Michael Simmons

Will Omicron overwhelm the NHS? The crucial missing data

Will Omicron overwhelm the NHS? The crucial missing data
Text settings

If you catch Omicron your risk of ending up in hospital is between 50 to 70 per cent less likely than if you’d had Delta. That’s according to a new analysis released this evening by the UK Health Security Agency. It’s another blow to the case for lockdown.

That case for lockdown goes like this: Omicron is growing exponentially and its casualties will overwhelm the NHS unless action is taken to slow the growth. The cautious course of action is to wait until we know more about crucial unanswered questions, such as:

  • What is the limit to Omicron’s growth?
  • How much milder than Delta is it?
  • Is Imperial’s figure of 45 per cent fewer hospitalisations robust? And independently verified? Is it liable to be revised up until we get the 80 per cent seen in South Africa?
  • Might Omicron halve the average hospital stay, as it did in South Africa?
  • If so, how would that affect the Sage models for NHS capacity?
  • Are NHS infections distorting the ‘Covid hospitalisation’ headline figure?

Every day brings some fresh data, and adds more pieces to the jigsaw. Here’s what we know now:

1. Omicron’s growth is finding limits, in Gauteng and Lambeth Cases are still rising fast, but are we really looking at exponential growth that can only be checked by lockdown? South Africa did not lock down but cases are falling fast in the epicentre of Gauteng (home to Joburg, Pretoria and a quarter of South Africans). At The Spectator’s data hub, we have taken a cautious approach and regarded recent data (in grey) as being liable to upwards revision. But there is now no doubt about the direction of travel:

      London is Britain’s Gauteng and Lambeth is the city’s epicentre. But curiously, Omicron’s growth seems to be decelerating: it’s struggling to displace the last ten or so per cent of Delta infections.

      This is the first UK indicator that Omicron’s growth will not be extreme: perhaps thanks to people changing their behaviour in response to the case numbers (a phenomenon seen in developed countries world over). The Lambeth figures are tentative, and it’s only one council. But worth keeping an eye on.

      2. Average hospital stay: a potential game-changer The latest Sage minutes admit its working assumption — still — is that Omicron patients will stay just as long in hospital. But this is a massive variable. South Africa found the average stay was 2.8 days, two thirds less than the 8.5 days of the prior 18 months. The Sage document does not mention this massive downside risk in its assessment of hospital stay, but does spell out what this could mean for fears of overwhelming the NHS. ‘A reduced length of stay would allow more capacity within hospitals to manage this, and, to a first approximation, this would scale linearly with the change, i.e. halving the length of stay would permit double the admissions,’ they say. And, presumably, transform calculations about NHS capacity.

        3. Omicron’s hospitalisation rate: any advance on 45 per cent milder? After case numbers and average stay, the third most important metric is how less likely Omicron is to hospitalise. South Africa has said 80 per cent, Scotland has said 66 per cent and Denmark has said 60 per cent — so the 45 per cent calculated by Imperial College is now, by these standards, a low estimation. Imperial has made this calculation via access to daily patient data that is not openly available: it’s so important that it needs to be updated every day as more figures emerge. (So far, all UK data will be on a small dataset). If this all-important ratio does end up closer to South Africa’s 80 per cent that will, obviously, have huge implications for NHS caseload. The data needed to make this ratio ought to be made openly available, as much could hang on it.

        4. Understanding Covid hospital growth in London. The headline figure is shooting up — from 100 to 300 in just 18 days. But as the below graph shows, about a quarter of Covid cases were acquired in hospital. This growing gap hints that the number of patients being treated ‘with’ rather than ‘for’ Covid could be increasing.

          Today, that’s been confirmed. Each week NHS England publish the ‘primary diagnosis supplement’ which shows the primary reason that Covid patients end up in hospital. Since June this has been roughly flat, with around a quarter of Covid patients being treated ‘with’ rather than ‘for’ Covid. But since Omicron was discovered the ratio has shot up. The latest figure was 29 per cent on Tuesday.

          The ratio is highest in the Midlands: some 39 per cent of Covid patients in hospital are being treated for something else. In the South West it’s 44 per cent. Compared to the summer wave — when the NHS was far from capacity — there are considerably more of these ‘incidental’ admissions.

          This mirrors what was seen in South Africa, which had a higher number of admissions that were incidental. Perhaps this is not surprising given how transmissible Omicron is. It infects people in the community more easily so it spreads in hospitals too. What we don’t know is whether the ‘incidental’ patients end up ill from Covid. So it’s harder to use these crude ‘Covid hospitalisation’ figures as a predictor of whether hospitals will be overrun.

          5. The ICU caseload

          At the last count, Covid patients accounted for 6 per cent of general and acute beds, vs 25 per cent in the January peak. The data offers an interesting glimpse at something else: the number of patients on ventilators has been falling in recent days. As a result the percentage is falling even faster:

          6. The boosters: a success story This is, in part, thanks to boosters and immunity. The ONS confirmed today 95 per cent of over 16s have antibodies. You might argue that this has less meaning given how Omicron has shown itself able to infect the double-jabbed, but those with boosters have far greater immunity. The stand-out figure — and the great UK success story — is the boosters. Of the age group most likely to die (the over-60s) an astonishing 92 per cent are now boosted. For those in their 50s (a borderline risk group) it’s 80 per cent. No other country comes close to these figures.

          All the data is pointing to Omicron resulting in less severe outcomes. But still, as the Sage professors are keen to warn: ‘a smaller portion of a bigger number is still a bigger number.’ The focus then returns to cases and transmission. If a reduced hospitalisation rate buys us more headroom, just how high will those cases go?