To increase the accuracy of routine blood test report by improving the auto-verifying system.

Three auto-verifying rules of blood routine examination were established. The rule A contains pure blood count results, the rule B contains pure blood count results and blood cells scatterplot abnormal alarm information, and the rule C contains blood count results, blood cells scatterplot abnormal alarm information and clinical preliminary diagnosis information. Ten thousand blood specimens were verified by the three self-verifying rules, respectively. The turn around time (TAT), false positive rates and false negative rates were compared.

There were no significant differences between the rules C and the rules A in TAT time (16.6 min vs 15.9 min, q=2.11, P>0.05). However, compared with rules A, the rules C lead to significantly reduction of the false positive rates (8.23% vs 5.96% χ²=4.89, P<0.05), and significantly reduction of the false negative rates (3.94% vs 1.91% χ²=5.68, P<0.05). Meanwhile, there were no significantly different false positive rates between the rules C and the rules B (8.23% vs 7.46%, χ²=0.41, P>0.05), as well as significantly different false negative rates (3.94% vs 2.89%, χ²=1.07, P>0.05). The audit consistency of the rules C was 94.62%, and was higher than that of the rules A (92.38%) and the rules B (93.19%). The improved rules C could detect parasites or malignant hematopathy in the blood but the hemogram of which was almost normal.

The improved auto-verifying system increased the accuracy of the test results and ensured the quality of the report.