Calculating preventable risk fractions for exposure-reducing interventions.

Area of Science:

• Environmental Health
• Epidemiology
• Causal Inference

Background:

• Traditional metrics like Population Attributable Fractions (PAFs) have limitations for interventional causal questions.
• Misuse of association-based metrics can lead to inaccurate risk assessments.

Purpose of the Study:

• To introduce novel causal metrics for quantifying preventable risk fractions from environmental and occupational exposures.
• To provide tools for scenario-specific, individual-level risk reduction predictions based on mechanistic causality.

Main Methods:

• Introduction of Interventional Probability of Causation (IPoC), Causal Assigned Shares (CAS), and Preventable Risk Fraction (PRF) curves.
• Application of methods to case studies: benzene and AML, smoking and lung cancer, blood lead and mortality.
• Utilizing Monte Carlo simulations for inter-individual variability and scenario analyses for practical thresholds.

Main Results:

• PRF curves effectively quantify risk-reduction benefits from exposure reductions at individual and population levels.
• Demonstrated ability to handle uncertainty and heterogeneity in exposure-risk relationships.
• Identified thresholds where further exposure reductions yield diminishing returns.

Conclusions:

• The proposed causal framework shifts focus from attribution to prevention of harm.
• Offers a transparent, rigorous approach for evidence-based risk assessment, policy development, and legal decision-making.
• Enables quantification of causally effective interventions and risk-reduction benefits.