京津冀地区110家检验结果互认医疗机构糖化血红蛋白的室间质量评价和室内质量控制结果分析

External quality assessment and internal quality control analysis of glycated hemoglobin A1c in laboratories of 110 medical institutions with mutual recognition agreement in Beijing-Tianjin-Hebeiregion

目的通过分析京津冀地区110家检验结果互认医疗机构糖化血红蛋白（HbA1c）的室间质量评价（EQA）和室内质量控制（IQC）结果，评价互认实验室HbA1c检测结果的可比性。方法将5个批号HbA1c质控品通过邮政特快专递分发到参评实验室，实验室在规定时间内对样品进行检测后，通过网络上报检测结果，同时提交2016年1月的IQC数据。对所有参加本次EQA计划的互认实验室不分组，分别采用传统统计方法和简单稳健统计方法对EQA结果进行评价；对上报IQC数据的互认实验室，按照检测系统分组，评价当月和累积在控变异系数（cv）满足《糖化血红蛋白检测》（WS／T461．20151中推荐的室内精密度标准的情况。结果132家互认实验室有110家参加了本次HbA1cEQA计划，95家回报了IQC数据。EQA结果：传统统计方法显示，HbA，。EQA实验室合格率为90．9％（100／110）；简单稳健统计显示，对于5个批号质控品，简单稳健z比分数绝对值（1z1）≤2的实验室所占比例均高于92．3％（107／110），lzl〉／3的实验室不到4．0％，出现两次或以上lzl／〉3的实验室所占比例为3．6％（4／110）。4家实验室两种统计方法均显示EQA成绩不合格。IQC结果：只有57家（60．0％）实验室检测了两个浓度水平质控品；分别以CV〈2％和CV〈3％为评价标准，总体cv通过率分别为51．2％-66．2％和77．4％-85．3％。结论参与本次研究的实验室仍有少部分不能满足互认要求，其中EQA结果较为理想，IQC性能还有待提高，建议督促未达到标准的实验室整改甚至取消检验结果互认资格。

Objective To evaluate the comparability of glycated hemoglobin A1c （HbA1c） testing results among mutual recognition laboratories by analyzing the results of external quality assessment （EQA） and internal quality control （IQC） of 110 mutual recognition laboratories in Beijing-Tianjin-Hebei region. Methods Five lots of EQA materials were distributed to the participant laboratories by express mail services （EMS）. Laboratories should test the samples within a specific time, and submit results via web-based EQA software. Meanwhile, IQC data in January 2016 should also be reported though web. Then the EQA results without grouping were evaluated by traditional statistical method and simple robust statistical method. Finally, the current coefficient of variation （CV） and cumulative CV of different groups divided by different analytical systems were evaluated against imprecision criteria for IQC recommended by Measurement of Hemoglobin A1c （WS/T461-2015）. Results Total of 110 mutual recognition laboratories participated in this EQA scheme and 95 laboratories submitted IQC information. As for EQA results, the percentage of laboratories of which the EQA results were acceptable was 90.9% （100/110） by use of traditional statistical method. When simple robust statistical method were applied, more than 92.3% （107/110） of the laboratories got a simple robust statistical Z score ｜Z｜ ≤2; less than 4.0% of the laboratories got ｜Z ｜ ≥3; only 4（3.64%） laboratories got a ｜Z ｜≥3 in two or more lots. EQA results of 4 （3.6%） laboratories were unacceptable according to the two kinds of statistical methods. As for IQC data, only 57（60.0%） laboratories tested IQC material of two concentration levels. If the criteria of CV〈2% and CV〈3% were employed, the range of overall pass rates was 51.2%-66.2% and 77.4%-85.3%, respectively. Conclusions Some laboratories still cannot meet the requirements for mutual recognition, the EQA results are satisfactory, but IQC performance should be improved. It is suggested that laboratories with unacceptable performance should take corrective action or their qualifications for mutual recognition could be canceled.