艾昆纬-解锁有效利用现实世界数据的关键（英）-2025_32页_4mb

报告摘要

Real World Data (RWD): Refers to patient-level data from real-world settings, such as electronic health records, claims, and hospital data. It complements randomized clinical trials (RCTs) by providing insights from diverse settings, enhancing evidence generation for drug development and regulatory decisions. RWD helps address questions RCTs cannot answer, like long-term outcomes and real-world efficacy.
Main RWD Sources:
- EMRs: Electronic medical records from primary and specialist care, useful for patient characteristics, treatments, and outcomes.
- LRx Data: Longitudinal prescription data from pharmacies, tracks drug usage patterns and adherence.
- Hospital Data: Administrative and clinical data from hospitals, covering diagnoses, procedures, and costs.
- Claims Data: Health insurance and medical claims data, useful for cost and utilization analysis.
- Multi-Omics Data: Linked clinical-genomic data, enabling personalized medicine and biomarker studies.
- Registries and Patient Cohorts: Academic or disease-specific data, often for rare disease research.
RWD Uses: RWD can be used throughout the drug lifecycle, from protocol design and safety monitoring to market access and commercial analytics. It supports real-world evidence (RWE) generation, such as for regulatory submissions, demonstrating value in conditions where RCTs are impractical (e.g., rare diseases).
Challenges and FAIR Principles: RWD faces issues like heterogeneity, interoperability gaps, and bias. Adhering to FAIR principles (Findable, Accessible, Interoperable, Reusable) is crucial for effective use. This involves metadata cataloging, standardization, and robust governance to ensure compliance and data reuse, especially with the rise of AI in data analysis.
Innovative Approaches: Technologies like AI/ML enable advanced analytics, including predictive modeling, data enrichment, and pattern recognition. Federated data networks allow collaboration across multi-country studies while maintaining data sovereignty. AI also aids in automating data transformation to analytical-ready data, accelerating evidence generation.
Regulatory Acceptance and Examples: Regulators increasingly accept RWE for regulatory decisions, supported by guidelines from agencies like FDA, EMA, and CHMP. Case studies include multi-country cohort analyses and AI-driven platforms for healthcare insights, with AI-pragmatic governance to ensure ethical use.
Key Takeaways: RWD offers transformative potential for evidence generation, but requires significant investment in data quality, governance, and infrastructure. Its integration with AI and multi-source linking enhances insights, supporting drug development and healthcare decision-making.