IHME Projections

The Institute revised its UK COVID-19 model.

The Institute for Health Metrics and Evaluations estimate deaths from COVID-19 by country. For the UK, there was major news coverage of our country’s IHME estimates.

This article looks at these IHME projections, and their revisions.

The model and the range

IHME made models to estimate peak demand of healthcare systems. This work focuses on US states. Also, IHME produced estimates for deaths, and for European countries.

The original IHME projection was that the UK would have around 66,000 COVID-19 deaths. One online article in The Guardian had over 18,000 shared links.

How does this fatality model work? Their model fits a curve to the series of reported deaths in each country. This is an empirical model: informed by the shape of COVID-19 deaths in each country.

It is like seeing the initial arc of a thrown ball, then predicting how high and far that ball will travel.

There are extra weights, which seek to model policies like stay-at-home orders. This model does not seek to estimate the spread of COVID-19 within each population.

There is high uncertainty in these estimates. In the original publication, the central prediction for UK deaths by early August was 66,314. The uncertainty interval around that prediction was from about 55,000 to 80,000.

Sensitive predictions

With four days more information — including of peaks elsewhere — the UK prediction changed. The original model overestimated COVID-19 deaths within hospitals.

This is a major revision. (Image: IHME)

Their central prediction fell from around 66,300 COVID-19 deaths to about 37,500. The new interval for death counts suggests a plausible range from around 26,100 to 62,500.

This uncertainty interval does not include the older figure of 66,300.

There is still a wide range of uncertainty. (Image: IHME)

This is a major revision. The empirical model appears to be very sensitive to new information and its assumptions. This updated model also overestimated reported COVID-19 deaths within hospitals, for April 10-12th.

As more information comes in, the Institute could revise their model again.

Journalists should take care when reporting these models. The IHME model for the UK appears to be sensitive to new data. Different models have different purposes, involving many layers of uncertainty.

This blog looks at the use of statistics in Britain and beyond. It is written by RSS Statistical Ambassador and Chartered Statistician @anthonybmasters.

Get the Medium app

A button that says 'Download on the App Store', and if clicked it will lead you to the iOS App store
A button that says 'Get it on, Google Play', and if clicked it will lead you to the Google Play store