More than half of Covid hospitalisations are patients who only tested positive after admission, leaked data reveal.
The figures suggest vast numbers are being classed as hospitalised by Covid when they were admitted with other ailments, with the virus picked up by routine testing.
Experts said it meant the national statistics, published daily on the government website and frequently referred to by ministers, may far overstate the levels of pressures on the NHS.
The leaked data – covering all NHS trusts in England – show that, as of last Thursday, just 44 per cent of patients classed as being hospitalised with Covid had tested positive by the time they were admitted.
The majority of cases were not detected until patients underwent standard Covid tests, carried out on everyone admitted to hospital for any reason.
Overall, 56 per cent of Covid hospitalisations fell into this category, the data, seen by The Telegraph, show.
Crucially, this group does not distinguish between those admitted because of severe illness, later found to be caused by the virus, and those in hospital for different reasons who might otherwise never have known that they had picked it up.
Last month, health officials instructed NHS trusts to provide "a breakdown of the current stock of Covid patients", splitting it into those who were in hospital primarily because of the virus and those there for other reasons. So far, NHS England has failed to publish this data.
Watch: Coronavirus in numbers - UK deaths rise by 86
However, the patterns shown in the leaked figures – with the vast majority of hospital Covid cases being diagnosed after admission, in some cases weeks later – suggest it includes large numbers likely to have been admitted for other reasons.
The breakdown of daily Covid hospital diagnoses shows that of more than 780 hospitalisations dated last Thursday, 44 per cent involved people who tested positive in the 14 days before hospital entry.
A further 43 per cent were made within two days of admission, with 13 per cent made in the days and weeks that followed, including those likely to have caught the virus in hospital.
Experts said the high number of cases being detected belatedly – at a time when PCR tests were widely available – suggested many such patients had been admitted for other reasons.
Prof Carl Heneghan, director of the Centre for Evidence-Based Medicine at the University of Oxford, said: "This data is incredibly important, and it should be published on an ongoing basis.
"When people hear about hospitalisations with Covid, they will assume that Covid is the likely cause, but this data shows something quite different – this is about Covid being detected after tests were looking for it."
Prof Heneghan urged the Government to publish clearer data, including whether or not the virus was the primary cause of hospital admission.
"This needs to be fixed as a matter of urgency," he said, adding that the published data could lead the public "towards false conclusions", exaggerating the true levels of pressures on hospitals.
Sir Graham Brady, the chairman of the 1922 committee of Tory MPs, said: "Nearly 18 months into the Covid crisis, it is absurd that data breaking down hospital admissions still isn't publicly available on a regular basis.
"Counting all patients who test positive as Covid hospitalisations is inevitably misleading and gives a false picture of the continuing health impact of the virus."
Watch: How the world could be better after COVID
Greg Clark, the chairman of Commons science and technology select committee, on Monday night said he would write to Sajid Javid, the Health Secretary, asking him to publish the breakdown on a regular basis following The Telegraph's disclosure.
"If hospitalisations from Covid are a key determinant of how concerned we should be, and how quickly restrictions should be lifted, it's important that the data is not presented in a way that could lead to the wrong conclusions being drawn," he said.
"While some of these people may be being admitted due to Covid, we currently do not know how many. And for those who are not, there is a big distinction between people who are admitted because of Covid and those are in for something else but have Covid in such a mild form that it was not the cause of their hospitalisation."
The leaked statistics come from NHS daily situation reports, collected by all hospital trusts in England.
One NHS data expert said the published statistics distorted the true picture, saying: "It creates an impression that all these people are going into hospital with Covid, and that simply is not the case. People are worried and scared and not really understanding the true picture – that is what I find despicable."
An NHS spokesman said: "Many patients are admitted to hospital because of their Covid symptoms and complications, which are then confirmed with a post-admission Covid test, and for others they may initially be presymptomatic or asymptomatic."
On Tuesday, NHS leaders said the health service is as stretched now as at the height of the pandemic in January and made a plea for extra funding.
In a letter to the Prime Minister, Chancellor and Health Secretary, NHS Providers, which represents hospital trusts, raised fears the situation could get worse before it gets better.
They said "very different pressures" - including a "massive backlog" of patients in need of care, high numbers of staff self-isolating or on annual leave, meant the strain on the service now is just as bad as at the start of the year.