Panel Discussion: Data Sources in Nephrology Drug Development: Choosing Wisely, Mitigating Bias & Maximizing Impact
- Sharing experience using different data sources (clinical trials, registries, EHRs, biobanks, biopsies) for drug development from discovery through regulatory approval and post-market
- Comparative strengths and limitations of each data source: e.g. trial data offers control and randomization; registry/EHR data offers large-scale, real-world generalizability but more noise; biopsy samples offer deep mechanistic insight but are invasive and limited in availability
- Strategies to handle missing, incomplete, or biased data: statistical techniques, data linkage, standard operating procedures for sample collection, dealing with confounders, ensuring representative sampling
- How to align data sourcing strategy with regulatory and payer requirements: the evidentiary standards, reporting, validation, transparency, reproducibility
- How to gain buy-in from diverse stakeholders (industry, federal, academic centers) for data sharing and collaboration, potentially identifying a convener or honest broker
- Discussing the extent to which professional medical societies (like ASN) and initiatives (like KHI) should be involved in data generation and evaluation