Beyond Total Error: An Observational Comparative Study of Sigma Metrics in Long-Term Arterial Blood Gas Analyzer...

Accurate respiratory and metabolic status evaluation requires valid and reliable arterial blood gas (ABG) analyzers. This study aimed to evaluate the analytical performance of three (ABG) analyzers using total error (TE) and sigma metric (SM) to establish a robust framework for clinical device performance monitoring.

The performance of three ABG analyzers (A, B, and C) was evaluated using third-party quality control materials from Randox® across three levels for pH, PCO₂, and PO₂. Each level was analyzed 20 times per analyzer over five days. TE and SM were calculated using standard formulas, with total allowable error (TEa) derived from the 2024 Clinical Laboratory Improvement Amendment (CLIA) guidelines. Mean values were compared to manufacturer-provided target ranges, and performance was assessed against accuracy specifications and Sigma metrics.

All analyzers achieved acceptable mean values within target ranges. Accuracy (TE < TEa) was met for PO₂ and PCO₂ across all analyzers but failed for pH in analyzer A (levels 1, 2), analyzer B (all levels), and analyzer C (levels 1, 3). Sigma metrics showed suboptimal performance for pH (1.6-2.06), PCO₂ ranged from Unaccepted (1.61) to Good (4.34), and PO₂ indicated Good to World-Class performance. No significant differences were observed among analyzers for SM or TE (p > 0.05), but accuracy strongly correlated with performance consistency (Cramér's V = 0.918, p < 0.001).

This study underscores the limitations of relying solely on TE and highlights the importance of integrating SM to ensure reliable and clinically relevant ABG analyzer evaluations.