Evaluation of the trueness of serum alkaline phosphatase (ALP) measurement in a group of Italian laboratories

The reference measurement procedure for ALP published in 2011 by the IFCC allowed to define the metrological traceability chain for the standardization of ALP measurement. This paper reports the results of an EQA experiment conducted to evaluate the level of ALP standardization among different Italian laboratories enrolled for a scientific project with the final aim to derive ALP traceable reference intervals for pediatric population. Three frozen serum pools with a target value assigned by the IFCC reference procedure were distributed to 13 centers and analyzed in triplicates for 3 different days. Only 3 laboratories averagely fulfilled the desirable goal of bias (<=+/- 5.5%) at all 3 concentrations (59.9 U/L, 186.9 U/L and 401.5 U/L), but only one provided data with a dispersion always within the uncertainty of the target result. The different ability to meet the goal clearly depended on the analytical system used. Focusing on the two most used analytical platforms, the Cobas systems (Roche Diagnostics) underestimated the ALP values, while the AU systems (Beckman Coulter) overestimated them. The regression parameters between the average values obtained by laboratories and the target values indicate that it would be possible to correct the results of all analytical systems and make them unbiased by a simple recalibration approach. The analysis of the commercial calibrator package inserts of the IVD companies involved in this study showed that, with the exception of Roche still aligned to the old Tietz method published in 1983, all companies offer at least two options, sometimes (e.g., Beckman AU) both not in line with the recommended standardization approach.