About Health Schema
The Problem We're Trying To Solve
Schema.org provides excellent vocabulary for tagging basic medical entities—conditions, treatments, symptoms, risk factors. But when we started marking up medical content at Creative Touch, we quickly hit a wall.
Medical content isn't just a list of entities. It's about relationships:
- Causal mechanisms: "Drug A inhibits protein B, which causes symptom C through pathway D"
- Temporal progressions: "At 2 months you'll experience X, at 6 months expect Y, full recovery by 18 months"
- Conditional logic: "If ferritin < 70 AND female AND postmenopausal, THEN risk increases 14-fold"
- Multi-factorial causation: "Condition X results from factors A + B + C, which interact synergistically"
Schema.org can say "drug X has risk factor Y." It cannot capture the how or why. It cannot represent multi-step biochemical pathways, stage-based timelines, or complex decision trees for population-specific recommendations.
The gap: Medical knowledge isn't just vocabulary—it's the causal chains, temporal dynamics, and conditional logic that connect concepts. This is exactly what large language models need to truly understand medical content, not just match keywords and entities.
Important note: We've researched schema.org's medical vocabulary extensively and believe these gaps are genuine. However, we stand on the shoulders of the brilliant founders and contributors behind schema.org, and we welcome feedback if there are existing properties or patterns we've overlooked. Please let us know if you're aware of ways to achieve this semantic richness using standard schema.org alone.
Our Solution
Health Schema extends schema.org with types and properties specifically designed to capture:
- Mechanistic pathways through
CausalPathwayandCausalStep - Time-based progressions through
TemporalProgressionandTemporalStage - Complex causation through
MultifactorialConditionandFactorInteraction - Conditional recommendations through
ConditionalRecommendationand population criteria types
These extensions don't replace schema.org—they augment it. We use standard schema.org properties wherever possible, and introduce custom vocabulary only where it adds genuine semantic value.
Why This Matters
Most schema markup helps search engines identify what entities appear on a page. Health Schema helps AI systems understand how medical knowledge works:
For Large Language Models
Rich semantic relationships enable genuine comprehension. An LLM can trace causal pathways, understand temporal dynamics, assess evidence quality, and provide population-specific guidance—not just keyword matching.
For Knowledge Graphs
Structured representation of medical knowledge enables sophisticated reasoning, relationship discovery, and integration across sources. Build knowledge graphs that capture how things work, not just what things are called.
For Content Creators
Express the full depth of medical content semantically. Show the mechanisms behind treatments, the timelines of conditions, the interactions between factors—structured in a way machines can process.
For Healthcare Sites
Demonstrate technical sophistication and authority. Rich semantic markup signals that your content is structured, evidence-based, and professionally curated.
Development History
Health Schema emerged from real-world need. While marking up medical and aesthetic treatment content at Creative Touch, we encountered explanations that schema.org simply couldn't represent:
- Multi-step biochemical mechanisms (e.g., how GLP-1 medications trigger hair loss through cortisol → Gas6 inhibition → follicle dormancy)
- Treatment timelines with distinct phases (e.g., hair loss progression with 2-4 month onset, 3-6 month peak, 6-18 month recovery)
- Risk stratification with complex criteria (e.g., postmenopausal women with low ferritin and rapid weight loss)
- Multi-factorial conditions with interacting causes (genetic + hormonal + nutritional + environmental factors)
Rather than settle for inadequate markup, we built extensions that could properly represent these relationships. After testing on our own content, we're releasing it as an open-source project for the wider community.
Design Philosophy
Health Schema follows clear principles:
- Schema.org first: Use standard schema.org vocabulary whenever possible. Custom extensions only when necessary.
- LLM-optimized: Designed specifically to enable AI systems to understand medical content deeply, not just process it superficially.
- Evidence-based: Properties for evidence levels, citations, and confidence enable proper epistemic representation.
- Practical: Developed through real-world implementation on live medical content.
- Convention-aligned: Follows schema.org's design patterns and naming conventions for easy adoption.
- Open source: Free to use, modify, and extend under Creative Commons licensing.
Who We Are
Health Schema was created by Creative Touch, a UK aesthetics clinic committed to providing evidence-based information to clients and the wider aesthetics and medical community.
The project reflects our belief that medical information online should be:
- Structured in ways machines can understand
- Rich in semantic relationships, not just keywords
- Evidence-based and properly cited
- Accessible to both humans and AI systems
Current Status
Version: 0.8 (December 2025)
Status: Active development. The vocabulary is being refined based on real-world implementation.
License: Creative Commons Attribution 4.0 International (CC BY 4.0)
Using Health Schema?
We'd love to hear about it! Share your implementation, ask questions, or suggest improvements:
Future Development
Planned enhancements include:
- Additional types for drug interactions and adverse event modeling
- Enhanced representation of diagnostic reasoning and differential diagnosis
- Support for intervention comparisons and treatment decision trees
- Community feedback integration and vocabulary refinement
- Validation tools and implementation guides
This is an evolving project. As we discover new patterns in medical content that need semantic representation, we'll extend the vocabulary accordingly.
Get Involved
Health Schema is open source and community-driven. Ways to contribute:
- Use it: Implement Health Schema in your medical content and share your experience
- Provide feedback: Tell us what works, what doesn't, and what's missing
- Suggest extensions: Identify new patterns that need semantic representation
- Share examples: Show us your implementations so we can learn and improve
- Report issues: Help us identify ambiguities or gaps in the vocabulary
Contact us at contact@healthschema.org to discuss collaboration, contributions, or just to say hello.
Acknowledgments
Health Schema builds upon the excellent foundation of schema.org, maintained by a community of contributors from major search engines and semantic web organizations. We're forever grateful for their work in establishing standards for structured data on the web.