Equivalence testing using existing reference data: An example with genetically modified and conventional crops in...

Area of Science:

• Agricultural Science
• Biostatistics
• Toxicology

Background:

• Assessing the safety of regulated products, such as genetically modified organisms (GMOs), is crucial.
• Historical data from studies with assumedly safe reference products can inform current safety assessments.
• Traditional methods may not fully leverage historical data for robust equivalence testing.

Purpose of the Study:

• To describe a novel equivalence testing method for regulated product safety assessment.
• To utilize historical data from reference products to evaluate current products.
• To propose a statistical framework for integrating historical and current study data.

Main Methods:

• Developed an equivalence testing method using data from historical and current studies.
• Illustrated the method with animal feeding studies comparing genetically modified and reference maize.
• Selected study-corrected distribution-wise equivalence as appropriate for the case study.
• Proposed an equivalence test using generalized fiducial inference to integrate uncertainties.

Main Results:

• The proposed method effectively assesses the safety of regulated products by comparing them to historical reference data.
• Study-corrected distribution-wise equivalence was identified as a suitable criterion for the maize feeding study.
• The generalized fiducial inference approach successfully integrated uncertainties from both historical and current data.

Conclusions:

• The described equivalence testing method provides a robust framework for evaluating the safety of regulated products.
• This statistical approach enhances safety assessments by leveraging historical data and accounting for uncertainty.
• The method is applicable to various regulated products, including genetically modified crops.