目的:建立和验证尿液分析检验结果的自动审核规则。方法:2019年12月至2020年3月收集来自吉林大学第一医院、空军军医大学西京医院、上海市第一人民医院、四川省人民医院、厦门大学附属第一医院、昆明市延安医院、山东大学第二医院共7家医院检验科的尿液分析常规样本共7 405份：其中用于规则建立组5 246份，规则验证组2 159份。应用Sysmex UC-3500尿液全自动干化学分析仪以及Sysmex UF-5000尿液有形成分分析仪检测尿液标本，同时按照第四版操作规程中尿沉渣镜检标准流程进行尿液样本的离心、镜检，将临床信息、仪器参数结果、仪器报警信息、仪器提示信息和镜检结果汇总，利用统计学方法筛选自动审核条件，将自动审核规则录入仪器软件，分别计算真阴性率（有效审核）、假阴性率（无效审核）、真阳性率（有效拦截）、假阳性率（无效拦截）、镜检率、自动审核通过率和通过正确率。应用自动审核规则模拟审核7家医院1个月内的尿液分析常规检测结果，比较自动审核规则应用前后TAT变化。结果:（1）通过优化组合，选定了用于自动审核的拦截规则16条，拦截后的关联处理需镜检的有8条，需人工审核的有8条。（2）自动审核规则建立组中5 246份样本其假阴性率为2.4%（128/5 246），镜检率16.5%（864/5 246），自动审核通过率为60.4%（3 169/5 246），通过正确率为72.1%（3 780/5 246）。（3）规则验证组中2 159份样本其假阴性率为3.3%（71/2 159），镜检率19.6%（423/2 159），自动审核通过率为62.9%（1 357/2 159），通过正确率为75.5%（1 630/2 159）。（4）应用尿液自动审核规则分别对7家医院一个月内的标本进行TAT模拟分析，结果显示TAT明显缩短，下降率分别为45.30%、52.85%、43.13%、38.44%、41.34%、51.77%和30.2%。结论:成功建立了尿液分析检验结果的自动审核规则，并通过了临床标本验证。“,”Objective:To establish and verify the autoverification rules for urine analysis.Methods:To establish the autoverification rules. From december 2019 to March 2020, 7 405 routine urine samples were collected from the laboratory of First Hospital of Jilin University, the laboratory of Xijing Hospital of Air Force Military Medical University, the laboratory of Shanghai First People′s Hospital, the laboratory of Sichuan people′s Hospital, the laboratory of the First Affiliated Hospital of Xiamen University,the laboratory of Yan′an Hospital of Kunming City, the laboratory of the second hospital of Shandong University. Among them, 5 246 cases were used for rule establishment group and 2 159 cases for rule verification group. Sysmex fully automated urine chemistry analyzer UC-3500 and urine sediment analyzer UF-5000 were used to test the urine sample. Urine sediment smears were done following the standard procedure of centrifugation and microscopy mentioned in National Guide to Clinical Laboratory Procedures (4th edition).Clinical information, instrument parameter results were summarized, instrument alarm information, instrument clinical prompt information and microscopy results. Statistical methods are used to screen autoverification rules. Autoverification rules were input into software. True negative rate(valid verification rate), false negative rate(invalid verification rate), true positive rate(valid interception rate), false positive rate(invalid interception rate), microscopy rate, autoverification pass rate, and correct passing rate were calculated respectively. Autoverification rules were used to imitate and autoverificate the routine urine analysis results of 7 hospitals within one month, and the TAT changes before and after the application of these rules were compared.Results:(1)After optimization and combination, 16 autoverification rules were selected for interception, including 8 rules for artificial microscopic examination, and another 8 rules to be released manually.(2)Following these autoverification rules, the false negative rate was 2.4% (128/5 246) among 5 246 samples in the rule establishment group, the microscopic examination rate was 16.5% (864/5 246), and the autoverification pass rate was 60.4% (3 169/5 246),correct passing rate was 72.1%(3 780/5 246).(3)Verification of 2 159 cases in autoverification validation group, the false negative rate was 3.3% (71/2 159), the microscopic examination rate was 19.6% (423/2 159), and the autoverification pass rate was 62.9% (1 357/2 159) correct passing rate was 75.5%(1 630/2 159).(4)The samples from 7 hospitals within one month were simulation analyzed by using the autoverification rules, the results showed that TAT were significantly shortened, with a decrease rate of 45.30%, 52.85%, 43.13%, 38.44%, 41.34%, 51.77% and 30.2%, respectively.Conclusions:The autoverification rules for urinalysis results were established successfully, and they were verified by clinical samples.