How can we measure the true scale of UK Covid hospital admissions?
The number of patients in hospital with Covid is rising, but how can we tell if they have been admitted because of the virus?
The number of patients in hospital with Covid is rising, but not all are admitted because of Covid.
Chris Hopson, chief executive of NHS Providers, which represents hospitals in England, has cautioned against overinterpreting hospital admissions data. “What our guys are saying is that incidental cases are about 25% to 30% of cases arriving … They are seeing an increase in the number of hospital admissions, but it’s not going up in an exponential way,” he said.
But just how many people are actually being hospitalised because of Covid, and how should the numbers be counted?
What are the current figures for hospital admissions?
The daily reported figure for Covid hospital admissions covers the total reported admissions to hospital and diagnoses while in hospital.
According to data from NHS England, the figure is rising, at 1,374 on 26 December compared with 926 on 19 December – a 48% increase over seven days.
Some reports suggest the majority of new hospital Covid patients are not “true cases”. Is that right?
What is clear is that while numbers are increasing, admissions are not rising as fast as cases, even taking into account the time lags between infection and becoming severely ill.
However hospital admissions data is not separated into those admitted with Covid or because of it.
This means that getting a true handle on the proportion of hospital admissions because of Covid is tricky, but is it possible to gain some insights from the figures already available.
The most recent available NHS England data reveals that on 26 December there were 1,338 new hospital cases. Of these 1,082 were either admitted in the previous 24 hours for the first time with Covid or were diagnosed in hospital in the previous 24 hours where the test was within seven days of admission.
In other words, 256 of the reported Covid hospital admissions – about 20% – were actually due to hospital-acquired infections, meaning the patients were originally admitted for different reasons.
This approach does not take into account those already infected with Covid but admitted for other reasons.
Dr David Strain, senior clinical lecturer at the University of Exeter medical school, said in his experience the vast majority of Covid patients are admitted because of Covid rather than the infection being an incidental finding.
And Hopson also suggested hospital bosses were reporting a figure of between 25 and 30%, so not the majority.
Should we be reassured that so many people are being admitted with Covid, rather than because of it?
Not really. Even those patients admitted ‘with’ rather than ‘because of’ Covid may only have required hospital care because of the disease itself.
“We [have] seen many other people who have been otherwise stable [with] chronic diseases such as heart failure, ulcerative colitis etc that caught Covid and had a rapid deterioration. Although they are regarded as “incidental Covid”, this sudden deterioration in otherwise stable disease can be attributed to the virus,” said Strain, adding people who have Covid also have poorer outcomes after surgery.
An alternative is to look at the total number of patients admitted, not just those with Covid, with Hopson noting some trusts have not seen an overall rise in admissions. However Strain said his experience is that overall unplanned admissions have risen. What’s more, he said, the approach has drawbacks. “The problem with looking at all admissions is, if the hospital is full, we don’t admit the surgical patients or people who we should be bringing in for elective procedures,” he said.
Is there another way to look at the situation?
As Hopson has noted, it is better to use numbers of Covid patients in hospital rather than new Covid admissions, as the data is more up to date. It also gives a more accurate picture of the pressures on the health service.
NHS England publishes a breakdown every week of the total number of Covid patients in hospital, together with the number who are being treated primarily for Covid.
The latest data gives figures up to 21 December – new figures are expected in the coming days. The data shows that of the 6,245 Covid patients in acute trusts in England on 21 December, 4,432 – or over 70% – were being treated primarily for Covid.
The remaining 30% of Covid patients in hospital were primarily being treated for another issue; however that does not necessarily mean they were not receiving some care for Covid as well.
Dr Julian Tang, a professor of respiratory sciences at the University of Leicester, said that the proportion of people in hospital with Covid, rather than because of it, will vary a lot between different hospitals depending on factors including the local incidence of Covid.
NHS data reveals the situation varies by region, with about 22% of Covid patients in hospital on 21 December primarily being treated for another issue in the south-west, compared with 39% in the Midlands.
Should patients in hospital or admitted with Covid, not because of it, be discounted from the Covid data?
Some have suggested including figures for patients with Covid is misleading as they may, for example, simply be receiving treatment for a broken leg.
But Covid can lead to more disease in those admitted for another reason, longer hospital stays, more frequent readmissions or outpatient visits and even – in some cases – death. Patients with Covid also need to stay in Covid wards, putting pressure on bed numbers, and meaning elective surgery is postponed or cancelled.
Covid patients also pose a risk of passing the infection to others, including other patients and staff. At present the NHS is hard hit by staff sickness and isolation due to Covid.
“When it comes to burden on the healthcare system, we should absolutely not discount them,” said Strain.
Source: https://www.theguardian.com/world/2021/dec/29/how-can-we-measure-the-true-scale-of-uk-covid-hospital-admissions
i've always had a lot of time for Neil Oliver.
