Creatinine Assays: Time for Action?

Abstract

Background Major differences in creatinine results between different laboratories and apparent inaccuracies when using commercial lyophilized standards were noted. In order to assess the variation and accuracy of the different methods we conducted a local audit. Methods To establish the variation between methods, plasma creatinine was measured on 47 human plasma samples by nine different laboratories using four different methods (Roche, Ortho, Olympus, modified Olympus). To establish the accuracy of the different methods, plasma creatinine was also determined on 16 of the plasma samples by tandem mass spectrometry (MS). In addition, all the laboratories measured the creatinine concentration on a commercial authenticated sample. Results All four methods gave significantly different ( P< 0·0001) plasma creatinine results when compared with each other. Generally, creatinine results produced by the Ortho method were considerably higher than those of the other methods, especially at higher creatinine concentrations (differences across methods between the lowest and highest result for the same sample ranged between 8% and 33%). All four methods generally gave higher results than those determined by tandem MS for samples with creatinine concentrations of < 250 μmol/L. Above this concentration the Olympus and Roche methods produced creatinine results that were lower then the tandem MS results, whereas results from the Ortho method were higher. Major matrix problems were found when a commercial lyophilized standard was used for creatinine estimation. Conclusion No method gave good agreement with the tandem MS results, and there were major differences in measured plasma creatinine concentrations (up to 30% difference) between the various methods. We suggest that efforts should be made to standardize plasma creatinine measurement across all laboratories to minimize these problems.