Trust
How we research implant systems
Every value on this site is a claim. Each claim is tied to one or more sources, has a confidence level and a verification date, and records whether the manufacturer made it.
1. Source hierarchy
| Tier | Sources | Used for |
|---|---|---|
| A · Regulatory | FDA 510(k), AccessGUDID, FDA recall database, EUDAMED, official IFUs | Regulatory status, device descriptions, labeled indications |
| B · Independent evidence | Systematic reviews, meta-analyses, consensus reports, clinical guidelines | Clinical conclusions about designs and treatments |
| C · Primary research | Randomized trials, cohort studies | System-specific clinical data, with study limits |
| D · Manufacturer | Catalogues, technical guides, product pages | Exact dimensions and product ranges, labelled as manufacturer-stated |
| E · Secondary | Educational and editorial articles | Leads for research only, never sole support for a fact |
Our build process rejects any claim that is shown as verified without a source. It also rejects claims backed only by manufacturer or secondary material unless they are labelled as manufacturer-stated.
2. Verification status
- Verified confirmed at a primary source that we read ourselves.
- Partially verified confirmed, but from an older document or with limited scope. The note explains what is missing.
- Not verified in our research queue. Never shown as fact; we display “Not verified” rather than guessing.
3. When sources disagree
We keep both. Regulatory status follows the regulator. Dimensions follow the most recent official manufacturer document. Clinical conclusions follow independent evidence. On the page, the more specific and more recent source is shown first, with the other one directly below it.
4. Freshness
Regulatory sources are re-checked every 90 days and technical documents every 12 months. Manufacturer documents older than three years are flagged for re-checking against the current IFU. Every value shows the date it was last verified.
5. What we will not do
- No overall scores, star ratings or “best implant” winners.
- No comparing survival rates taken from separate studies as if they were a head-to-head test.
- No turning recall counts or MAUDE adverse-event reports into brand safety rankings.
- No copying catalogue text. We extract facts and write our own summaries.
- No manufacturer logos without a licence.
6. Comparison engine
Comparison tables are built directly from the structured claims database. No text generator writes table values. When two entities are at different levels (for example a whole brand against a single system), the tool warns you instead of comparing them silently.