Performance evaluation of four detection systems for total bilirubin

Objective To evaluate the performance of four detection systems for total bilirubin. Methods The precision,trueness, anti-interference,analytical measuring range( AMR) were evaluated according to EP5-A2,EP15-A2,EP7-A2,EP6-A and EP9-A2 guidelines of Clinical and Laboratory Standards Institute( CLSI) and the correlation and deviation of the results were compared with Diazo J-G of Beckman coulter DXC800 system. The EQA samples( N-A and N-B) from College of American Pathologist( CAP) were used to assess the trueness of D system. The qualitative evaluation content was traced back to the reference method program setting of clinical chemistry. The regression was analyzed using Passing-Bablok method and the linear regression was detected by Cusum method. Pearson correlation analysis and Bland-Altman deviation analysis were performed. Results The precision of the four detection systems( A-D) were less than the standard of each manufacturer statement when the concentration of total bilirubin was in 17 to 258 μmol/L. The measurements of CAP EQA samples showed that the relative deviation of the four detection systems were less than the standard of each manufacturer statement. Hemoglobin,triglyceride and vitamin C may cause interference to different extents for the four systems. The upper limits of AMR for A,B,C and D system were 733. 58,230. 18,710. 09 and 459. 78 μmol / L respectively. The A,B,C system showed better correlation with D system. Compared with D system,the Bland-Altman results showed that the average absolute deviation were- 16. 6,3. 4 and- 0. 8 μmol / L,and the average relative devation were- 25. 32%,- 3. 49% and- 11. 75% for the A,B and C system,respectively. Conclusion Certain differences of performance among the four systems for total bilirubin detection were found. The results of Vanadate oxidation were lower than those of Diazo J-G method.