Why Health Information Is Broken; and What We Do Differently
There are two categories of health information available to patients today: overwhelming clinical papers written for specialists, and oversimplified content written for traffic. AI chatbots have added a third category: confident-sounding answers that are demonstrably and dangerously wrong.
The AI accuracy problem (2025–2026 research data): A Yale University study found that AI chatbots ordered unnecessary laboratory tests in 91.9% of chronic disease scenarios and prescribed potentially inappropriate medications in 57.8% of cases. Stanford researchers found severe errors in 22% of AI-generated medical answers. ECRI (the global patient safety organisation) ranked AI chatbot misuse as the #1 health technology hazard for 2026. Google was forced to remove AI-generated health summaries in January 2026 after they were found to provide dangerous misinformation about liver disease.
Lucent Guides exists to fill the gap between academic clinical guidelines (which patients cannot read) and AI-generated content (which patients cannot trust). Every guide we produce is traceable to specific, named, peer-reviewed international guidelines; the same documents that govern specialist clinical practice worldwide.
How Each Guide Is Created
-
1Identify the authoritative international guidelines for the condition. We prioritise the highest-impact guideline-producing organisations: ACC/AHA for cardiovascular, NICE for general medicine, ESC for European cardiology, WHO for global standards, KDIGO for nephrology, EAU for urology, and so on. Multiple guidelines are synthesised for each guide.
-
2Extract the evidence-based key clinical facts; the specific recommendations, evidence levels, threshold values, and treatment targets that the guidelines define. Only Level A (highest) and Level B evidence is used for primary recommendations. The guideline source is cited on every guide.
-
3Distil to plain language at approximately eighth-grade reading level, while preserving clinical precision. Medical jargon is explained on first use. Numbers and thresholds are kept specific; vague language ("may help") is never used where the evidence is clear.
-
4Structure for action. Every guide includes: what the condition is, how it is diagnosed, what the evidence-based treatments are, what the patient should ask their doctor, and what warning signs require urgent care. The goal is a patient who arrives at their next appointment better prepared, not a patient who avoids their physician.
-
5Cite every claim. Every guide lists the specific international guidelines and trial citations on which it is based, with the year the guideline was published or updated. When major guidelines change, we update the guide.
The Organisations Behind Our Guides
Our guides synthesise recommendations from over 55 international clinical guideline-producing organisations. These represent the consensus of tens of thousands of specialist physicians across decades of clinical research.
Why Not Just Use an AI Chatbot?
According to OpenAI, over 230 million people ask ChatGPT health and wellness questions every week. This is both the scale of the need and the scale of the problem. AI chatbots produce fluent, confident answers; but those answers are not reliably accurate for medical content.
| Feature | AI Chatbots (ChatGPT, Gemini etc.) | Lucent Guides |
|---|---|---|
| Evidence source | ✗ Opaque training data; not traceable | ✓ Named international guidelines cited on every guide |
| Accuracy verification | ✗ No systematic fact-checking; confabulation is common | ✓ Synthesised directly from peer-reviewed guideline documents |
| Currency | ✗ Training data has a cutoff; may miss recent guideline updates | ✓ Updated when major guidelines change; date shown on every guide |
| Printable / shareable | ✗ Session-based; cannot easily be shared or printed | ✓ Downloadable PDF designed for printing and sharing at appointments |
| Regulatory status | ✗ Not regulated as a medical device (ECRI 2026) | ✓ Educational content with explicit guideline attribution |
| Tailored to patient education | ✗ Generic answers; not structured around patient decision-making | ✓ Every guide includes "Questions to Ask Your Doctor" section |
| Available offline | ✗ Requires internet and account | ✓ PDFs work offline; PWA works offline after first visit |
What Lucent Guides Is Not
Lucent Guides is patient education material. It is not a substitute for professional medical advice, diagnosis, or treatment. Clinical guidelines describe best practice for typical presentations; your physician's judgement about your individual circumstances is always the governing factor in your care.
Guides reflect the evidence as of the date shown. Medicine evolves: new trials are published, guidelines are updated, and recommendations change. We update guides when major guideline changes occur, but there will always be a lag between new evidence and updated educational content.
Our guides do not cover all conditions, all medications, or all clinical scenarios. They are a starting point for informed conversations with your healthcare provider; not an endpoint.
Questions About Our Methodology
If you believe a guide contains a factual error, has been superseded by new guideline evidence, or could be improved, please contact us at hello@lucentguides.com. We take accuracy corrections seriously and will review and update guides promptly when credible concerns are raised.