Mapping Emergency Medicine Data to the Observational Medical Outcomes Partnership Common Data Model: A Gap Analysis of the American College of Emergency Physicians Clinical Emergency Data Registry

Inessa Cohen, Zihan Diao, Pawan Goyal, Aarti Gupta, Kathryn Hawk, Bill Malcom, Caitlin Malicki, Dhruv Sharma, Brian Sweeney, Scott G. Weiner, Arjun K. Venkatesh, Richard A. TaylorJournal of the American College of Emergency Physicians Open 2025 · DOI

TL;DR

This paper tackles the challenge of standardizing emergency medicine data by analyzing the compatibility between the American College of Emergency Physicians' Clinical Emergency Data Registry (CEDR) and the Observational Medical Outcomes Partnership Common Data Model (OMOP-CDM). The findings show high compatibility, suggesting OMOP-CDM is a viable framework for enhancing emergency medicine research and analytics.

What Problem Does This Paper Address?

The rapid growth of electronic health records (EHRs) presents a massive opportunity to improve healthcare surveillance, identify disparities, and advance research. However, a major hurdle is the lack of standardized data formats, which hinders data sharing, interoperability, and efficient analysis across different institutions and research platforms. The Clinical Emergency Data Registry (CEDR), a national database for emergency medicine, aims to harness this data but faces limitations due to its non-standardized schema. To overcome this, the authors investigate the feasibility of mapping CEDR data to the Observational Medical Outcomes Partnership Common Data Model (OMOP-CDM), a widely adopted standard for observational health data research. This mapping is crucial for enabling CEDR to participate in larger research networks, facilitating automated queries, and improving the overall utility of emergency medicine data for research and public health initiatives.

Methods

The researchers employed a structured gap analysis approach. They created a comparison matrix to systematically map fields from the CEDR schema to the OMOP-CDM schema. Each CEDR field was evaluated for its compatibility with OMOP-CDM, categorizing it into one of three types:

  • Direct Match: The CEDR field has a direct equivalent in the OMOP-CDM based on both description and data type.
  • Requires Transformation: The CEDR field has a similar description but needs conversion (e.g., data type change, value recoding, or combining fields) to align with OMOP-CDM standards.
  • No OMOP-CDM Equivalent: The CEDR field does not have a corresponding element in the OMOP-CDM schema.

The mapping process was informed by consultations with the Observational Health Data Sciences and Informatics (OHDSI) community forums and existing documentation. For ambiguous fields, the OHDSI forums and the Athena tool (for querying OMOP-CDM vocabularies) were used to ensure consistency. The analysis involved descriptive statistics and the creation of a Sankey diagram to visualize the flow and mapping between CEDR and OMOP-CDM categories.

Datasets

The study utilized the Clinical Emergency Data Registry (CEDR) schema as the source data. The target schema for mapping was the Observational Medical Outcomes Partnership Common Data Model (OMOP-CDM), specifically version 5.4. No new datasets were introduced; the focus was on schema-level comparison.

Key Insights and Ideas

  • High Compatibility is Achievable: The most significant insight is that a substantial portion of emergency medicine data, as captured by CEDR, can be effectively standardized using the OMOP-CDM. This validates OMOP-CDM as a promising framework for this domain.
  • Transformations are Necessary but Manageable: While direct matches are common, a significant number of fields require transformation. This highlights the need for robust ETL (Extract, Transform, Load) processes and a clear understanding of data nuances.
  • Unique Data Elements Pose Challenges: Certain fields, particularly those related to protected health information (PHI) like names and social security numbers, do not have direct equivalents in OMOP-CDM due to its de-identification focus. This necessitates careful handling and potential exclusion or anonymization strategies.
  • Context is Key for Mapping: The study emphasizes that mappings are rarely one-to-one. Fields from a single CEDR table can map to multiple OMOP-CDM tables depending on the specific context, underscoring the complexity of data harmonization.
  • Community Collaboration is Vital: The reliance on OHDSI forums and tools like Athena demonstrates the importance of community-driven efforts in standardizing data and establishing best practices.

Key Results

  • Over 90% of CEDR fields (244 out of 269) could be successfully mapped to the OMOP-CDM.
  • Of these, 173 fields were direct matches.
  • 71 fields required transformations.
  • Approximately 9.5% (25 out of 269) of CEDR fields had no equivalent in the OMOP-CDM. These primarily included PHI and certain billing details.
  • Tables like "Plan of care" and "Vital signs" showed 100% direct mapping.
  • Tables like "Medication" and "Location" had a higher proportion of fields requiring transformations.

Limitations and Future Work

Limitations:

  • Schema-Based Analysis: The gap analysis was based on schema comparisons, not actual patient-level data. This limits the ability to capture the full complexity and nuances of real-world data, potentially leading to under- or overestimations of compatibility.
  • Inherent Limitations of Comparison: The comparison of field descriptions might not identify all possible matches or required transformations, as some fields can be versatile.
  • Operational/Technical Challenges: The study did not assess the practical difficulties or technical requirements for implementing these transformations.
  • Generalizability: The findings may not be directly generalizable to other emergency medicine databases or different healthcare disciplines due to unique schema structures and content.
  • Schema Evolution: The analysis is based on specific versions of CEDR and OMOP-CDM; future updates could impact relevance.
  • Feasibility of Implementation: The study did not assess the resources (time, expertise, financial investment) required for implementing a standardized data schema across multiple emergency departments.

Future Work:

  • Pilot Implementation: The authors suggest piloting the OMOP-CDM schema in emergency departments contributing to CEDR to validate the mapped fields with real-world data and ensure data integrity and clinical relevance.
  • Refining Mapping Process: Continued collaboration with the OHDSI community is essential for refining the mapping process, addressing specialty-specific data needs, and extending the OMOP-CDM schema.
  • Phased Implementation: A phased implementation strategy, potentially including an intermediate data mart, is recommended to manage scalability and the practicality of standardizing diverse data sources.
  • Developing Tailored Solutions: Addressing identified gaps will require developing specific solutions for emergency medicine data and potentially creating informatics community-driven extensions to OMOP-CDM.

Summary updated

Abstract

The OMOP-CDM presents a promising framework for standardizing emergency medicine data, thereby enhancing future query automation, analytics, and cross-institutional collaboration. Despite the potential challenges in capturing unique CEDR fields and addressing necessary transformations, most emergency department data can be standardized within the OMOP-CDM, fostering broader insights and applications in research and public health.
BibTeX
@article{cohen2025mapping,
  author = {Inessa Cohen and Zihan Diao and Pawan Goyal and Aarti Gupta and Kathryn Hawk and Bill Malcom and Caitlin Malicki and Dhruv Sharma and Brian Sweeney and Scott G. Weiner and Arjun K. Venkatesh and Richard A. Taylor},
  title = {{Mapping Emergency Medicine Data to the Observational Medical Outcomes Partnership Common Data Model: A Gap Analysis of the American College of Emergency Physicians Clinical Emergency Data Registry}},
  journal = {Journal of the American College of Emergency Physicians Open},
  year = {2025},
  doi = {10.1016/j.acepjo.2024.100016},
}
Figure