Causal diagrams in systems epidemiology.

Area of Science:

• Epidemiology
• Systems Science
• Causal Inference

Background:

• Diagrammatic modeling has advanced significantly.
• Epidemiology traditionally focuses on single causal links, neglecting
• causes of causes
• and system-level interactions.

Purpose of the Study:

• To advocate for system-wide causal diagrams in epidemiology.
• To highlight the limitations of single-link analyses.
• To demonstrate the utility of comprehensive causal modeling.

Main Methods:

• Developing and applying system-wide causal diagrams.
• Integrating knowledge of world function with statistical evidence.
• Utilizing concepts from infectious disease epidemiology modeling.

Main Results:

• System-wide models can characterize entire research areas and integrate diverse datasets.
• They facilitate the use of instrumental variables and natural experiments.
• These models aid in detecting and understanding feedback mechanisms.

Conclusions:

• Adopting system-wide diagrammatic modeling enhances epidemiological analysis.
• It allows for a more holistic understanding of disease determinants and their interrelationships.
• This approach unlocks new possibilities for causal inference and data integration.