Establishment and evaluation of the review criteria of automatic urine analysis workstations / 中华检验医学杂志

ABSTRACT

Objective To establish the proper review rules for the microscopic screening of urine samples tested by automatic urinalysis work station.Methods A total of 3 154 random urine samples were enrolled to establish and validate review rules .All the samples were collected from the inpatients and outpatients of Shanghai First People′s Hospital from March to May 2013 and tested by urinalysis work station.Three thousands one hundred and fifty four urine samples were firstly tested by urinalysis work station,including both urine dry chemical analyzer and urine sediments analyzer .Then each urine sample was examined microscopically by two technicians-in-charge using double-blind method.The average results from the two technicians were used as review results .Compared with review results ,the review rules were set up.According to different test methods by automatic urinalysis work station , four microscopic review protocols were defined(1)Protocol 1based on chemistry results only ,microscopy review was performed when any of WBC,RBC,PRO and NIT was positive;(2)Protocol 2based on urine sedimental analysis only ,microscopy review was performed when any of WBC ,RBC and CAST count was over upper limit of the reference range;(3)Protocol 3if any of BLD ,RBC,LEU,WBC was different between two systems ,or quantitative results had two or more than two gradient differences ,microscopy review was performed;(4) Protoco1 4if any of BLD, RBC,LEU ,WBC was different between two systems , or CAST was over upper limit of the reference range , or alarm appeared , microscopic review was performed .400 randomly selected urine samples were tested to validate the review rules .Omission diagnostic rate and review rate were used to evaluate the rules .Results According to the review rules,the positive samples rate was 43.47%(1 371/3 154) and the negative rate was 56.53%( 1 783/3 154 );Positive samples were composed of RBC ( 55.58%) , WBC ( 59.66%) and CAST(6.42%).The review rates of four protocols were 44.48%(1 403/3 154),45.69%(1 441/3 154), 26.09%(823/3 154),28.95%(913/3 154),respectively.The false negative rates (omission diagnostic rates)were 7.13%(225/3 154),4.53%(143/3 154),2.73%(86/3 154) and 1.02%(32/3 154), respectively .Protocol 4 was selected as an ideal plan.Additional 400 urine samples were tested using protocol 4 in order to confirm the review rule.The review rate, false negative rate were 26.25%(105/400), 0.75%( 3/400 ), respectively.After image review revised, the review rate was 14.50%(58/400).Conclusion This study formulates that the automatic urine analysis workstation review rules have clinical maneuverability and validity.