Abstract Background Serum bilirubin could be used to assess liver functions. Due to clinical importance, it is highly desired to have an accurate bilirubin test. However, there are circumstances where error could occur in bilirubin test results. In particular, the hemolysis interference is one of the major problems that could mislead the direct bilirubin results because of their relatively high occurring frequencies in the clinical laboratories. To resolve this issue and improve the accuracy for the direct bilirubin, we developed a new reagent formulation for DBIL assay that minimizes hemolysis interference. In addition, this reagent also shown good performance characteristics on Mindray BS-2800M Clinical Chemistry analyzer. Methods The new DBIL assay is a vanadate-oxidation method that bilirubin is gradually transformed to biliverdin because of oxidizing agent vanadate, resulting in decreasing of absorbance at 450 nm. Two-level calibration is used for this assay. DBIL concentrations are obtained from the calibration line with signals from the patient samples in the assay. Following CLSI protocols, basic assay performances were evaluated: interference, precision, linearity, method comparison and low end detection (LOB, LOD, LOQ). Results Hemolysis interference was tested in parallel with DBIL reagents from different IVD companies, Roche, Wako and Chinese domestic main-player companies, A, B, and C. The results showed that the reagents of A and B companies were significantly suffered from hemolysis interference with maximum bias exceeded -30%. Although there are some effects on the anti-hemolysis with DBIL reagents of Wako and C company, the DBIL reagent of Mindray was the best with less than -8% of impact when hemolysate hemoglobin was up to 1000 mg/dL (H index). The Diazo DBIL assay from Roche shown issue with hemolysis interference. The interference to our new DBIL assay from lipid (up to 1000 mg/dL) and ascorbic acid (up to 30 mg/dL) were also with less than ±10% impact. The precision study was run according to EP05A3 with two commercial controls (∼ 10 and 30 µmol/L) and three serum sample pools (∼ 6.8,17.1 and 342 µmol/L). The CVs of repeatability were ranged from 0.54% to 0.58% for controls and 0.24% to 0.88% for serum pools, respectively. The within-lab precision was below 2.5%. The analytical linearity was determined as from 0.66 µmol/L to 439.66 µmol/L. For low-end detection, the LOB, LOD and LOQ were 0.5,1.0 and 1.7 µmol/L, respectively. In addition, the Mindray (y) correlated well with the DBIL assay of Wako (x) reagent on Mindray BS-2800M system (y): y =0.9847 x+0.1301 (r =0.9997, n =400, sample range: 1.08 - 430.00 μmol/L). Conclusions The new Mindray direct bilirubin assay shown a good anti-hemolysis interference performance on BS-2800M system and demonstrated good performances in other assay characteristics. It is therefore suitable for use in clinical laboratories.
Read full abstract