During the Downing Street Press Conference on Saturday, a graph was shown that modelled a reasonable worst case planning scenario, titled ‘Winter scenarios from early working analysis: England daily deaths if no changes in policy or behaviour’
Prof James Naismith FRS FRSE FMedSci, Director of the Rosalind Franklin Institute, and University of Oxford, said:
“As more data become available, our understanding improves. It also has to be understood that no one model for daily number of new infections, be it Nowcasting, ONS, REACT or Zoe is a precise measurement. All of these models have strengths and weaknesses, all report their models with confidence intervals (that is what range of outcomes fits to the data).
“It is correct that the Nowcasting model changed on the 29th October, the Nowcasting model changes every week, as do other models. Predicting the future always has uncertainty.
“What even the most optimistic models agree on is that we will see around 500 deaths per day in two to three weeks (best case). We know these deaths will happen because of the number of people infected last week.
“What matters now is how many people are going to infected each day this week and next week. This will determine how many people die in 4 weeks time.
“We can choose do nothing and wait to be see if the existing measures start working. We don’t know for sure that they won’t work; we do know for sure that they have not yet started to work enough to make a difference.
“If the virus continues at the rate we saw last week, then taking the two or three weeks to prove beyond any doubt that the current measures have failed, then we will be unable to avoid over 1000 deaths a day (in a best case scenario) before Christmas. This is to put a hope over evidence and to risk at least 500 unnecessary deaths per day.
“Predicting the future is always hard, but the prediction I made above are based on simple well understood maths that anyone can do.
“It is for politicians to make and to accept responsibility for these decisions in a democracy.”
Prof Neil Ferguson, Director, MRC Centre for Global Infectious Disease Analysis, Imperial College London, said:
“SPI-M undertakes a wide range of modelling for government. The “up to 4000 deaths a day” scenarios represent preliminary work to generate a new reasonable worst case planning scenario to assist NHS and other government planning. The reasonable worst case is intended to be pessimistic, so these scenarios assumed an R value of 1.3-1.5 and that the tier system would have minimal impact. SPI-M also generate 6-week projections on a weekly basis. As Sir Patrick Vallance indicated on Saturday, the latest analyses indicate that while Tier 2 and 3 measures have reduced the rate of growth of the epidemic, the R value is still above 1 and hospital admissions are doubling every 17 days. Thus even allowing for the effects of the current tier system, the most recent SPI-M projections suggest that without further action, the second wave is set to exceed the first wave in hospital demand and deaths.”
Graph on page 3 here, entitled ‘Winter scenarios from early working analysis: England daily deaths if no changes in policy or behaviour’: https://assets.publishing.service.gov.uk/government/uploads/system/uploads/attachment_data/file/931775/Slides_to_accompany_coronavirus_press_conference-_CSA-__31_October_2020.pdf
All our previous output on this subject can be seen at this weblink: