A multi-center research on the establishment and validation of autoverification rules for urine analysis
张慧,王学军,樊爱琳,郝晓柯,刘家云,王刚强,李莉,李秋晨,张如霖,蒋黎,熊志刚,何贝,洪国粦,赖小华,蔡梅真,高辉,李红燕,程云,陈颖洁,周勇,杨倩文,单洪丽,宁书晗,黄晶,续薇
DOI: https://doi.org/10.3760/cma.j.cn114452-20200327-00313
2020-01-01
Abstract: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.