State-of-the-art of serum total calcium measurement as investigated by split-sample measurement with an ion chromatography candidate reference method.

We evaluated bias and inaccuracy of four frequently used routine test systems for serum total calcium, using a candidate reference method based on ion chromatography. The mean biases and 95% confidence intervals that we observed were 0.0 +/- 0.59% for Johnson & Johnson arsenazo(III), 1.3 +/- 0.62% for Beckman arsenazo(III), -0.4 +/- 0.44% for Beckman applying ion selective electrode measurement after sample dilution, and -1.9% +/- 0.42% for Boehringer o-cresolphthalein. The inaccuracy of all test systems was usually < 4.7% (calculated as deviation of singlicates from ion chromatography). Both bias and inaccuracy are discussed in the light of specifications set by expert groups or derived from the biological variation of serum total calcium. The study revealed that the intrinsic quality of commonly used test systems for serum total calcium satisfies even some of the more stringent criteria for method bias and inaccuracy.