Commutability of Possible External Quality Assessment Materials for Cardiac Troponin Measurement

BACKGROUND The measurement of cardiac troponin is crucial in the diagnosis of myocardial infarction. The performance of troponin measurement is most conveniently monitored by external quality assessment (EQA) programs. The commutability of EQA samples is often unknown and the effectiveness of EQA programs is limited. METHODS Commutability of possible EQA materials was evaluated. Commercial control materials used in an EQA program, human serum pools prepared from patient samples, purified analyte preparations, swine sera from model animals and a set of patient samples were measured for cTnI with 4 assays including Abbott Architect, Beckman Access, Ortho Vitros and Siemens Centaur. The measurement results were logarithm-transformed, and the transformed data for patient samples were pairwise analyzed with Deming regression and 95% prediction intervals were calculated for each pair of assays. The commutability of the materials was evaluated by comparing the logarithmic results of the materials with the limits of the intervals. Matrix-related biases were estimated for noncommutable materials. The impact of matrix-related bias on EQA was analyzed and a possible correction for the bias was proposed. RESULTS Human serum pools were commutable for all assays; purified analyte preparations were commutable for 2 of the 6 assay pairs; commercial control materials and swine sera were all noncommutable; swine sera showed no reactivity to Vitros assay. The matrix-related biases for noncommutable materials ranged from -83% to 944%. Matrix-related biases of the EQA materials caused major abnormal between-assay variations in the EQA program and correction of the biases normalized the variations. CONCLUSION Commutability of materials has major impact on the effectiveness of EQA programs for cTnI measurement. Human serum pools prepared from patient samples are commutable and other materials are mostly noncommutable. EQA programs should include at least one human serum pool to allow proper interpretation of EQA results.