Working Group Proposal


Overview

This page summarizes the formal proposal to launch the OHDSI Waveform Working Group. The complete proposal document is available in the Resources section of this repository.


Mission Statement

To enable integration of physiological waveform data (e.g., ECG, EEG, arterial blood pressure) with electronic health records in the OMOP Common Data Model, supporting observational research, AI model development, and clinical decision support.


Background & Motivation

The Challenge

Physiological waveform data represents a rich source of information about patient health status, yet this data remains largely siloed from electronic health records used in observational research. Current challenges include:

  • Format Heterogeneity: Dozens of proprietary and open formats make data sharing difficult
  • Temporal Misalignment: Synchronizing continuous waveform data with discrete clinical events
  • Metadata Loss: Signal characteristics often not preserved during data transfers
  • Limited Integration: Waveform data typically stored separately from EHR systems
  • Reproducibility Gaps: Lack of standardized feature definitions across institutions

The Opportunity

The OMOP Common Data Model has successfully standardized clinical data across hundreds of institutions worldwide. Extending OMOP to include physiological waveforms would enable:

  • Multi-Site Studies: Federated research using harmonized waveform representations
  • Integrated Analytics: Seamless analysis combining waveforms with diagnoses, medications, procedures, and outcomes
  • AI/ML Development: Training and validation of predictive models across institutions
  • Reproducible Science: Standardized feature definitions enable cross-institution validation


Proposed Scope

In Scope

The working group will focus on:

  1. Schema Development
    • Define OMOP CDM extensions for waveform data
    • Specify table schemas, relationships, and constraints
    • Ensure compatibility with core OMOP principles
  2. Vocabulary Development
    • Develop concepts for waveform types (ECG, EEG, ABP, etc.)
    • Map signal channels and derived features
    • Coordinate with OMOP vocabulary team
  3. Implementation Guidance
    • Create ETL specifications and best practices
    • Develop reference implementations for common formats
    • Provide data quality and validation frameworks
  4. Tool Development
    • Build ETL utilities and converters
    • Create visualization and analysis tools
    • Integrate with OHDSI HADES packages
  5. Community Support
    • Support institutions implementing the extension
    • Facilitate use case development
    • Enable multi-site collaborative studies

Out of Scope

The working group will NOT:

  • Develop new waveform file formats
  • Provide clinical interpretation services
  • Maintain proprietary device integrations
  • Offer individualized consulting services


Deliverables

Phase 1: Foundation (Months 1-6)

Phase 2: Implementation (Months 7-12)

Phase 3: Adoption (Months 13-18)


Governance Structure

Leadership Team

  • Working Group Lead: Responsible for overall direction and coordination
  • Co-Leads: Support specific workstreams (vocabulary, tools, implementation)
  • Technical Advisory Board: Provide expertise on standards and integration

Collaboration Model

  • Open Participation: All OHDSI community members welcome
  • Consensus-Based Decisions: Major changes require community discussion
  • Transparency: All work conducted in public forums (GitHub, Teams)
  • Use-Case Driven: Development priorities guided by research needs

Meeting Cadence

  • Monthly Working Group Calls: Progress updates and technical discussions
  • Office Hours: Ad-hoc support for implementers
  • Annual Symposium Presentations: Share results with broader OHDSI community


Resource Requirements

Personnel

  • Working Group Leadership: 20% FTE (coordination, specification development)
  • Technical Contributors: Variable (community volunteers + funded developers)
  • Vocabulary Experts: 10% FTE (concept development and mapping)

Infrastructure

  • GitHub Repository: Provided by OHDSI
  • Microsoft Teams: Provided by OHDSI
  • Documentation Hosting: GitHub Pages (free)
  • Compute Resources: Minimal (static website, code repositories)

Funding

  • Initial Development: Supported by CHoRUS Bridge2AI
  • Ongoing Maintenance: Community volunteers + grant-supported developers
  • Travel: Conference attendance for presentations (seek funding as available)


Success Metrics

Year 1

Year 2

Year 3


Risk Mitigation

Risk: Low Adoption

Mitigation: - Focus on high-impact use cases (ICU, cardiology) - Provide comprehensive implementation support - Demonstrate value through multi-site studies - Build easy-to-use tools and templates

Risk: Technical Complexity

Mitigation: - Develop reference implementations - Provide step-by-step guides and video tutorials - Offer office hours and implementation support - Start simple and iterate based on feedback

Risk: Vocabulary Challenges

Mitigation: - Engage OMOP vocabulary team early - Leverage existing standards (LOINC, SNOMED) - Use 2-billion concept range for waveform-specific terms - Plan for gradual vocabulary maturation

Risk: Data Privacy Concerns

Mitigation: - Provide clear de-identification guidance - Support aggregated feature sharing (not raw waveforms) - Align with OHDSI’s existing privacy frameworks - Develop privacy-preserving analysis methods


Alignment with OHDSI Strategy

The Waveform Working Group aligns with OHDSI’s strategic priorities:

1. Expand Data Coverage

Waveform data complements existing OMOP tables, enabling more comprehensive observational research.

2. Enable New Research

Integration of waveforms with EHR data enables novel research questions in critical care, cardiology, neurology, and AI/ML.

3. Foster Collaboration

The working group brings together data engineers, clinicians, researchers, and standards experts.

4. Maintain Standards Excellence

The extension follows OMOP CDM principles and integrates with existing vocabulary and tools.

5. Support Global Health

Standardized waveform data can improve care delivery worldwide, especially in resource-limited settings.


Next Steps

Immediate (Month 1)

  1. Formalize Working Group: Announce formation to OHDSI community
  2. Establish GitHub: Set up repository and initial documentation
  3. Recruit Participants: Invite interested community members
  4. Schedule Meetings: Set up regular working group calls

Short-Term (Months 2-3)

  1. Finalize Specification: Complete 4-table schema based on community feedback
  2. Launch Website: Publish documentation and implementation guides
  3. Develop Vocabulary: Submit initial concept proposals
  4. Identify Early Adopters: Support first implementations

Medium-Term (Months 4-6)

  1. Build Tools: Develop reference ETL and validation utilities
  2. Create Examples: Publish sample datasets and analysis notebooks
  3. Present at Symposium: Share progress with OHDSI community
  4. Expand Participation: Grow working group membership


Conclusion

The OHDSI Waveform Working Group represents a natural evolution of the OMOP CDM to encompass multimodal physiological data. By standardizing waveform integration, we can unlock new research opportunities, enable multi-site collaboration, and ultimately improve patient care through better evidence generation.

We invite the OHDSI community to join us in this exciting effort.


Contact & More Information