Evaluation of the commutability of control materials for C-peptide proficiency testing

Commutability refers to the consistency of measurement results between reference materials (RMs) and patient samples across different measurement procedures. C-peptide is essential for diabetes classification and management, however, inter-assay variability persists despite ongoing harmonization efforts. The use of commutable materials is therefore critical for calibration and external quality assessment (EQA). This study evaluated the commutability of pooled human serum and commercial control materials using CLSI- and IFCC-based approaches.

Commutability was evaluated using 30 clinical samples and six RMs, including three serum-based in-house prepared materials and three commercial materials. C-peptide was measured in triplicate using four major measurement procedures in Korea; Alinity i (Abbott Diagnostics), cobas e801 (Roche Diagnostics), UniCel DxI 800 (Beckman Coulter), and Atellica IM 1600 (Siemens Healthineers). Commutability was assessed using the CLSI EP14-ED4 Deming regression-based approach with 95% prediction intervals (PIs) and the IFCC difference-in-bias approach with expanded uncertainty, applying a total allowable error of 7.1% and a coverage factor of 1.9.

Three clinical samples were excluded from the analyses after being identified as outliers in scatter plots. Using the CLSI approach with the remaining 27 clinical samples, all RMs fell within the 95% PIs across all method pairs and were classified as commutable. In contrast, the IFCC approach, with concentration partitioning at 1.0 ng/mL distinguished differences in commutability among the RMs. Serum-based in-house prepared materials were classified as commutable or indeterminate, whereas commercial materials showed multiple indeterminate results and one non-commutable result.

Pooled human serum-based in-house prepared materials demonstrated more stable commutability than commercial materials, supporting their use in C-peptide EQA programs.