The impact of the analytical performance specifications of calcium and albumin on adjusted calcium.

Abstract

The generation of accurate, comparable results from traceable measurement procedures is a primary goal in harmonization efforts. In this study, the analytical performance of routine methods for calcium and albumin measurement is assessed to define the impact of the analytical bias of calcium and albumin on adjusted calcium equation performance and on reference intervals. In collaboration with the Wales External Quality Assurance Scheme, six months' worth of anonymized data that cover a concentration range of clinical interest were collected. The data were grouped by analytical platforms/methods. Albumin BCG methods are positively biased (8%) to BCP methods. The overall bias for BCP methods ranges from 5.1 to -4.3% and the overall bias for BCG methods is from 2 to -6.7%. Bias for both methods is higher than the allowable minimal bias for albumin. Calcium concentrations for Roche Cobas CPC and NM-BAPTA, Beckman Arsenazo III, Abbott Architect Arsenazo III were within bias of 1.5 to -1%. However, Siemens calcium methods CPC and Arsenazo III appear to suffer from concentration-dependent bias ranging from +3 to -6%, which exceeds even the minimal allowable limits for calcium (1.3%). Adjusted calcium shows significant bias of 11%. Even with the exclusion of Siemens Advia, the scatter of adjusted calcium results exceeds that for total calcium. This study shows wider than acceptable analytical variation for albumin and calcium. This variation may contribute to overall adjusted calcium equations variation and invalidate the application of a harmonized reference interval for calcium and albumin.