Validation of Bioavailability-Based Toxicity Models for Metals.

Area of Science:

• Environmental toxicology
• Aquatic toxicology
• Risk assessment

Background:

• Regulatory bodies are adopting bioavailability-based toxicity models for metal risk assessment.
• Model validation is essential for ensuring the reliability of these protective values (PVALs).

Purpose of the Study:

• To review existing guidance on validating bioavailability-based toxicity models.
• To recommend criteria and a framework for model validation in PVAL development.
• To propose new methods for evaluating model performance.

Main Methods:

• Review of current model validation guidance.
• Identification of key questions for validation studies.
• Discussion of study design and evaluation metrics.
• Development of a validation framework for PVALs.

Main Results:

• Model validation requires assessment of appropriateness, relevance, and accuracy.
• New methods for evaluating model performance were presented.
• A framework for integrating model validation into PVAL development was suggested.

Conclusions:

• Model validation should be rigorous but adaptable to specific user needs.
• Models can be sufficiently validated for a defined purpose, acknowledging inherent uncertainty.
• Validation should use consistent experimental designs and cover diverse water types.