Atom Grants
Discover
  • Dashboard
  • Search
  • Deep Research
  • Favorites
  • Projects

    Real-World Evidence Innovation Challenge

    Real-World Evidence Innovation Challenge supports individuals, teams, and entities validating clinical data analytic methods to establish reliable, reproducible performance across health systems.

    Browse similar grants
    AI in Healthcare Research Grants
    OverviewSimilar Grants
    OverviewSimilar Grants

    This grant was recently updated

    NIH replaced the pending prize announcement with a purse of up to $6 million: up to 12 teams will receive $50,000 each for Phase 1. Phase 2 anticipates up to three $1 million, three $600,000, and six $100,000 prizes, with new workflow execution and performance requirements.

    Funder: NIH Challenges

    Due Dates: February 2, 2027, 5:00 PM Eastern (Registration) | March 26, 2027, 5:00 PM Eastern (Phase 1) | September 2027 (Phase 2)

    Funding Amounts: Up to $6,000,000 total; Phase 1 prizes of $50,000; Phase 2 prizes up to $1,000,000

    Summary: A two-phase prize competition validating clinical data analytic methods for reproducibility, transportability, and stability across health systems.

    Key Information: Methods must support reproducible execution within the secure N3C enclave without external dependencies or site-specific modification.


    Description

    NIH’s National Center for Advancing Translational Sciences (NCATS) leads this national prize competition to establish validated benchmarks for analytic methods using real-world clinical data. Eligible approaches include predictive models, causal inference methods, phenotyping algorithms, survival models, risk scores, and other clinically meaningful analytic workflows. Cancer, kidney disease, and pediatrics are priorities, but other use cases are welcome.

    The competition uses the National Clinical Cohort Collaborative (N3C) secure enclave to evaluate methods across independent health systems. Phase 1 assesses methodological readiness; Phase 2 tests performance across individual sites and progressively pooled datasets. The goal is to define Good Algorithmic Practice: measurable expectations for reliable performance despite differences in populations, workflows, and data capture. Validated transportability and stability—not peak performance in one dataset—are central priorities.


    Atom

    See the full grant listing

    Sign in to view full eligibility details, sources, similar grants, and AI-powered analysis.