Self-Supervised Representation Learning for Longitudinal Polypharmacy Patterns

2026-08-06

AI Summary

Longitudinal multi-drug usage data pose significant challenges due to high dimensionality, sparsity, irregular sampling, and structured inter-drug relationships, yet a general-purpose representation learning framework remains lacking. This work proposes LOPEL, a two-stage self-supervised framework: in the first stage, it integrates the Anatomical Therapeutic Chemical (ATC) drug classification hierarchy with Gaussian processes to learn visit-level embeddings; in the second stage, it models the temporal dynamics and uncertainty of individual medication trajectories using Wasserstein distance to generate individual-level representations. LOPEL is the first to unify pharmacological similarity and temporal uncertainty within a self-supervised learning paradigm, yielding generalizable and interpretable embeddings of longitudinal polypharmacy patterns. Evaluated on both simulated and real-world HIV cohorts of older adults, LOPEL accurately recovers latent structures and identifies clinically meaningful heterogeneous subgroups differing in medication timing, composition, and progression, thereby supporting precise risk stratification and clinical decision-making.

Abstract

Polypharmacy, commonly defined as the concurrent use of multiple medications, is increasingly prevalent in aging populations and is associated with adverse health outcomes. Motivated by longitudinal studies of aging people with HIV (PWH), we study how medication use evolves over time and characterizes multimorbidity patterns. Longitudinal medication data present substantial methodological challenges, including high dimensionality, sparsity, irregular observation times, and structured pharmacologic relationships among medications. Existing approaches are typically task-specific and lack a unified framework for learning general-purpose representations of medication trajectories. We propose LOPEL (LOngitudinal Polypharmacy Embedding Learning), a two-stage self-supervised framework for learning low-dimensional representations of longitudinal medication data. In the first stage, visit-level embeddings are learned using a Gaussian process model that incorporates pharmacologic similarity through the Anatomical Therapeutic Chemical hierarchy. In the second stage, subject-level embeddings are constructed by modeling trajectories over time and defining similarity through a Wasserstein-based representation capturing temporal dynamics and uncertainty. Simulation studies demonstrate that LOPEL accurately recovers latent structure under realistic conditions with high-dimensional sparsity and irregular sampling. In an application to aging cohorts of PWH, LOPEL identifies clinically meaningful subgroups that differ in the timing, composition, and progression of medication use, highlighting heterogeneity relevant for risk stratification and clinical management.

Problem

• polypharmacy
• longitudinal data
• representation learning
• multimorbidity
• medication trajectories

Innovation

• self-supervised learning
• longitudinal medication data
• Gaussian process
• Wasserstein distance
• polypharmacy embedding