Background: Analytical validation is a sequence of operations aiming to evaluate the accuracy, reliability, and cost of analytical results for making informed decisions in method selection and meeting the requirements of regulatory institutions. Objective: This study aims to perform an analytical validation by comparing three different approaches: the accuracy profile, the uncertainty profile, and the conventional validation to assess the capability of each method in confirming the robustness of the results. Methods: The accuracy profile offers a comprehensive assessment of analytical performance and integrates systematic and random errors to determine if future results will satisfy the predefined acceptance limits. Meanwhile, the uncertainty profile, which is complementary and innovative, allows uncertainty estimation from validation data. These approaches were developed after conventional validation that relies on statistical methodologies based on separate evaluations of method criteria to provide a comparative framework for evaluating new methods. Results: This comparison will give recommendations for best practices related to analytical validation. The uncertainty profile is a graphical decision-making tool for determining full validation by integrating analytical validation and the estimation of measurement uncertainty, evaluating two statistical methods: β-expectation tolerance intervals and β-content, γ-confidence tolerance intervals, using a formula introduced by Saffaj and Ihssane, predicting that 95% of future results will fall within the acceptance limits of ±5%, revealing that the tolerance intervals for β-expectation are smaller than β-content, γ-confidence. Conclusion: The total error approaches offer robust recommendations for optimal methods for routine application. Highlights: This study highlights the critical need for appropriate analytical validation and the challenges arising from the absence of clear guidelines for that purpose. Different approaches emphasize the significant impact of the choice of an adequate method, which remains pivotal for providing accurate results in real-world scenarios. Concrete examples and simulations illustrate the viewpoints associated with different approaches to making decisions.
17 July 2020 in Journal of Pharmaceutical Innovation
Background: Azo dyes are synthetic dyes that can degrade to carcinogenic amines, therefore determining their exact concentration is crucial. Objective: The purpose of this study is to demonstrate that an HPLC method may be used to quantify 22 azo dye amines in a textile matrix. Methods: The HPLC–DAD (Diode Array Detector) technique for measuring 22 amines in a textile matrix was tested using a new graphical statistical strategy based on the building of β-content, γ-confidence tolerance intervals, which allows for good qualimetry of analytical procedures. In this way, we established the approach's ability to perform a full validation of the analytical technique as well as its ability to estimate measurement uncertainty in a simple and uncomplicated way with minimum effort, utilizing only the data collected during the analytical validation. Results: The findings suggest that this technique can properly assess the validity of the HPLC method for simultaneous determination of 22 amines, with 67–90% of compliance results falling within acceptable limits. Conclusion: Using the uncertainty profile methodology, the full validation of an HPLC method for the determination of 22 azo amines in a textile matrix was successfully established. Highlights: The development and evaluation of an efficient HPLC technique for the simultaneous determination of 22 azo amines in a textile matrix have been completed. The total reliability of the analytical procedure was established by evaluating its measurement uncertainty at each concentration level using the suggested uncertainty profile approach and confirming proportions 67 to 90% of compliance.
Background: Measurement uncertainty (MU) is a crucial parameter for ensuring the reliability of analytical methods and the validity of results, as required by ISO 17025:2017. Its estimation is particularly critical for quality control laboratories, where compliance decisions are based on a rigorous interpretation of uncertainties. Methods: In this study, we evaluated the uncertainty associated with an HPLC-UV method for the determination of Metopimazine (MPZ) in a pharmaceutical form, applying two complementary approaches: The ISO-GUM (Guide to the Expression of Uncertainty in Measurement) top-down approach and the Monte Carlo Simulation (MCS). Results: The results of both approaches showed excellent agreement, thus validating the robustness of the evaluation. The analysis of uncertainty sources revealed that the accuracy of the sample volume (VSample) and the calibration standard (Cx) were the dominant contributors, representing 39.9% and 36.2% of the total uncertainty, respectively. Combined, these two factors accounted for 76.1% of the variability, underscoring their critical impact on the assay’s precision. The expanded uncertainty (k = 2, 95% confidence level) was determined to be (99.41 ± 0.69)%, reflecting the method’s reproducibility. Conclusions: These results highlight the importance of rigorously controlling calibration standard preparation, sample volume, and repeatability conditions to optimize the reliability of the assay.