Driving · 2026-10-01
EHR2Trace: Auditable EHR Data Infrastructure for Patient World Models and Clinical Agents
Xinye Yang, Yuli Wang, Cheng Ting Lin, Harrison Bai
AI-assisted summary based on the paper. Consult the linked source for methods and limitations. Benchmark numbers appear after editor verification.
TL;DR
EHR2Trace is an auditable EHR conversion pipeline that turns heterogeneous EHR sources into source-linked patient events for OMOP and MEDS exports. It separates event time from information availability, distinguishes medication orders vs dispensing vs administration, and runs automated validation (55 checks). Across three clinical datasets it converted 846.4M events, detected all 28 injected faults, and produced reproducible byte-identical artifacts across build configurations.
Why it matters
Patient world models and clinical agents depend on correct event timing and on knowing what information was available at each decision. The authors report that backdating diagnoses or using later information at earlier prediction times inflates measured performance, and that using availability-filtered histories preserves accuracy under deployment-like evaluation. EHR2Trace provides explicit, inspectable data rules and auditability for those temporal choices.
Method
- Configurable conversion builds a shared patient-event representation with separate event time and availability time, plus medication action types (orders, dispensing, administration) and source links, then exports to OMOP and MEDS.
- Stores audit metadata (extraction dates, cohort membership) and links each exported record back to source files/rows; uses YAML dataset configs for merge/time/unit/terminology rules.
- Runs an automated validation suite (55 checks) comparing saved outputs against source and audit records, with cross-layer checks across ingest, canonical, OMOP, and MEDS artifacts.
Limitation
Evaluation scope is limited to three clinical datasets, a fault catalogue, a reviewed medication reference set, and a retrospective mortality-proxy task; broader evaluation with patient world models, clinical agents, and prospective deployment is suggested. Source completeness and clinical interpretation remain the responsibility of data owners and domain experts.
Abstract (from arXiv)
Patient world models and clinical agents aim to predict changes in patients' health and support clinical work. Developing these systems requires reliable histories of patient conditions, treatments, and the information available at each decision. Electronic health records (EHRs) contain these histories, but differences in how events are recorded make them difficult to use consistently. We present EHR2Trace, a system that converts EHRs from different sources into traceable patient events for model training and evaluation. It links events to source records, separates event time from information availability, and distinguishes medication orders, dispensing, and administration. A shared event representation supports both OMOP and MEDS exports, with automated validation and reproducible builds. Across three clinical datasets, EHR2Trace converted 846.4 million events, with every applicable check passing except one unit-consistency check on MIMIC-IV, and detected all 28 injected faults. A controlled prediction experiment showed that assigning later diagnoses to admission time substantially inflated measured performance, and that a model trained on such data lost accuracy when deployed on histories filtered by availability. EHR2Trace provides a reusable data foundation for patient world models and clinical agents, helping researchers inspect patient histories, check conversion decisions, and evaluate models with explicit data rules.
Related papers
- Forecasting Patient Trajectories from the Electronic Health Records with Action-Conditioned World Models
- PatientWM: Towards Generalizable Clinical World Models via Unified Patient States
- Same Predictions, Different Harms: Causal Auditing of Patient World Models
- Repeated-Measure Leakage, Distribution Shift, and Reliability under Partial Observation in Patient World Models
- Are Joint-Embedding Predictive Objectives Effective for Longitudinal EHRs?
Publication record
Bibliographic evidence describes where and when a record appeared. It does not validate scientific claims.
First public release: 2026-09-18 · Indexed by Pub-AI: 2026-10-01 UTC
EHR2Trace: Auditable EHR Data Infrastructure for Patient World Models and Clinical Agents
Preprint posted: 2026-09-18 · Original record ↗
- arxiv
- 2609.38193
Explicit source versions (1)
- v1 · 2026-09-18
- Preprint available
Source date: 2026-09-18 · First detected: 2026-10-25 UTC · Historical baseline
Source evidence (1)
- Official evidence ↗ · atom:published (official metadata)
Abstract · arxiv
Patient world models and clinical agents aim to predict changes in patients' health and support clinical work. Developing these systems requires reliable histories of patient conditions, treatments, and the information available at each decision. Electronic health records (EHRs) contain these histories, but differences in how events are recorded make them difficult to use consistently. We present EHR2Trace, a system that converts EHRs from different sources into traceable patient events for model training and evaluation. It links events to source records, separates event time from information availability, and distinguishes medication orders, dispensing, and administration. A shared event representation supports both OMOP and MEDS exports, with automated validation and reproducible builds. Across three clinical datasets, EHR2Trace converted 846.4 million events, with every applicable check passing except one unit-consistency check on MIMIC-IV, and detected all 28 injected faults. A controlled prediction experiment showed that assigning later diagnoses to admission time substantially inflated measured performance, and that a model trained on such data lost accuracy when deployed on histories filtered by availability. EHR2Trace provides a reusable data foundation for patient world models and clinical agents, helping researchers inspect patient histories, check conversion decisions, and evaluate models with explicit data rules.