Counterfactual Formulation of Patient-Specific Root Causes of Disease

TL;DR

This paper introduces a mathematically rigorous, counterfactual approach to identifying patient-specific root causes of disease, enabling more accurate and efficient detection from data.

Contribution

It advances prior work by formulating root causes using counterfactuals on Pearl's third causal rung and incorporates Shapley values for causal contribution scoring.

Findings

Provides a counterfactual definition aligned with clinical intuition.

Allows fast computation without counterfactual simulation.

Handles noisy labels and adapts to disease prevalence.

Abstract

Root causes of disease intuitively correspond to root vertices that increase the likelihood of a diagnosis. This description of a root cause nevertheless lacks the rigorous mathematical formulation needed for the development of computer algorithms designed to automatically detect root causes from data. Prior work defined patient-specific root causes of disease using an interventionalist account that only climbs to the second rung of Pearl's Ladder of Causation. In this theoretical piece, we climb to the third rung by proposing a counterfactual definition matching clinical intuition based on fixed factual data alone. We then show how to assign a root causal contribution score to each variable using Shapley values from explainable artificial intelligence. The proposed counterfactual formulation of patient-specific root causes of disease accounts for noisy labels, adapts to disease prevalence and admits fast computation without the need for counterfactual simulation.