Clinical and Analytical Impact of Moving from Jaffe to Enzymatic Serum Creatinine Methodology

Abstract Background Identification and monitoring of chronic kidney disease (CKD) requires accurate quantification of serum creatinine. The poor specificity of Jaffe creatinine methods is well documented, and guidelines recommend enzymatic methodology. We describe our experience of moving from Jaffe to enzymatic creatinine methodology. We present comparison of >5000 paired Jaffe and enzymatic creatinine results, examine interferences, and attempt to assess clinical consequences of changing methodology. Methods Overall, 5303 serum samples received for routine creatinine measurement were analyzed using Jaffe and enzymatic methods with an Abbott Architect autoanalyzer. Associated results for glucose, total bilirubin, triglycerides, total protein, and hemolytic, icteric, and lipemic indexes were extracted from the laboratory database. CKD staging was estimated for each sample to assess potential clinical effects. Results The methods correlated well (r = 0.996) and showed good agreement (Passing-Bablok fit, y = 0.935x + 0.074). Paired analysis, however, showed significant differences (P < 0.001), and approximately 20% of results differed by more than ±10%. Multivariate analysis demonstrated independent associations between difference in creatinine results, glucose (P < 0.0001), and hemolytic index (P = 0.009). Glucose demonstrated positive interference in the Jaffe method, and hemolysis produced negative interference in the enzymatic method. Little or no association was observed with other analytes. CKD staging differed in 4% of samples. Conclusions Differences between Jaffe and enzymatic serum creatinine results exceed the recommended 5% target for a significant proportion of samples, particularly at concentrations <1.13 mg/dL (100 µmol/L). Both glucose and hemolysis contribute to the variance in results. Although the clinical impact of these differences seems small, laboratories should continue moving toward enzymatic creatinine estimation to ensure the best estimate of renal function.