A total of 2 243 fresh urine samples (within 1h) from outpatients and inpatients were analyzed by traditional manual microscopy as a standard method, and the system was evaluated. The qualitative comparison between EH-2080B and manual microscopy results was evaluated by false positive (negative) rate.The 100 fresh urine samples with positive red blood cell (RBC)and white blood cell (WBC)were selected to do quantitative comparison of RBC and WBC. While low power single view of liquid filling static time period and visual persistence of low power and high power photographed conversion instant view were used to observe, living sperms and trichomonas vaginalis' dynamic observation messages by the artificial screen browsing were useful to evaluate the function extension of the instrument.

The within-run and between-run precisions of EH-2080B met the requirements [WBC<10%,RBC<10%, epithelial cell (EC)<30%]. The correlation with manual microscopy was good [correlation coefficient (r) for RBC=0.973,P=0.000; r for WBC=0.990, P=0.000; r for EC=0.959, P=0.000]. WBC, RBC, fungus and so on were of a certain degree of false positive (negative) rate, but the comprehensive analysis of combinating dry chemistry information, photo information and automatic urinary sediment analysis was reliable. The coincidence rate by random sampling was more than 85%, and the RBC and WBC quantitative comparison difference from fresh urine samples with RBC and WBC positivities had no statistical significance (P>0.05).The artificial browsing could be added to capture the dynamic information or suspicious dynamic information of living sperms and Trichomonas vaginalis'dynamic observation.

The results of EH-2080B and manual microscopy are consistent. EH-2080B has some recognition analysis for urine formed elements, but for the actual work or some special specimens, manual microscopy identification is needed. The complementary effect of different detection methods can improve the work efficiency and improve the detection quality.