Multicenter study of autoverification methods of hematology analysis

Abstract

This study was designed to establish and validate a set of autoverification methods for hematology analysis. One thousand and twenty-four samples were selected from Shanghai Ruijin Hospital and 999 from Beijing Hospital, China. False positive, false negative and autoverification pass rates were verified and the rules were then adjusted and confirmed according to the verification results. After confirmation, at least 10,000 sample cases were selected from Shanghai Ruijin Hospital, Beijing Hospital and China Armed Police General Hospital and checked automatically. The differences in the autoverification pass rate and average report delivery time before and after the application of the autoverification methods were compared between the three hospitals. Preliminary validation results showed that the false negative rates of the Shanghai Ruijin Hospital and Beijing Hospital were less than 2%. The false positive rates of these two hospitals were high, close to 18%. After rule adjustment, the false negative rate was basically the same as before adjustment, but the false positive rate declined obviously while the pass rate of autoverification improved significantly. The autoverification pass rates of the three hospitals were 76.4%, 85.1% and 84.2%, respectively. The turnover time (TAT, time from receipt of sample to report of the result) of the three hospitals decreased by 4.1 min, 8.8 min and 10.2 min, respectively. Autoverification systems using a Mindray BC-6800 auto hematology analyzer and labXpert were confirmed as being effective in reducing TAT and enhancing working efficiency on the premise of ensuring low false negative rate.