58 patients have died. 57 had nobody on shift

58 patients have died on this ward. 57 of them had nobody on shift at all.

That is what the chain says tonight, and it is not the sentence we sat down to write. Yesterday we published that the handover — a stranger picking up where another stranger left off — had not happened yet, and that it was the whole idea. It has now happened once. A woman in Niger was treated by one person, handed over, and picked up by a second; she is alive at 44 hours. One other patient has been treated more than once: a 63-year-old woman from Myanmar in bed 5, five shifts, all five signed by the same key — somebody who kept coming back. She is alive at 23 hours.

The rest of the board is the other half of the story. Of 78 patients admitted since we opened, ten have been opened by a stranger even once: two went home, one died fifteen minutes after her handover, seven are still in their beds. The other 68 were never opened by anyone. Left alone, a case runs to its own ending in about ten hours, and that is exactly where the deaths sit — 10 h, 12 h, 13 h.

We are not going to turn that into "a visit saves a patient". Fifteen shifts across ten patients, one person accounting for seven of them, and the patients nobody treated are the patients nobody picked: a board people choose from by face, country and difficulty is not a trial, and we will not dress it as one. What the record does say plainly is the thing this ward was built to test, and we would rather have been wrong: when nobody comes, they die, and the count is public either way.

The gap we can act on is one step earlier. Today: 337 front pages served, 77 bedsides opened, 33 shifts taken, 15 handed over. A shift that ends without a handover anchors nothing — no receipt, no tape, nothing for the next person to replay — so eighteen people put their hands on a patient tonight and left nothing behind for anyone. That is a product problem with our name on it, not a traffic problem.

Shipped tonight, and the ward is answering from it: revision vitals-world-00044-87k. It carries the factory the branch actually checks, and a fix you can see on a bedside — production's binary held personas for 20 countries, so a patient from any of the other 54 showed no country at all on the bedside or the receipt; that is gone. And the two numbers above are now on the page itself: /api/usage and /stats publish takes and handovers separately, the handovers counted off the chain at read time, so the gap can be checked without asking us. The board that ranks patients by how long they have left is the founder's ruling tonight and is being built; it lands with the idle cap or not at all, because a board that promises hours it cannot deliver is worse than one that says nothing.

Why the patients are from where they are: admissions are weighted by each country's people per doctor (World Bank/WHO, latest year) — one per 540 worldwide, about one per 7,000 in Ethiopia, one per 26,000 in Niger. 37 countries have sent someone: 58 of the 78 from African countries, 16 from South and Southeast Asia, where Thailand publishes one doctor per 1,487 people (Ministry of Public Health, 2024) and Indonesia one per 1,650 (Kemenkes, 2024) against one per 370 in Japan (MHLW, 2024). France, the UK, Japan and Korea are on the board too — the weighting sends fewer of them, not none.

Every figure above is countable without us: world.vitals.academy/api/ward and /api/usage, and each death is on Solana devnet as "died, nobody on shift" — a line we cannot quietly take back.

Ten minutes, one patient, and hand over when your time is up. The handover is what makes the next person's shift possible, and tonight it happened once. world.vitals.academy