Drug-disease Graph: Predicting Adverse Drug Reaction Signals via Graph Neural Network with Clinical Data

TL;DR

This paper introduces a graph neural network framework that leverages clinical data to predict adverse drug reaction signals, demonstrating improved performance and the ability to identify novel ADR pairs.

Contribution

The study presents a novel graph-based approach using healthcare claims data and GNNs for ADR signal detection, filling a gap in post-market surveillance methods.

Findings

Achieved AUROC of 0.795 and AUPRC of 0.775, outperforming other algorithms.

Successfully predicted ADR pairs not listed in existing databases.

Demonstrated the model's potential to supplement current ADR knowledge.

Abstract

Adverse Drug Reaction (ADR) is a significant public health concern world-wide. Numerous graph-based methods have been applied to biomedical graphs for predicting ADRs in pre-marketing phases. ADR detection in post-market surveillance is no less important than pre-marketing assessment, and ADR detection with large-scale clinical data have attracted much attention in recent years. However, there are not many studies considering graph structures from clinical data for detecting an ADR signal, which is a pair of a prescription and a diagnosis that might be a potential ADR. In this study, we develop a novel graph-based framework for ADR signal detection using healthcare claims data. We construct a Drug-disease graph with nodes representing the medical codes. The edges are given as the relationships between two codes, computed using the data. We apply Graph Neural Network to predict ADR signals, using labels from the Side Effect Resource database. The model shows improved AUROC and AUPRC performance of 0.795 and 0.775, compared to other algorithms, showing that it successfully learns node representations expressive of those relationships. Furthermore, our model predicts ADR pairs that do not exist in the established ADR database, showing its capability to supplement the ADR database.