Testing assay linearity over a pre-specified range.

Area of Science:

• Analytical Chemistry
• Statistical Modeling
• Regulatory Science

Background:

• Linearity validation is a critical regulatory requirement in analytical method development.
• Existing linearity testing methods present challenges in setting acceptable limits and are difficult for non-statisticians to implement.
• Current approaches often depend on the specific concentration levels used during the linearity experiment.

Purpose of the Study:

• To propose a novel statistical procedure for robust linearity testing.
• To overcome the limitations of existing methods, particularly regarding implementation and limit setting.
• To provide an easy-to-implement solution for validating assay linearity.

Main Methods:

• The proposed method employs a two one-sided test (TOST) of equivalence.
• It evaluates the potential bias introduced by approximating higher-order polynomial responses with a linear model.
• Utilizes orthogonal polynomials and generalized pivotal quantity analysis for a closed-form solution.

Main Results:

• The method provides a statistically sound procedure for linearity validation.
• Offers a closed-form solution, simplifying the implementation process.
• Addresses the difficulties in setting fit-for-purpose acceptable limits.

Conclusions:

• The proposed statistical procedure offers a practical and accessible solution for linearity testing.
• Enhances the reliability and ease of regulatory compliance for analytical methods.
• Facilitates accurate linearity assessment for a wider range of users.