到德国医学会的要求。使用含13s/22s质控规则的实验室比例由59.2%上升至76.0%。结论6年来，实验室室内质量控制CV随时间变化总体呈下降趋势，但部分实验室使用的质控物浓度水平和质控规则未满足管理要求，少数实验室Hct和PLT的室内质控CV未达到我国行业标准规定的最低要求，需进一步实施质量改进。

Objective To investigate current status and problems of internal quality control （IQC） of complete blood count in China so as to perform IQC normally.Methods The IQC data of complete blood count for five parameters were collected from laboratories participating in national external quality assessment during 2012-2017 （totally 12 times）, including WBC, RBC, Hb, Hct and PLT. After confirmation of all data, data for the 12 times were analyzed as follows.The proportions of using different levels of quality control materials were calculated.The 25th, 50th, 75th, 90th percentiles CV of data collected for the 12 times were calculated respectively and the trends of CV were observed over time.The difference of CV among laboratories running three control levels was compared.The CV of each parameter in 2017 was compared with precision requirements based on biological variation, health standards and German Medical Association Directive; The proportions of laboratories using different control rules were calculated.Results After invalid data was excluded from those IQC data of laboratories for the 12 times external quality assessment （up to 2 402, as low as 1 449） from 2012 to 2017, the residual data （up to 2 332, as low as 1 431, accounting for 96.0%-99.2%） was used for analysis. 61.9%-66.1%, 18.2%-23.6% and 14.3%-17.3% of laboratories ran one, two and three control levels respectively, and the proportions of laboratories running more than two control levels increased from 33.9% to 38.1%. The decrease trend of the 75th, 90th percentiles CV of WBC, RBC, Hb, Hct for three levels, PLT for normal level and the 90th percentiles CV of PLT low level had statistically significance over time （P〈0.05）; the decrease trend of the 75th percentiles CV of PLT low level and 75th, 90th percentiles CV of PLT high level had no statistically significance over time. The CV had significant difference between low and normal, low and high control level for WBC and PLT, while there were no difference between normal and high control levels. There were no significant difference of CV among three control levels for RBC, Hb, and Hct. Except for the CV of Hct low, normal level and PLT low level, 85% of laboratories for the other parameters could meet the minimum precision requirements based on biological variation; more than 85% laboratories met the requirements of health standards; except for the CV of PLT low level, more than 80% laboratories met the requirements of German Medical Association Directive. The proportion of laboratories using 13s/22s quality control rules increased from 59.2% to 76.0%.Conclusions During the past 6 years, the CV for IQC has shown a decrease trend over time. However, the control level and quality control rules used by some laboratories do not meet management requirements. The CV of Hct and PLT in a few laboratories do not meet the minimum requirements of the health standards, and need to implement quality improvements fatherly.