HADES

HADES performs large-scale analytics on electronic health records and administrative claims converted to the Observational Medical Outcomes Partnership (OMOP) Common Data Model to enable population characterization, population-level causal effect estimation, and patient-level prediction across federated data networks.

Key Features:

• OMOP Common Data Model requirement: Requires source data to be converted into the Observational Medical Outcomes Partnership (OMOP) Common Data Model for analysis.
• Population Characterization: Enables comprehensive analysis of demographic and clinical characteristics at scale.
• Population-Level Causal Effect Estimation: Provides analytical tools for estimating causal effects within populations.
• Patient-Level Prediction: Supports predictive modeling at the individual patient level.
• Federated data network operation: Executes analyses across federated networks while keeping patient-level data localized and sharing only aggregated statistics between nodes.
• Data types supported: Processes electronic health records and administrative claims data.
• Software quality practices: Employs continuous integration and a suite of unit tests to maintain software reliability.
• Origin: Developed and maintained by Observational Health Data Sciences and Informatics (OHDSI).

Scientific Applications:

• Evidence generation from EHR and claims: Enables large-scale analyses of electronic health records and administrative claims for observational evidence generation.
• Support for published research: Has been used in published OHDSI studies.
• Regulatory and policy impact: Has contributed to analyses that influenced regulatory decisions and can inform patient care and public health policy.

Methodology:

Requires data conversion to the OMOP Common Data Model and implements population characterization, population-level causal effect estimation, and patient-level prediction methods, with support for federated execution that shares only aggregated statistics between nodes.