Foundation Models to Unlock Real-World Evidence from Nationwide Medical Claims

2026-05-04

AI Summary

This work proposes ReClaim—the first generative foundation model for healthcare trained on nationwide-scale insurance claims data—designed to extract generalizable clinical and economic insights from over 200 million beneficiaries and 43.8 billion medical events spanning 2008–2022. Built on a from-scratch Transformer architecture (140M–1.7B parameters), ReClaim models longitudinal trajectories of diagnoses, procedures, medications, and expenditures, enabling multitask disease prediction and real-world evidence generation. Across more than 1,000 disease prediction tasks, it achieves an average AUC of 75.6%, outperforming LightGBM and Delphi by 9.3 and 6.2 percentage points, respectively. It also improves expenditure prediction, raising explained variance from 0.28 to 0.37, and reduces average systematic bias in target trial emulation by 72%, substantially surpassing existing approaches.

Abstract

Evidence derived from large-scale real-world data (RWD) is increasingly informing regulatory evaluation and healthcare decision-making. Administrative claims provide population-scale, longitudinal records of healthcare utilization, expenditure, and detailed coding of diagnoses, procedures, and medications, yet their potential as a substrate for healthcare foundation models remains largely unexplored. Here we present ReClaim, a generative transformer trained from scratch on 43.8 billion medical events from more than 200 million enrollees in the MarketScan claims data spanning 2008-2022. ReClaim models longitudinal trajectories across diagnoses, procedures, medications, and expenditure, and was scaled to 140 million, 700 million, and 1.7 billion parameters. Across over 1,000 disease-onset prediction tasks, ReClaim achieved a mean AUC of 75.6%, substantially outperforming disease-specific LightGBM (66.3%) and the transformer-based Delphi model (69.4%), with the largest gains for rare diseases. These advantages held across retrospective and prospective evaluations and in external validation on two independent datasets. Performance improved monotonically with scale, and post-training added 13.8 percentage points over pre-training alone. Beyond disease prediction, ReClaim captured financial outcomes and improved real-world evidence (RWE) analyses: for healthcare expenditure forecasting it increased explained variance from 0.28 to 0.37 relative to LightGBM, and in a target trial emulation it reduced systematic bias by 72% on average relative to Delphi. Together, these results establish administrative claims as a scalable substrate for healthcare foundation models and show that learned representations generalize across time periods and data sources, supporting disease surveillance, expenditure forecasting, and RWE generation.

Problem

• real-world evidence
• healthcare foundation models
• administrative claims
• disease prediction
• medical expenditure forecasting

Innovation

• foundation model
• real-world evidence
• medical claims
• transformer
• disease prediction