Hacker News Reader

Comment by ppod | Hacker News Reader

People who are making claims about hospital capacity should compare them to the same times in previous years. The data is all available online:

https://www.england.nhs.uk/statistics/statistical-work-areas...

View on HN

This is a consistent taking point among the more egregious parts of the “lockdown skeptic” crew and it is, of course, absolute bullshit - it’s based on shallow comparisons of numbers that are not comparable.

If you do decide to promote this talking point anyway, flag it with the appropriate disclaimer from the NHS:

…caution should be exercised in comparing overall occupancy rates between this year and previous years. In general hospitals will experience capacity pressures at lower overall occupancy rates than would previously have been the case.

https://www.england.nhs.uk/statistics/statistical-work-areas...

The NHS depends on the government and is therefore not a neutral source.

Well but the NHS is the source of the data to begin with. And they communicate limitations along with the data. You can't just ignore those and use their data. You'd have to find another source that offers actually comparable numbers.

>In general hospitals will experience capacity pressures at lower overall occupancy rates than would previously have been the case.

Why?

Hospital capacity has had to be organised in new ways as a result of the pandemic to treat COVID-19 and non-COVID-19 patients separately and safely in meeting the enhanced Infection Prevention Control measures. This results in beds and staff being deployed differently from in previous years in both emergency and elective settings within the hospital.

Literally from the link.

Do we know how effective those measures have actually been in practice? There's been some reporting that large numbers of people have caught it in hospital (and indeed I have a personal interest here as a close relative of mine caught it in hospital just before Christmas, though a very minor case and now fully recovered).

It makes sense to try to stop spread in hospitals but unless I've missed something there seems to have been very limited work done on if any of the protocols as actually implemented (ie not just in a theoretical world where they were designed) - and especially the costlier ones in terms of bed space - have worked, or at least worked enough to offset their cost in health terms.

Does anyone know if there's any data on this?

>it is, of course, absolute bullshit

I'm sorry that I briefly lowered the tone by linking to data, thanks for your help.

So, January 2019 in London Commissioning Region hospitals:

1033 open adult critical care beds; 882 in use. 85% utilisation.

Best-case forecast Jan 2021:

2078 critical care beds required, 1,661 available (increase of ~600). Forecasted 125% utilistation.

Beat me to it. And according to this report, London currently has 46 free ICU beds.

I took the term open to mean unoccupied which put them at 46% utilization.

The columns in the spreadsheet are "Sum of Number of adult critical care beds open" and "Sum of Number of adult critical care beds occupied".

I don’t think beds really work in a way that allows 125% utilization.

They do when they start converting other wings into ICU space

Yes, that’s the point.

The data you provided isn't easily compared to the data in the article. I did a pivot off the Jan 2019 data and got this...

    LONDON COMMISSIONING REGION
    Adult Critical Care Beds:
        Open:     1033
        Occupied:  882
        --------------
        Total:    1915

According to the Article London has...

    BEST CASE
    Adult Critical Care Beds:
        Unmitigated: 1392
        Mitigated:   1661

So they lost 300-600 beds capacity since 2019?

Austria and Germany also report lower total ICU capacities than 1 year ago before the start of the pandemic. Weird...

We all got to save money. If you have 60% of your beds open you clearly don’t need all of them, so maybe that’s why they were first to go?

AFAIK you have misunderstood the data; the “open” figure is not the number of unoccupied beds, but the total number of beds available for patients (1033) of which 882 were occupied by patients already (meaning 151 unoccupied).

I disagree. The word open is superfluous if that's the case. What's the meaning of the word open if not unoccupied or available?

Of course that begs the question, why did they use open instead of unoccupied?

If open is meant to mean bed that can be utilized, then we don't have a true number of beds or a reason why a bed would be closed.

Either way it's confusing but generally the term open as the opposite of occupied for filled. For example, open positions in a company aren't employees, they're vacancies.

You can disagree all you want you really are categorising these numbers incorrectly; the data as supplied lists the categories “open”, “occupied”, and “% of open beds occupied” - which seems pretty clear.

There is inherently no “true number” of beds because this number varies in response to things like hospital space configuration, staffing levels, and demand.

AFAIK the term “open” is analogous to that in “open for business” - there may be other physical “beds” available, but without staff or without some equipment or some other constraint that makes them not open.

I do appreciate that it’s confusing, because I had to doh me check myself. But it’s not without reason that it’s described that way.

> You can disagree all you want you really are categorising these numbers incorrectly; the data as supplied lists the categories “open”, “occupied”, and “% of open beds occupied” - which seems pretty clear.

I went back and looked and you're right. I didn't look at the XLS file, I just pulled the raw data from the CSV file which doesn't offer the %.

It's clear from the XLS that the % is derived from occupied/open and not occupied/(open+occupied) as I was thinking.

Open means the bed and staff are present, but not necessarily available because it might have a patient in it.

This is compared to beds that are not available because eg the hospital is in critcon4.

The link you provided is for last year, and that site doesn’t have any data after February 2020 anyway. I think this is the best source:

https://www.england.nhs.uk/statistics/statistical-work-areas...

Reply on news.ycombinator.com