Evaluation of bilirubin interference and accuracy of six creatinine assays compared with isotope dilution-liquid chromatography mass spectrometry.

Abstract

OBJECTIVES: The aim of this study was to estimate bilirubin interference and accuracy of six routine methods for measuring creatinine compared with isotope dilution-liquid chromatography mass spectrometry (ID-LC/MS). DESIGN AND METHODS: A total of 40 clinical serum samples from 31 patients with serum total bilirubin concentration >68.4μmol/L were collected. Serum creatinine was measured using two enzymatic reagents and four Jaffe reagents as well as ID-LC/MS. Correlations between bilirubin concentration and percent difference in creatinine compared with ID-LC/MS were analyzed to investigate bilirubin interference. Bias estimations between the six reagents and ID-LC/MS were performed. Recovery tests using National Institute of Standards and Technology (NIST) Standard Reference Material (SRM) 967a were also performed. RESULTS: Both the enzymatic methods showed no bilirubin interference. However, three of the four Jaffe methods demonstrated significant bilirubin concentration-dependent interference in samples with creatinine levels <53μmol/L, and two of them showed significant bilirubin interference in samples with creatinine levels ranging from 53.0 to 97.2μmol/L. Comparison of these methods with ID-LC/MS using patients' samples with elevated bilirubin revealed that the tested methods failed to achieve the bias goal at especially low levels of creatinine. In addition, recovery test using NIST SRM 967a showed that bias in one Jaffe method and two enzymatic methods did not achieve the bias goal at either low or high level of creatinine, indicating they had calibration bias. One enzymatic method failed to achieve all the bias goals in both comparison experiment and recovery test. CONCLUSIONS: It is important to understand that both bilirubin interference and calibration traceability to ID-LC/MS should be considered to improve the accuracy of creatinine measurement.