Assessing interactive causes of an occurred outcome due to two binary exposures

Area of Science:

• Epidemiology
• Biostatistics
• Causal Inference

Background:

• Causal attribution analysis identifies key factors responsible for observed outcomes, differing from treatment effect evaluation.
• Assessing the likelihood of an outcome resulting from individual exposures versus their interaction is crucial, especially with multiple risk factors.
• Identifying retrospective causal interactions between exposures is challenging, even in randomized controlled trials.

Purpose of the Study:

• To define and establish the identifiability of posterior probabilities for characterizing interactive causes of observed outcomes.
• To develop a method for assessing the causal contribution of individual exposures and their interactions.
• To apply the proposed method to the classic case of smoking and asbestos exposure in lung cancer.

Main Methods:

• Defined posterior probabilities to quantify the contribution of interactive causes to an observed outcome.
• Established identifiability of these posterior probabilities using a secondary outcome variable.
• Applied the developed method to analyze lung cancer data involving smoking and asbestos exposure.

Main Results:

• The study found that for lung cancer patients with both smoking and asbestos exposure, the disease is primarily attributable to the synergistic effect between these two exposures.
• Individual contributions of smoking or asbestos exposure alone were less significant compared to their combined interactive effect.
• The proposed method successfully characterized the interactive causes of lung cancer in the studied population.

Conclusions:

• The synergistic effect between smoking and asbestos exposure plays a primary role in causing lung cancer.
• Posterior probabilities provide a robust framework for evaluating interactive causes of observed outcomes.
• This methodology offers a valuable tool for causal attribution analysis in complex etiological research.