Abstract:

Objective　Based on the Westgard Sigma rules, the quality control data of 12 chemiluminescence items were analyzed with 3 different biases, the method performance was evaluated, and the best personalized quality control plan was established through this exploratory study. Methods　To determine the total clinical allowed error (TEa) based on the analytical quality requirements of the US Clinical Laboratory Improvement Amendments (CLIA) Capacity Verification Program, the cumulative coefficient of variation (CV) of internal quality control from January to December in 2019 was used to reflect the imprecision of the method, and the bias of clinical laboratory center of the Ministry of Health in 2019, the bias accumulated by Bio-Rad peer group in internal quality control, and the biases of 12 inter laboratory comparison data of Bole External Quality Assurance Services (EQAS) in the whole year were used to reflect the inaccuracy of the method, so we cold calculate the σ value and quality goal index (QGI); the σ method performance verification chart was used to evaluate the performance of 12 chemiluminescence quality control items, and the optimal quality control scheme was obtained according to three different biases. Results　Among the 12 items tested, the σ value obtained by using 3 kinds of biases; the numbers of items with σ≥6 (world-class) of the 3 kinds of biases were 22, accounting for 61.11% (22/36), with no statistical difference (P<0.05); the Ministry of Health's inter room quality assessment had 3 items (8.33%, 3/36) of biases whose 2≤σ<3 (unacceptable), the indoor quality control had 1 (2.78%, 1/36), and the Bole EQAS had 1 (2.78%, 1/36), with statistical differences between the latter two and the former one (both P<0.05). According to the lowest level of σ value and QGI among the three kinds of biases as the criteria, the σ values of free triiodothyronine (FT3), thyroid stimulating hormone (TSH), follicle stimulating hormone (FSH), luteinizing hormone (LH), prolactin (PRL), testosterone (TSTO), and estradiol (E2) were over 6 σ, and only the 1_3s rule needed to be used; total triiodothyronine (TT3), free tetraiodothyronine (FT4), and insulin (INS) needed to be improved for precision, and used the 1_3s2_2sR_4s4_1s8_x rule; total thyroxine (TT4) and progesterone (PRG) used the 1_3s2_2sR_4s rule. Conclusions　The application of different biases can more comprehensively help laboratories choose the best and most appropriate quality control plan; Westgard Sigma rules are practical management tools for laboratory quality control, improve the efficiency of quality control, and make continuous quality improvement.

Key words: Biases, Westgard Sigma rules, σ method performance verification diagram

摘要： 目的　基于Westgard西格玛规则，应用3种不同偏倚对12个化学发光项目的质量控制数据分析，评价其方法性能，通过此探索性研究建立最优的个性化质量控制方案。方法　依据美国临床检验改进修正计划（Clinical Laboratory Improvement Amendments，CLIA）能力验证计划的分析质量要求确定临床允许总误差（allowed total errors，TEa））以2019年1月至12月室内质量控制的累积变异系数（coefficient variation，CV）反映方法的不精密度，应用卫生部临床检验中心2019年室间质评的偏倚、室内质量控制Bio-Rad对等组积累的偏倚和伯乐国际室间质量评价服务（external quality assessment service，EQAS）全年12次的室间比对数据的偏倚反映方法的不准确度，计算出σ度量值和质量目标指数（quality goal index，QGI），并绘制σ方法性能验证图，评价12个化学发光质量控制项目的检验性能，以及根据3种不同偏倚得到最优的质量控制方案。结果　测定的12个项目中，应用3种偏倚所得的σ值，σ≥6（世界一流）项目的个数，3种偏倚均为22个，占61.11%（22/36），差异无统计学意义（P<0.05）；2≤σ<3（不可接受水平）项目个数分别为卫生部室间质评偏倚3个［8.33%（3/36）］、室内质量控制偏倚1个［2.78%（1/36）］和伯乐EQAS偏倚1个［2.78%（1/36）］，后两类偏倚与卫生部室间质评偏倚相比，差异均有统计学意义（均P<0.05）。根据3种偏倚中最低水平的σ值和QGI作为判断标准，游离三碘甲状腺原氨酸（free triiodothyronine，FT3）、促甲状腺激素（thyroid stimulating hormone，TSH）、卵泡刺激素（follicle stimulating hormone，FSH）、促黄体生成素（luteinizing hormone，LH）、催乳素（prolactin，PRL）、睾酮（testosterone，TSTO）和雌二醇（estradiol，E2）的σ值均超过6σ水平，只需要使用1_3s规则；总三碘甲状腺原氨酸（total triiodothyronine， TT3）、游离四碘甲状腺原氨酸（free tetraiodothyronine，FT4）和胰岛素（insulin，INS）需要改进精密度，使用1_3s2_2sR_4s4_1s8_x规则；总甲状腺素（total thyroxine，TT4）和孕酮（progesterone，PRG）使用1_3s2_2sR_4s规则。结论　应用不同偏倚可以更全面地帮助实验室选择最优的质量控制方案；Westgard西格玛规则是实验室开展质量控制实用有效的管理工具，提高质量控制效率，进行质量的持续改进。

关键词: 偏倚, Westgard西格玛规则, σ方法性能验证图