甲状腺功能五项检验报告自动审核系统的建立与优化  

Establishment and optimization of an autoverification system for thyroid function reports

作　　者：崔俊华 朱晶 邵文琦 杨静 潘柏申 王蓓丽 郭玮 Cui Junhua;Zhu Jing;Shao Wenqi;Yang Jing;Pan Baishen;Wang Beili;Guo Wei(Department of Laboratory Medicine,Zhongshan Hospital,Fudan University,Shanghai 200032,China)

机构地区：[1]复旦大学附属中山医院检验科,上海200032

出　　处：《中华检验医学杂志》2025年第2期207-213,共7页Chinese Journal of Laboratory Medicine

基　　金：国家重点研发计划(2022YFC3602302,2022YFC3602301);国家自然科学基金(82172348);上海市临床重点专科建设项目(shslczdzk03302)。

摘　　要：目的建立并利用检验历史数据优化甲状腺功能5项检验报告自动审核系统。方法依据我国《临床实验室定量检测结果自动审核》文件和美国临床和实验室标准化协会AUTO 10-A和AUTO 15指南,结合人工审核经验建立甲状腺功能5项检验报告自动审核系统。收集2021年193860份甲状腺功能5项报告用于原有系统的评估。收集2022年210097份甲状腺功能5项报告和2023年的299198份甲状腺功能5项报告用于自动审核系统的优化。使用2024年上半年160666份甲状腺功能5项报告进行优化后系统的人机比对验证。结果2021年基于原有甲状腺功能5项自动审核系统的通过率为69.56%(134849/193860)。利用2022年和2023年历史数据优化后的系统涵盖了21种模式规则,并根据不同模式建立对应的审核范围规则、差值校验规则和复检规则。2024年上半年数据以人工审核为标准,优化后系统的敏感度和特异度分别为100%(499/499)和81.57%(130646/160167),假阴性率为0,自动审核与人工审核的符合率为81.63%(131145/160666),通过率为81.32%(130646/160666)。结论建立并利用检验历史数据优化甲状腺功能5项自动审核系统,根据不同模式制定其对应的审核规则能应用于临床实践工作,保证检测报告准确性的同时进一步提高工作效率,使系统不断优化和完善。Objective To establish and optimize an autoverification system for thyroid function test reports of 5 items using historical test data.MethodsBased on the docoment'Autoverification of Clinical Laboratory Quantitative Test Results′,CLSI AUTO 10-A and AUTO 15 guidelines,an autoverification system for thyroid function test reports of 5 items was established combining with manual verification experience.A total of 193860 thyroid function test reports of 5 items in 2021 were collected for the assessment of the original system.Totally 210097 thyroid function test reports of 5 items in 2022 and 299198 reports in 2023 were collected for the optimization of the autoverification system.There were 160666 thyroid function test reports of 5 items from the first half of 2024 for the manual and autoverification comparison after optimization.ResultsThe pass rate of the autoverification system based on original thyroid function report in 2021 was 69.56%(134849/193860).The optimized system utilizing historical data from 2022 and 2023 covered 21 pattern rules and established verification for different patterns including range rules,delta check rules,and review rules.Taking manual verification as the standard for the data from the first half of 2024,the sensitivity and specificity of the optimized system were 100%(499/499)and 81.57%(130646/160167),respectively,with a false-negative rate of 0.The concordance rate between autoverification and manual verification was 81.63%(131145/160666),and the pass rate was 81.32%(130646/160666).ConclusionEstablishing and optimizing the autoverification system for thyroid function tests of 5 items using historical test data,and formulating verification rules for different patterns can be applied to clinical practise,which not only ensures the accuracy of test reports but also improves work efficiency,allowing continuously optimized and perfected of the system.

关 键 词：甲状腺功能试验 自动审核系统 临床实验室技术 

分 类 号：R581[医药卫生—内分泌] R446[医药卫生—内科学]