Skip to main content
Medical data demands accuracy. Here’s how we maintain it.

​
Confidence Scoring

Every entity and relationship in the ontology carries a confidence level:
LevelMeaning
HighValidated by medical advisor, peer-reviewed sources
MediumCross-referenced from 2+ public databases
LowCommunity-contributed, pending expert review
Developers can filter by confidence level depending on their use case — consumer wellness apps may accept medium confidence; clinical-adjacent tools can restrict to high only.

​
Review Process

  1. Data enters from public medical databases or community contribution
  2. Automated cross-referencing against ICD-10, PubMed
  3. Medical advisor review for clinical accuracy
  4. Confidence score assigned
  5. Published to ontology with provenance metadata

​
Error Reporting

Found an inaccuracy? Our community correction pipeline accepts flagged issues with expert review gates. No unvalidated change reaches the production ontology.

​
Transparency

Every API response will include:
  • Ontology version number
  • Last review date for the returned entity
  • Confidence level
  • Source attribution