graphPAF

graphPAF estimates, infers, and displays population attributable fractions (PAF) and impact fractions through an R package implementation.

Key Features:

• Standard PAF and Impact Fraction Inference: Performs estimation and inference for standard population attributable fractions and impact fractions.
• Multi-Risk Factor Visualization: Displays attributable fractions over multiple risk factors using fan-plots and nomograms.
• Continuous Exposure Calculations: Calculates attributable fractions for continuous exposures.
• Pathway-Specific Attributable Fractions: Enables inference for attributable fractions appropriate for specific risk factor → mediator → outcome pathways.
• Bayesian Network-Based Calculations: Performs Bayesian network-based calculations and inference for joint, sequential, and average population attributable fractions in multi-risk factor scenarios.

Scientific Applications:

• Population Attributable Fraction Estimation: Quantifies the proportion of disease incidence in a population attributable to specific risk factors.
• Impact Fraction Analysis: Estimates the potential impact of risk factor modification on population health outcomes.
• Multi-Risk Factor Attribution: Disentangles contributions of multiple correlated risk factors to disease burden using joint, sequential, and average PAF measures.
• Causal Pathway Analysis: Identifies attributable fractions along specific risk factor → mediator → outcome pathways to understand intermediate mechanisms.
• Continuous Exposure Assessment: Evaluates attributable fractions for risk factors measured on continuous scales.

Methodology:

Bayesian network-based calculations support inference for joint, sequential, and average population attributable fractions.