Beyond the ATE: Interpretable Modelling of Treatment Effects over Dose and Time

2025-07-09

AI Summary

This work addresses the limitation of the average treatment effect (ATE) in causal inference—its inability to capture the dynamic evolution of treatment effects across dosage and time—by proposing an interpretable, dynamic causal modeling framework tailored for clinical applications. Methodologically, it adapts the SemanticODE framework to causal inference, explicitly decoupling the learning of effect trajectory shapes from clinically meaningful semantic definitions (e.g., onset time, peak magnitude, duration), thereby enabling domain-knowledge injection, post-hoc editing, and verifiable analysis. Leveraging a proxy-learning strategy, the method estimates smooth and robust dose–time effect surfaces without requiring ground-truth treatment effects. Evaluated on both synthetic and real-world healthcare datasets, it achieves superior estimation accuracy while producing clinically interpretable, editable, and verifiable dynamic effect trajectories. These advances significantly enhance transparency and trustworthiness in high-stakes causal decision-making.

Technology Category

• Machine Learning: Causal Learning
• Reasoning under Uncertainty: Causality
• Cognitive Modeling & Cognitive Systems: Conceptual Inference and Reasoning

Application Category

• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions
• User Modeling, Personalization and Recommendation: Explainable and interpretable methods for personalization
• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs

Abstract

The Average Treatment Effect (ATE) is a foundational metric in causal inference, widely used to assess intervention efficacy in randomized controlled trials (RCTs). However, in many applications -- particularly in healthcare -- this static summary fails to capture the nuanced dynamics of treatment effects that vary with both dose and time. We propose a framework for modelling treatment effect trajectories as smooth surfaces over dose and time, enabling the extraction of clinically actionable insights such as onset time, peak effect, and duration of benefit. To ensure interpretability, robustness, and verifiability -- key requirements in high-stakes domains -- we adapt SemanticODE, a recent framework for interpretable trajectory modelling, to the causal setting where treatment effects are never directly observed. Our approach decouples the estimation of trajectory shape from the specification of clinically relevant properties (e.g., maxima, inflection points), supporting domain-informed priors, post-hoc editing, and transparent analysis. We show that our method yields accurate, interpretable, and editable models of treatment dynamics, facilitating both rigorous causal analysis and practical decision-making.

Problem

• Modeling treatment effects varying with dose and time
• Ensuring interpretability and robustness in causal inference
• Extracting clinically actionable insights from treatment dynamics

Innovation

• Models treatment effects as dose-time surfaces
• Adapts SemanticODE for causal inference
• Decouples trajectory shape from clinical properties