Assessing the agreement between two quantitative assays with repeated measurements

Area of Science:

• Analytical Chemistry
• Biostatistics

Background:

• Assay validation is critical for reliable scientific measurements.
• Evaluating agreement between two measurement methods requires robust statistical approaches.
• Existing methods may lack the power to definitively assess equivalence.

Purpose of the Study:

• To propose a novel statistical test for evaluating the equivalence or agreement between two assay methods.
• To assess agreement by considering both the conditional mean and variance of measurement differences.
• To provide a statistically assured conclusion on the practical meaningfulness of differences between assay methods.

Main Methods:

• The proposed method utilizes repeated measurements from sampled subjects.
• It assesses agreement based on the conditional mean and variance of the difference between assay values (x - y) given the true analyte concentration (z).
• The test is distribution-free and model-independent, requiring no assumptions on the underlying data distribution or conditional models.

Main Results:

• The statistical test provides a robust evaluation of assay agreement.
• It allows for conclusions regarding the practical equivalence of two assay methods.
• The approach is analogous to methods used in bioequivalence testing for pharmaceuticals.

Conclusions:

• The developed statistical test offers a reliable tool for assay validation.
• It enhances confidence in concluding assay equivalence when differences are not practically meaningful.
• This method contributes to rigorous instrument validation and comparative analytical studies.