Who publishes this

Drug Eye Africa is a register of published regulatory actions on medicines and medical devices in Africa. The regulatory action is the unit of record; an article is a view of it.

Every record links the primary regulator notice, or states explicitly that no published notice could be located. There is no third state.

Records are superseded, never deleted. A corrected record keeps its URL and its reference number, and the old version stays visible in its version history.

• It is not a product-safety database. An absence of records for a product means no authority we monitor has published an action against it — nothing more.
• Manufacturer profiles count published regulatory actions. They are not a finding about a company.
• Our own indicative severity, where present, is labelled as ours and kept separate from any classification the authority issued.
• Coverage is incomplete and the gaps are published — see the coverage matrix.

A register is easy to start and hard to keep honest. Collecting notices is the straightforward part; deciding what a notice actually says is where the Drug Eye Africa team has spent its time.

• 616 batch numbers read out of PDFs, scanned circulars and third-party-hosted letters, one record at a time, and corroborated before any of them reached a page. A lot number is what sends a pharmacist to check a shelf, so none is published because software believed it saw one.
• 297 translations of notices published in Arabic. The original text is stored word for word, the translation sits beside it, and anything produced by a machine says so on its face.
• 266 records carry a month rather than a day, because Egypt’s authority stopped printing the day in 2024. Each of those dates was read off the circular by a person and stored at the precision the authority actually published — not rounded to a day we would be inventing.
• All 55 African states have a row on the coverage matrix whether we hold anything for them or not, and every row carries the date somebody last looked at that authority’s site.
• All 742 records carry an archived copy of the document they came from and its SHA-256. Regulator links rot; several of these already point at pages that have moved.

The collection runs itself every day. The judgement does not, and is not allowed to: the pipeline is forbidden from publishing anything it extracted on its own. Reading a batch number out of a document is a claim about a medicine, so someone on the Drug Eye Africa team signs off every one. That limit is deliberate, and it is the reason this register grows at the pace it does.

The register came first because everything else depends on it. Three things are open and deliberately thin rather than padded out.

• The newsroom — articles are written from records, never the other way round, and most records will never become one. It is open and has published nothing yet.
• The blog — signed opinion from people who work in African medicines regulation, pharmacy, manufacturing and supply. The standards went up before the first piece, which is the only order in which a standards page means anything. Applications are open.
• Country pages — one for every state, which will deepen as coverage does. Where something is contested we leave it empty rather than guess: regional-bloc membership is blank on every country page for that reason, because a wrong badge on a country hub is the first thing a regulator would notice.

The named editor, the editorial advisory board and the registered legal person are not yet settled, and this page will not name anyone before they are. An MOU signed by “the Newsroom” would not clear a regulator’s internal review, and a placeholder name here would be worse than an empty section.