首页|患者数据指数加权移动均值法在甲状腺功能检测质量控制中的应用

患者数据指数加权移动均值法在甲状腺功能检测质量控制中的应用

扫码查看
目的 探讨患者数据实时质量控制(PBRTQ C)的指数加权移动均值法(EWMA)在甲状腺功能检测质量控制中的应用价值.方法 收集2022年12月1日至2023年4月30日山东大学第二医院住院及门诊患者的甲状腺功能检测结果进行数据的正态修正、参数设置、程序建立和实时运行.以2023年5月1日至8月31日所有进行甲状腺功能检测患者的结果作为验证数据集,统计患者甲状腺功能检测结果的EWMA估计值以及4个月的累积变异系数(CV),将累积CV与精密度质量目标(取1/3TEa)和室内质量控制(IQC)的CV进行比较,采用Westgard 2-2s、1-3s规则进行警报设定,通过现场质量记录和仪器信息验证来监控EWMA实时控制图在甲状腺功能相关项目检测中的预警情况,并分析预警原因.所有数据采用DxLab Mind软件进行正态分布统计,对检测结果进行Kolmogorov-Smirnov检验.结果 患者血清甲状腺功能相关项目均呈正偏态分布,通过Box-Cox方法对数据进行修正后,游离三碘甲状腺原氨酸(FT3)、游离甲状腺素(FT4)的PBRTQC数据呈正态分布,其4个月内EWMA的累积CV分别为6.26%和2.86%,均小于该项目的精密度质量目标8.33%.但促甲状腺激素(TSH)、甲状腺过氧化物酶抗体(TPOAb)和甲状腺抗球蛋白抗体(TgAb)的数据经修正后仍呈正偏态分布,EWMA累积CV分别为13.16%、15.31%和16.77%,分别大于TSH、TPOAb和TgAb的精密度质量目标8.33%、10%和10%.通过EWMA实时控制图可发现不同情况的失控报警,包括由于样本来源集中导致的FT4假报警以及试剂性能变化引起TSH结果偏移出现的失控报警.此外,EWMA实时控制图也能够发现不同仪器之间FT3结果的差异.结论 基于PBRTQC专业智能软件工具建立的患者数据EWMA程序可实时、连续监控检测系统的患者数据,动态识别和监测分析过程中造成的误差并进行智能预警.可作为甲状腺功能项目,特别是FT3、FT4日常质控品质量控制的有益补充,具有较好的应用价值.
Application of patient-based real-time quality control using exponentially weighted moving average method on quality control procedures of thyroid function tests
Objective To investigate the application of patient-based real-time quality control(PBRTQC)using exponentially weighted moving average(EWMA)method in internal quality control(IQC)procedures of thyroid function tests.Methods The serum thyroid function test results of outpatients and inpatients in the Second Hospital of Shandong University from December 1,2022 to April 30,2023 were collected.Based on the PBRTQC professional intelligent software system,normality correction,parameter setting,program preparation and real-time operation of test data were carried out.The results of all patients who underwent thyroid function testing between May 1,2023 and August 31,2023 were used as the validation dataset.The estimated EWMA value of thyroid function test results and the cumulative coefficient of variation(CV)over 4 months were calculated.The cumulative CV was compared with the criteria of precision quality standard(1/3TEa)and the CV of IQC.Westgard 2-2s and 1-3s rules were used for alarm setting.The early warning information of the EWMA quality control program were recorded and the potential causes of performance changes were analyzed.DxLab Mind software was used to conduct normal distribution statistics for all data,and the Kolmogorov-Smirnov test was performed on the test results.Results The items related to serum thyroid function of the patients were all positively skewed.After data correction by Box-Cox method,the PBRTQC data of free triiodothyronine(FT3)and free thyroxine(FT4)were normally distributed,and their cumulative precisions(CV)of EWMA within 4 months were 6.26%and 2.86%,respectively,both of which were lower than the precision quality target of 8.33%.However,the data of thyroid-stimulating hormone(TSH),thyroid peroxidase antibodies(TPOAb)and thyroglobulin antibody(TgAb)were still positive skewed after modification.The EWMA cumulative CV of TSH,TPOAb and TgAb were 13.16%,15.31%and 16.77%,which were higher than the precision quality targets of 8.33%,10%and 10%,respectively.The EWMA QC program can detect different out-of-control alarms,including FT4 false alarms due to sample source concentration and TSH result bias caused by changes in reagent performance.In addition,the EWMA QC program can also detect differences in FT3 results between different DXI800 fully automated chemiluminescence instrument instruments.Conclusions The EWMA program based on PBRTQC professional intelligent software tools can monitor the patient data of the detection system in real time and continuously,dynamically identify and monitor the errors generated during the analysis process and give early warning.It can be used as a useful supplement for the daily IQC of thyroid function items,especially FT3 and FT4,and has good clinical application value.

Quality controlPatient-based real-time quality controlExponentially weighted moving average

赵晓晴、牛爱军、王小玲、任灵杰、毛海婷

展开 >

山东大学第二医院检验医学中心,济南 250033

贝克曼库尔特商贸(中国)有限公司,上海 200335

质量控制 患者数据实时质量控制 指数加权移动均值法

2024

中华检验医学杂志
中华医学会

中华检验医学杂志

CSTPCD北大核心
影响因子:1.402
ISSN:1009-9158
年,卷(期):2024.47(5)