Selecting experimental data for use in quantitative risk assessment: an expert-judgement approach

Area of Science:

• Toxicology and Risk Assessment
• Environmental Health Sciences
• Biostatistics

Background:

• Current cancer risk assessments often use generic guidelines for experimental data selection, which can be arbitrary and exclude valuable scientific insights.
• This can lead to suboptimal statistical modeling and potentially inaccurate low-dose risk estimations.

Purpose of the Study:

• To implement and evaluate an alternative approach for data selection in cancer risk assessment using expert scientific judgment.
• To assess the impact of expert-driven data selection on low-dose extrapolation for formaldehyde.

Main Methods:

• Elicited data selection judgments from eight practicing scientists using a formal interview protocol.
• Applied the multistage model to selected data sets for dose-response assessment and low-dose extrapolation of formaldehyde.
• Compared expert-driven data selection with previously published risk assessment methodologies.

Main Results:

• Low-dose risk estimates for formaldehyde were found to be sensitive to the specific data sets chosen, particularly regarding the inclusion of benign tumor data.
• Expert recommendations for data selection differed from those derived from generic guidelines used in prior assessments.
• The expert judgment approach demonstrated potential advantages in data selection for risk assessment.

Conclusions:

• Scientific judgment can effectively augment generic guidelines in cancer risk assessments when a diverse range of experimental data is available.
• Incorporating expert insights into data selection may lead to more robust and scientifically grounded risk characterizations.
• The expert judgment approach offers a valuable alternative for improving the reliability of toxicological risk assessments.