CIDGMed: Causal Inference-Driven Medication Recommendation with Enhanced Dual-Granularity Learning

TL;DR

The Causal Inference-driven Dual-Granularity Medication Recommendation method (CIDGMed) leverages causal inference to uncover the relationships between diseases/procedures and medications, thereby enhancing the rationality and interpretability of recommendations.

Abstract

Medication recommendation aims to integrate patients' long-term health records to provide accurate and safe medication combinations for specific health states. Existing methods often fail to deeply explore the true causal relationships between diseases/procedures and medications, resulting in biased recommendations. Additionally, in medication representation learning, the relationships between information at different granularities of medications, coarse-grained (medication itself) and fine-grained (molecular level), are not effectively integrated, leading to biases in representation learning. To address these limitations, we propose the Causal Inference-driven Dual-Granularity Medication Recommendation method (CIDGMed). Our approach leverages causal inference to uncover the relationships between diseases/procedures and medications, thereby enhancing the rationality and interpretability of recommendations. By integrating coarse-grained medication effects with fine-grained molecular structure information, CIDGMed provides a comprehensive representation of medications. Additionally, we employ a bias correction model during the prediction phase to further refine recommendations, ensuring both accuracy and safety. Through extensive experiments, CIDGMed significantly outperforms current state-of-the-art models across multiple metrics, achieving a 2.54% increase in accuracy, a 3.65% reduction in side effects, and a 39.42% improvement in time efficiency. Additionally, we demonstrate the rationale of CIDGMed through a case study.

Figures (8)

• Figure 1: A case where a method based on co-occurrence relationships leads to erroneous recommendation results.
• Figure 2: CIDGMed Flowchart: The Relationship Mining module on the left side of the figure utilizes causal discovery and causal inference on the EHR to generate causal graphs and causal effect matrices. The Representation Learning module at the top of the figure builds on this foundation with dual granularity at the medication level and the molecular level to learn patient representations and recommend preliminary medication probabilities. The Bias Correction module at the bottom of the figure corrects the recommended probabilities for each medication based on the causal effects and recommends the final combination of medications.
• Figure 3: A real example of using causal discovery to correct incorrect relationships from co-occurrence relationships to generate true relationships.
• Figure 4: A real example of DAC, we analyze the causal position of each node in the causal graph $G^d_{v_t}$ to gauge its influence on the patient's current state. For instance, if $d_3$ is identified as a causal disease, it is classified as ${D}^1_t$.
• Figure 5: Comparison with recent outstanding works across all metrics in MIMIC-III.