High-Fidelity Longitudinal Patient Simulation Using Real-World Data

TL;DR

This paper introduces a generative model trained on over 200 million clinical records to simulate realistic patient trajectories, enabling personalized treatment planning and virtual trials with high fidelity.

Contribution

The study presents a novel, scalable framework that leverages real-world clinical data to empirically model and generate detailed patient timelines.

Findings

High-fidelity future patient trajectories closely match real data.

Accurate estimation of future event probabilities with ratios near 1.0.

Model trained on 200 million records demonstrates scalability and realism.

Abstract

Simulation is a powerful tool for exploring uncertainty. Its potential in clinical medicine is transformative and includes personalized treatment planning and virtual clinical trials. However, simulating patient trajectories is challenging because of complex biological and sociocultural influences. Here, we show that real-world clinical records can be leveraged to empirically model patient timelines. We developed a generative simulator model that takes a patient's history as input and synthesizes fine-grained, realistic future trajectories. The model was pretrained on more than 200 million clinical records. It produced high-fidelity future timelines, closely matching event occurrence rates, laboratory test results, and temporal dynamics in real patient future data. It also accurately estimated future event probabilities, with observed-to-expected ratios consistently near 1.0 across diverse outcomes and time horizons. Our results reveal the untapped value of real-world data in electronic health records and introduce a scalable framework for in silico modeling of clinical care.