Comprehensive Comparison of ISO-GUM and Monte Carlo Simulation for Uncertainty Assessment in HPLC Statin Quantification with single and multi-point calibrations

Abstract

Abstract Background Reliable measurement uncertainty estimation is essential for building trust in analytical results and ensuring they meet regulatory expectations. In HPLC chromatography, the chosen approach for uncertainty estimation can strongly influence both the results and their interpretation. Objective This study compares two widely used strategies, the ISO-GUM method and Monte Carlo simulation applied with either single-point or multi-point calibration, for the simultaneous quantification of four statins at the National Laboratory for Drug Control (LNCM). Methods In the single-point approach, all sources of uncertainty were combined into one global model, while the multi-point approach evaluated variability across different calibration levels. Both strategies were applied within the ISO-GUM framework and through Monte Carlo simulation, allowing a side-by-side comparison of the resulting uncertainty budgets. Results The single-point method tended to underrepresent the variability present in actual measurements. In contrast, the multi-point method, especially when paired with Monte Carlo simulation, delivered uncertainty estimates that better reflected real-world analytical conditions. The ISO-GUM framework combined with multi-point calibration also proved to be a robust and practical option for routine use. Conclusions Multi-point calibration, whether combined with Monte Carlo simulation or the ISO-GUM method, offers a more representative and reliable way to assess uncertainty in HPLC statin analysis.