December 21, 2020

expert reaction to latest COVID-19 Nowcast and Forecast, by the MRC Biostatistics Unit at the University of Cambridge

The MRC Biostatistics Unit at the University of Cambridge have released their latest nowcast and forecast of COVID-19.

Prof James Naismith, Director of the Rosalind Franklin Institute, and University of Oxford, said:

“The Nowcasting group at Cambridge have an established track record, report their methods and the underlying data. 

“Their estimate of the number of new daily cases makes sobering reading 91, 000 (95 % CI, 69, 000 to 119, 000).

“There is no reason not to accept these as fairly robust estimates, even at the lower end, it would confirm that the virus is rapidly spreading in the UK.

“There are several possible reasons for these, one is that social distancing has not been effectively implemented or has occurred too late. One explanation is that the new strain spreads more rapidly, it will take time to be sure which is true. Rushing to judgement helps no one, we must know the cause to tackle to problem.

“Whatever happens January and possibly February will be difficult months, with death tolls once again exceeding 500 per day on some days.

“This is a very serious moment, we can all do our part, washing hands, wearing a mask, socially distancing and following government guidance.

“Track and trace has not achieved its goal. The system is at best catching just less than half the actual cases. It is disheartening that such an obvious flaw remains under reported.

“It’s important for people who make statements on radio, social media, TV etc which say “show me the data” to be challenged on them in real time. The data are all here in the Nowcast, anyone can offer a different interpretation…This is entirely proper and provided the different interpretation is offered with the same level of transparency of methods as the Nowcasting team, then this is valuable contribution. There may be better ways to use these data, that dramatically change the interpretation, it’s highly unlikely but this is how science is conducted.

“However, pretending the data are not available is not science. It is in fact dishonest and / or incompetent, in either case the person making such statements is adding no value whatsoever and a period of silent study on their part will save more lives.

“I note that the ZOE KCL data have not tracked the apparent spike in cases and rather closely followed the official test numbers. It will be interesting to see the ONS data in January.

“We have always believed that wide spread effective vaccination alone will end the pandemic, the good news today is the progress on the roll out of the vaccines.”

Prof Sheila Bird, Formerly Programme Leader, MRC Biostatistics Unit, University of Cambridge, said:

“Today’s nowcasting and forecasting by my former colleagues at MRC Biostatistics Unit is timely: particularly for what it reveals about East of England, London and the South East.

“First, a caution for too-quick readers: the vertical scale changes dramatically between side-by-side plots. For example, East Midlands (where I live) is side-by-side with London but 5,000 infections on the East Midlands’ scale lines up against 40,000 infections for London. The estimated median number of infections on 18 December was 7,800 for East Midlands (uncertainty; 4,500 to 12,900), 16,000 for London, 14,500 for East of England and 17,200 for the South East. See plots for the uncertainty which qualifies each of the other three medians.

“Second, the regional plots – East of England, London and the South East excepted – show a notable dip in infection-incidence during the November lock-down with rebound thereafter in December. The three exceptional regions are now in Tier 4, but the reason is as much genomic as the absence of impact from lockdown. Nonetheless, the two may be partially, behaviourally confounded.

“Third is the gloomy short-term forecast – made ahead of, and so without accounting for, Tier 4 restrictions – that on New Year’s Day England might record 750 COVID-mention deaths (95% uncertainty: 620-900), almost as many as at the April peak.

“The Tier 4 regions have pre-restriction forecasts  for 1 Jan 2021 as follows: London 70 COVID-mention deaths  (40 to 115) versus 1st wave peak of around 200; East of England 150 COVID-mention deaths(70 to 200) versus 1st wave peak of around 100; South East 160 COVID-mention deaths (100 to 245) versus 1st wave peak of around 125.

“Should I be as worried in the East Midlands as those in the South West, neither in Tier 4, that our median forecasts for COVID-mention deaths on 1 January 2021 also exceed the region’s 1st wave peak?

“On Saturday, Scotland put the entire Scottish mainland into Tier 4, thereby invoking legal powers to close its border with England.”

https://www.mrc-bsu.cam.ac.uk/nowcasting-and-forecasting-21st-december-2020/

All our previous output on this subject can be seen at this weblink:

www.sciencemediacentre.org/tag/covid-19

Declared interests

Prof Sheila Bird: SMB is a former programme leader at MRC Biostatistics Unit.

None others received.