April 12, 2020

expert comments on the updated IHME model and the potential for the UK to have the highest number of deaths from COVID-19 in Europe

The Institute of Health Metrics and evaluation (IHME) have release their latest model for how the UK COVID-19 out break will progress, along  comments on the potential for the UK to have the highest number of deaths from COVID-19 in Europe.

Prof Sylvia Richardson, University of Cambridge & President Elect of the Royal Statistical Society & co-chair of the Royal Statistical Society Task Force on COVID-19, said:

“Countries like South Korea and Germany, which are able to pursue testing of acute COVID-19 infections followed by isolation on a massive scale, the universally agreed first line policy for controlling an epidemic, have experienced better trajectories so far. This was the initial strategy followed by PHE in the UK but was not sustained as we did not have the testing capacity at that time.  Ramping up testing capacity as outlined by the government is key to being able to restore such a strategy.”

Prof Babak Javid, Principle Investigator, Tsinghua University School of Medicine, Beijing and Consultant in infectious diseases at Cambridge University Hospitals, said:

“The IHME model is extremely sensitive to the base underlying assumptions.  A few days ago, when the predicted UK deaths were about 67,000, one base assumption was that the model assumed only about 800 ICU beds available in the UK, which was a massive shortfall. Checking today (with the revised downward projection of 37,000 deaths), the base assumption is that the UK has 6800 ICU beds available. Comparing France/UK in the models, there is huge divergence on assumptions about general medical beds available too. I can’t comment how accurate these assumptions are: but it seems clear that the model is extremely sensitive to them. So the predictive power of the model is likely to be highly dependent on how accurate those assumptions about ICU beds are.  I would want to make clear that I am not an expert in modelling, but the fact that underlying assumptions in the model vary by such magnitudes over short spaces of time should be investigated.”

Prof Sir David Spiegelhalter, Chair, Winton Centre for Risk and Evidence Communication, University of Cambridge, said:

“Although reporting delays make it difficult to track the epidemic with confidence, we appear to be following a slightly worse trajectory than Italy and it is plausible that we might end up with the most in-hospital Covid deaths in Europe. 

“However the IHME projections still seem unduly pessimistic, even after their recent dramatic change.  For example, their model is saying that today (April 12th) we should be needing 62,000 hospital beds while there are only 18,000 available –  although our hospitals are undoubtedly under a lot of pressure, this does not seem to fit reality.  Their curve-fitting procedure currently seems too sensitive to small changes in data and assumptions to make it reliable for the UK.”

Prof Keith Neal, Emeritus Professor in the Epidemiology of Infectious Diseases, University of Nottingham, said:

“The IHME model has been running for nearly a week since first published.  Re-doing their prediction in under a week strongly suggests major flaws in their models. The number of deaths each day in their original calculations have been just above their lower limit so they have been seriously overpredicting UK deaths since I first saw their model.  This is not the first model to be shown to have got their projections seriously wrong.  Although this is a pandemic the epidemiology in each country is different in different countries and different within countries.  Compare London, Lombardy and New York to other parts of the respective countries. 

“The daily trend in the UK appears to be slowing but we need another week to be sure especially with possible reporting delays over the holiday period.

“The number of deaths by country within Europe where health services have not been totally overloaded except in Italy reflect how long their epidemic has been going on for, (not how long it has been recognised), and other factors such as social mixing and family structures.

“It is likely that the UK will have one of the largest number of total deaths solely because we the second largest population in Western Europe and EU countries, only Germany has a larger population. The important figure is the death rate per million and the total number of deaths.  On this count Belgium seems to be heading for a serious problem like Italy and Spain.” 

https://covid19.healthdata.org/united-kingdom

All our previous output on this subject can be seen at this weblink: www.sciencemediacentre.org/tag/covid-19

Declared interests

None received.