Bed occupancy is counted at midnight, by a nurse walking the ward with a screen, and it drives everything from staffing to whether the hospital declares a critical incident. Ward 6 reports twenty two of twenty two with two empty beds in it.
I am an information manager for an acute trust. I do not treat anybody. I make sure the numbers the trust reports are the numbers the trust has.
The midnight count is the one that matters most.
At midnight every ward's nurse in charge confirms occupancy on a screen: which beds have a patient in them, which are empty and available, and which are empty and closed for a reason like an infection or a broken bed.
It is not a form somebody fills in later. It is done at the bedside, on the ward, from the board, and it flows into the trust's operational picture within minutes.
That number decides the next day. Whether elective surgery goes ahead. Whether the emergency department can move anybody. Whether we tell the region we are in trouble.
Ward 6 is a twenty two bed medical ward.
Since about 2019 it reports twenty two occupied at midnight far more often than any other ward in the hospital. Ninety one per cent of nights over the last three years, against a hospital average of about seventy eight.
That on its own is not wrong. Some wards run hotter.
What is wrong is that on a lot of those nights there are empty beds on it.
I found it as everything in this job is found, by two systems disagreeing.
The patient administration system holds admissions and discharges, and every patient has a bed. If you take the list of patients on ward 6 at midnight and count it, you should get the occupancy figure.
For about a hundred nights a year it comes to twenty or twenty one, and the midnight count says twenty two.
The obvious answer is a discharge not entered, and that is the commonest data error in a hospital by a distance. A patient goes home at four and nobody types it until the morning, so the system thinks a bed is full when it is empty.
That produces the opposite error. It makes the system count HIGHER than the ward, not lower.
Here it is the ward reporting higher than the system, which means a nurse standing on that ward at midnight, looking at the board and at the beds, is confirming a patient the hospital has no record of.
I went and stood on it.
Not once. Eleven times over two years, at midnight, which my director knows about and approved, because a data manager who never goes to a ward is a data manager who believes screens.
Four of those eleven nights, the count said twenty two and there were two empty beds.
Made up. Not stripped, not being cleaned. Made up the way a bed is made up when it is ready for somebody, with the blanket folded at the foot.
Bays four and six, on the left as you go down, both times both nights.
I asked the nurse in charge, the first time, carefully, at midnight, at the end of a hard shift.
She looked at the screen and she looked down the ward and she said twenty two, and she said it in the tone of somebody confirming something obvious.
I asked about bay four. She said that bed is occupied.
I said there is nobody in it.
She said, and I wrote this down in the corridor afterwards because I wanted the words: I know, but it is occupied.
Then she went back to what she was doing, and there was nothing unkind or strange in it, and she has been a band six on that ward for eleven years and she is one of the best nurses in this building.
I have asked four others since, on four different nights, all of them separately.
Three of them said a version of the same thing. One of them, a younger one, said that they leave those two made up, and when I asked why she said because they are.
None of them was uncomfortable. It is not a secret and it is not a story they tell. It is a fact about their ward, like where the linen is.
What the data does.
Ward 6's two beds are, in effect, permanently unavailable. Over a year that is about seven hundred bed days.
Seven hundred bed days is a lot. It is the difference between a hospital that cancels operations in January and one that does not, and I have said so, in writing, twice.
The first time it went to a capacity meeting and the outcome was that ward 6's establishment and occupancy be reviewed. The review found the ward well run and appropriately staffed.
The second time, in 2023, a director of operations went and stood on the ward himself at midnight, which I did not expect and which I respect.
His email to me the following morning says he is satisfied the ward is being managed correctly and asks me to close the query.
He did not say what he saw. I did not ask.
What I do about it.
I stopped reporting it as a data quality issue in 2023 and started treating it as a capacity fact.
Ward 6 is modelled as a twenty bed ward in every plan this trust makes. The bed base on paper says twenty two, and the model says twenty, and the model is right, because it predicts what actually happens.
Nobody has ever queried the difference. Every hospital's models have adjustments in them for wards that do not behave, and mine is a line in a spreadsheet called W6 ADJ that has been there for two years.
The one thing I have changed for myself is that I no longer do the midnight visits.
I have all the data I need and going up there was never going to produce more of it, and the last time, in the November, the nurse in charge asked me, quite kindly, whether I wanted to know which bay had been occupied longest.
I said no.