A Framework for Evaluating Synthetic Electronic Health Records.

Area of Science:

• Health Informatics
• Data Science
• Biostatistics

Background:

• Synthetic data generation for Electronic Health Records (EHRs) enhances patient privacy.
• A wide variety of evaluation methods exist for synthetic data, leading to challenges in consistent assessment.
• Current methods often fail to preserve variable dependencies and temporal dynamics in synthetic EHRs.

Conclusions:

• A standardized and comprehensive evaluation framework is crucial for assessing synthetic EHR data quality.
• The proposed framework addresses the limitations of current methods by incorporating variable dependency and temporal analysis.
• This work facilitates more reliable and trustworthy synthetic EHR data generation and utilization.