Causal criteria: time has come for a revision

Area of Science:

• Epidemiology and Public Health
• Causal Inference
• Scientific Methodology

Background:

• Epidemiologists aim to establish causation for disease prevention, but correlation does not always imply causation.
• Established causal criteria, such as the Bradford Hill criteria, were developed 50-60 years ago and have not kept pace with advances in epidemiologic theory and infrastructure.
• Current criteria may not adequately guide the translation of observed associations into actionable causal links for public health interventions.

Purpose of the Study:

• To advocate for a critical re-evaluation and modernization of existing causal criteria used in epidemiology.
• To propose the integration of a 'consequence' criterion alongside traditional procedural criteria for assessing causality.
• To provide a framework for determining when evidence of association warrants action, considering social and ethical implications.

Main Methods:

• Review and critique of existing epidemiological causal criteria, particularly the Bradford Hill criteria.
• Conceptual development of a new 'consequence' criterion for evaluating the implications of causal associations.
• Discussion on the integration of procedural and consequence criteria for robust causal inference.

Main Results:

• Existing causal criteria require updating to reflect advancements in epidemiologic methods and theory.
• A 'consequence' criterion is proposed to assess the social and ethical dimensions of causal associations, influencing the threshold for action.
• The proposed framework aims to better separate the discussion of causality from the discussion of appropriate public health actions.

Conclusions:

• The established causal criteria in epidemiology need revision to enhance their robustness and relevance.
• Incorporating a 'consequence' criterion is essential for a more comprehensive approach to causal inference and decision-making in public health.
• A collaborative effort involving experts, civil society, and the state is recommended for establishing updated, interconnected procedural and consequence criteria.