Why the FDA’s new real-world evidence guidance ends the era of structured-data-only submissions

This excerpt outlines architectural consequences for healthcare data systems: the need to ingest and preserve multiple data modalities (text, FHIR, HL7, DICOM, PDFs), apply healthcare-specific language models that handle negation, assertion status and temporality, maintain auditable terminology mapping, and implement a reconciliation layer to surface conflicts between sources. It emphasizes that while capturing data is straightforward, proving the correctness of each fact is harder and that new guidance makes both capabilities mandatory, requiring sponsors to redesign pipelines.