Evaluation of the commutability of lombardy region EQA control materials for enzymes, calcium, and creatinine

External Quality Assessment (EQA) is essential for ensuring inter-laboratory comparability, yet the interpretability of EQA data depends critically on the commutability of control materials (CM). This study assessed the commutability of CM used in the Lombardy Regional EQA program for selected clinical chemistry measurands.

A multicenter cross-sectional study across ten clinical laboratories, evaluated the commutability of three lyophilized human serum control materials supplemented with non-human enzymes and inorganic/organic constituents, analyzed in six replicates, and fifteen human serum pools analyzed in triplicate. Eight measurands were included: alkaline phosphatase (ALP), alanine aminotransferase, aspartate aminotransferase, creatine kinase, lactate dehydrogenase, cholinesterase (CHE), creatinine, and calcium. Commutability was evaluated pairwise among 312 method comparisons using the IFCC "difference-in-bias" framework. Acceptance limits were predefined as 0.5 × allowable total error, derived primarily from the EFLM Biological Variation database.

Patterns of commutability varied markedly across measurands and methods. Within-vendor comparisons generally supported commutability, whereas cross-manufacturer comparisons frequently demonstrated non-commutability, with isolated cases of level-dependent behavior. Full commutability across all CM was observed only in selected pairings (e.g., Abbott-Siemens for ALP; Roche-Werfen for enzymatic creatinine). Specific matrix-method interactions (e.g., dry-slide systems for CHE; NM-BAPTA versus arsenazo III for calcium) accounted for systematic non-commutability.

The study confirms that commutability of these CM is highly variable and measurand-dependent, limiting the interpretation of accuracy assessments in many cross-method settings. These findings underscore the need for rigorous verification of commutability in EQA design to support meaningful assessment of inter-method bias and harmonization.