How this was built

The rule everything follows

Every factual claim on this site links to a named source. If a statement cannot point at one, it does not go on the page.

Where sources disagree, the site says so rather than picking the tidier answer.

Where the information comes from

Sources are public, free to check, and either government-run, peer-reviewed or otherwise accountable.

• US government health sources for the condition, its genetics and how it is diagnosed — NINDS, the Genetic and Rare Diseases Information Center, and MedlinePlus Genetics.
• GeneReviews for detail on individual genetic types. Expert-authored, peer-reviewed, written for clinicians.
• ClinicalTrials.gov for studies, queried live. Status, dates and enrolment come straight from the registry.
• Europe PMC for recent published literature, also queried live.
• openFDA and Drugs@FDA for what is and is not approved. The treatments page queries the FDA label database live rather than asserting the answer.
• Company regulatory filings for drug programme status — filings preferred over press releases, because filings carry legal accountability that press releases do not.
• Patient organisations for anything about daily life and support.

Every source is listed individually, labelled by what kind of evidence it is — because a government database and a company announcement are not equivalent, and it matters which one a claim rests on.

How sources were weighed

Three habits, applied consistently.

• Primary over secondary. A regulator’s own document beats a news story about that document. A filing beats a press release about the filing.
• Results and regulatory status travel together. Trial numbers never appear without what the regulator concluded about them, because the second thing changes how to read the first.
• Sponsor claims are labelled as sponsor claims. Registry descriptions are written by the organisations running the studies. Their dates are dependable; their enthusiasm is not evidence.

Where this could be wrong

A site claiming reliability without naming its failure modes is less trustworthy, not more. So:

• It goes out of date. Regulatory decisions and trial statuses change. The live feeds refresh automatically; the written pages do not. Check the date on anything that matters to you.
• No clinician has reviewed it. The sources are reliable. The summarising was done by someone without medical training, so errors of emphasis and nuance are possible.
• It is US-weighted. The FDA data and most of the regulatory material are American. Approvals, care pathways and available support differ by country.
• It generalises across dozens of conditions. Something true of the SCAs as a group may not hold for a specific type. For one type, GeneReviews and MedlinePlus Genetics are better than this.
• It is not a substitute for the people who do this properly. Patient organisations have decades of accumulated knowledge and actual communities. Go and find them.

Nothing is collected

People researching a health condition are in a vulnerable position, and this site has no business knowing anything about them.

• No accounts, sign-ups, forms, comments or email collection.
• No advertising, no advertising cookies, no third-party tracking scripts.
• Live data from ClinicalTrials.gov, Europe PMC and the FDA is fetched by the server when pages are built, not by your browser. Those organisations never see your visit.
• Videos link out rather than embedding, because an embedded player would load trackers into the page.
• Aggregate page-view counts are recorded through cookieless analytics — how many people read a page, never who they are.

Because no personal data is collected, there is nothing to request, correct or delete. That is the design, not an oversight.

Corrections

If something here is wrong it should be fixed, and that goes double for anything a family might act on. Every claim links to its source, so an error should be traceable — and if a claim cannot be traced to a source, treat that as a fault in the page.