Editorial Policy

Research and first drafts on WhichMedicine are AI-assisted. We use AI to gather source material quickly and structure it clearly, but every article is editorially reviewed by a human before it is published — checked for accuracy, Australian relevance, and plain English before it goes live.

Nothing on this site is medical advice. Our guides, comparisons and symptom checker provide general information only and are not a substitute for advice from a doctor or pharmacist who knows your circumstances. See our full Medical Disclaimer for details.

We source from Australian authorities first: the Therapeutic Goods Administration (TGA), Healthdirect Australia, the Australasian Society of Clinical Immunology and Allergy (ASCIA), and the Australian Department of Health.

Every checkable claim in an article must trace back to a live, authoritative source — we do not publish drug scheduling, dosage, age limits, or availability information we cannot verify. Where a post makes claims like these, it carries a References section listing the sources used.

Before publication, externally checkable medical claims — drug scheduling, doses, age limits, product availability, and first-aid protocols — are verified against authoritative Australian sources (TGA, Healthdirect, ASCIA) and recorded in an internal verified-claims ledger. Each entry notes what was checked, the source, and when it needs to be rechecked, so facts that move over time (like scheduling changes) don't go stale unnoticed.

Articles that carry checkable claims display a "Facts checked" date alongside the last-updated date, so you can see when the medical detail was last verified, not just when the page text was edited.

If you spot an error, please tell us via hello@whichmedicine.com.au or our contact form. We correct errors promptly once verified.

Material corrections — those that change a recommendation, a dose, or a safety detail — are tracked in our internal verified-claims ledger against the specific claim that was wrong. This means a corrected error is flagged if it is ever reintroduced, rather than silently returning.