A Generalized Pivotal Quantity Approach to Analytical Method Validation Based on Total Error.

Area of Science:

• Analytical Chemistry
• Statistical Modeling
• Regulatory Science

Background:

• Method validation is crucial for ensuring the safety, efficacy, and quality of medicinal products.
• Traditional validation criteria do not directly relate to the intended purpose of guaranteeing accurate future test results.
• Existing statistical tests for 'fit for purpose' validation may fail to adequately protect against accepting unsuitable methods, leading to consumer risk.

Purpose of the Study:

• To propose a novel statistical test for analytical method validation.
• To provide enhanced protection against the risk of accepting unsuitable analytical methods.
• To evaluate the performance of the proposed method against existing approaches.

Main Methods:

• Development of a new statistical test based on generalized pivotal quantity inference.
• Comparison of the proposed method with five existing approaches through simulation studies.
• Assessment of Type I error control and consumer risk protection.

Main Results:

• The proposed generalized pivotal quantity method and the beta-content (0.9) method effectively control Type I error and consumer risk.
• The generalized pivotal quantity method is less conservative for biased methods and more conservative for unbiased methods compared to beta-content (0.9).
• The generalized pivotal quantity method demonstrates superior asymptotic properties over current methods.

Conclusions:

• The generalized pivotal quantity method offers robust protection against accepting unsuitable analytical methods.
• The choice between the generalized pivotal quantity and beta-content (0.9) methods depends on the specific analytical method's accuracy.
• This novel statistical test improves the reliability of analytical method validation.