Precision psychiatry needs causal inference.

Area of Science:

• Psychiatry
• Statistical Methods
• Computational Psychiatry

Background:

• Psychiatric research utilizes both causal inference and prediction frameworks.
• Recent trends advocate for prioritizing prediction for "precision psychiatry" and individualized treatments.
• This perspective critically evaluates these proposals.

Purpose of the Study:

• To critically appraise the roles of causal inference and prediction in psychiatric research.
• To highlight the necessity of causal inference for individualized treatment decisions.
• To defend the importance of causal inference in advancing precision psychiatry.

Main Methods:

• Outlining strengths and weaknesses of causal inference and prediction frameworks.
• Describing the link between clinical decision-making and counterfactual predictions (causality).
• Identifying key causal structures and prediction pitfalls that can lead to erroneous interpretations.

Main Results:

• Both prediction and causal inference are vital in psychiatric research.
• The relative importance of each framework is context-dependent.
• Causal inference is indispensable when individualised treatment decisions are required.

Conclusions:

• Causal inference is fundamental for achieving the goals of precision psychiatry.
• This perspective advocates for the continued and essential role of causal inference in psychiatric research.
• Integrating causal inference is key to advancing personalized mental healthcare.