Covid model implies IFR is ~0.5%

At the beginning of March Neil Fergusson modelled the impending Covid-19 epidemic for the UK  which was expected to begin in early April and consequently reach a peak in May, all based on the infection rate assumed at that time. His influential “Report9” circulated and published on March 16 reflected these underlying assumptions. The lockdown measures that he proposed  to suppress infections and thereby allow the NHS to cope with the predicted surge in demand for ICU beds were also timed for April. There seems to have been a widespread belief in SAGE at the time that the UK was 3 to 4 weeks behind Italy, so consequently we had ample time to plan measures accordingly. We can now see that this simply wasn’t true and infections were spreading fast in early March. This was mainly because of open borders to Europe and a lack of testing capability. Figure 1 compares the recorded deaths in Italy, France and England as of June 16.

Figure 1. Daily recorded deaths in Italy, France and in NHS hospitals in England. Note the overall similarity between  England to France. The criteria for recorded deaths vary between countries. Here  I am just using hospital deaths registered by  day of death for England (population: 56 million)

In Figure 1.  I am plotting just the hospital deaths in England as recorded by the NHS because this probably best represents the dynamics of community wide infections. (The tragic deaths that have also occurred in care homes are another story!). Note how the epidemic begins simultaneously in England and France, both of which lag about 2 weeks behind Italy.  However by the end of April  all 3 countries then tail off together at almost the same rate ending together.  The Italian epidemic just seems to have lasted longer.

Ferguson originally released his code for “CovidSim” on GitHUB in May, together with the parameters files that describe the UK, updated to describe the emerging epidemic data. The unmitigated run he used on GitHUB show that he increased Ro to 3 because it was discovered that early UK infections were doubling every 3-4 days. His suppression  simulation was really based on the consequent  March 16 Household Quarantine (HQ) and limited Social Distancing (SD) measures as announced by Boris Johnson, then followed 7 days later on March 23rd by stringent Place Closures(PC). On March 23rd Boris Johnson announced that all pubs, restaurants,  shops, schools, Universities would close forthwith and that everyone should stay at home as far as possible.  March 23rd is therefore considered to be the real lockdown date for the UK, forgetting that a week earlier social distancing had already been already been introduced. As a result of all this the parameter file that Fergusson uses to simulate what really would happen during the UK epidemic is called PC7_CI_HQ_SD, where “PC7” symbolises the 7 day delay between March 19 and March 23 in implementing full lockdown.

Firstly we look at the unmitigated predictions. Figure 2 shows the  comparison between the original Report9 Neil Fergusson model predictions as made in early March with those from his updated May version.

Figure 2. A direct comparison between Fergusson’s Report 9 simulation with R0=2.4 and the later UK simulation in May following the start of the outbreak.

The most striking feature here is the nearly one month anticipation of the real infection peak compared to the earlier Report9 simulations. Note also in passing that the original Report9  runs were made just for Great Britain and for some reason excluded Northern Ireland. This however can only explain a small proportion of the increase in  infections. The major reason for this increase was essentially due to using the observed increase in R0 to 3.0. SAGE were estimating that infections were doubling every 3 days.

Secondly we can compare the new May CovidSim “lockdown” simulations of deaths against the actual statistical data as recorded up to June 16th (Figure 3). I am using COVID-19 deaths as a measure of infection rates mainly  because these are the only consistent numbers across time.

Figure 3. A comparison between the Ferguson model simulation of the UK lockdown with the actual death data vas recorded across the UK including Care Homes. The total unmitigated deaths reach 620,000. The recorded cumulative UK deaths were 41736 on 16 June.

The new model is clearly predicting more deaths than were actually recorded even if we now also include care home deaths.  These care home deaths were more a consequence of NHS lockdown measures rather than community infections as simulated by Fergusson. So what does all this mean ?

Ferguson had assumed an IFR (Infection Fatality Rate) of 0.9% and as a consequence his model predicted that 70,000 deaths would occur by June 16. In reality 41736 total deaths have occurred across all settings by then. It is likely that he calculated the infection rate correctly but simply used a too high value for IFR.

Therefore I would  conclude that IFR is  ~0.5%, and probably less than that once you exclude hot spots of infections such as those occurring  inside care homes.

This entry was posted in Public Health and tagged . Bookmark the permalink.

10 Responses to Covid model implies IFR is ~0.5%

  1. Andrew Carey says:

    Oh dear, IFR – that’s infection fatality rate right, the ratio of infections to the number of deaths. But what counts as an infection? For sure if 3000 virons get in to your body following a 2 hour liaisob in the same room as a carrier and these virons start replicating, then you are definitely infected and you will have a bad few days based on influenza data. If your viral dose is say 30 virons from a 1 minute conversation outdoors, then who would even know? Were you infected – you probably were – did your immune system get on the case before you even noticed – probably yes. But nobody would know to include you in the denominator.
    There is a way to get a better idea of IFR in due course, but it would involve serology data, and also what the conditions are when the virus goes away of its own accord. But speculation about IFR now is just fanciful.

    • Clive Best says:

      I agree but government policies are unfortunately dictated by “experts” . The serology data shows only up to 17% of people with antibodies in London. Perhaps some people already have immunity but don’t develop immunity. The problem right mow is that all governments are all scared to release lockdown measures too fast in order not to get the blame if a second peak arrives. Hence the reliance on “experts”.

  2. Mike Greig says:

    Clive. Have you seen this paper from the Oxford Centre for Evidence Based Medicine? https://www.cebm.net/covid-19/covid-19-mortality-over-time-ons-update-2nd-june/
    It reports that the number of excess deaths in the U.K. was 29,000 versus the ONS’ 51,000 at the time. I think the CEBM’s logic is correct and their figure a better view of excess deaths. The main reason for the difference is that the ONS compares to an average of the last 5 years whereas the CEBM uses a harmonic progression forecast of deaths in 2020 which takes in to account: a) The population has been increasing and ageing and there has been a trend for the absolute number of deaths to increase over time. b) that there is a cyclic effect on deaths – if you have more deaths o e year you will have less next year and vice-versa (and 2019 was definitely low). I calculated a linear regression for a) over 10 years and got an increase in deaths for the first 21 weeks of the year of 3,247 p.a. and a base figure for 2020 which was 7,000 higher than the ONS’ 235,293.
    I am very surprised no one else has picked up on this.

    • Clive Best says:

      No I hadn’t. Thanks

      I think there are three separate COVID “epidemics”
      1. Community wide infection – i.e. the standard model of epidemics
      2. Localised infections within Hospitals and Care Homes.
      3. Knock on deaths caused by lockdown – Suicides, delayed cancer treatment, heart operations, strokes etc.

      It is also rue that all countries count deaths differently so a direct comparison is not straightforward. For example Spain reduced the total number of deaths by more than 100 in one day.

      • Mike Greig says:

        Clive in your third category there was an initial reduction of deaths in the initial phases of lockdown people got less flu like illnesses and drove less.

    • MarkR says:

      Those sound like corrections for factors that need to be addressed. Using year-to-date deaths in the model is a bit strange to me, and performance tests are missing though.

      I took the 11 available weeks from 11th March and compared versus the 4 weeks just before in ONS data. This tries to get before/after big COVID effects.

      From 2010–2019 the “most positive” change was -105/week (in 2010) and the most negative was -2,027 (2018). As you’d expect, weekly deaths tend to drop in winter-spring transition, generally it was more negative recently. I don’t know whether that’s coincidence or e.g. related to aging.

      By taking this year’s change and using the typical changes from the past decade to represent range I get 58,871 (49,115–71,412). I nudged the lower end down a bit.

      It’s hard to understand how ~30,000 coincidental and additional deaths beyond the Oxford blog post happened over 11 weeks but I’m not an epidemiologist.

      If the fraction of those excess deaths attributable to COVID-19 is “f” then with Clive’s 0.5% IFR you get about 11.5f million COVID-19 infections in the UK as of sometime before 5th June. Or another way: sometime before 5th June then Clive’s IFR implies almost f*17% infection rate in the UK population.

      I suspect “f” is pretty high, which raises plenty of questions.

      • Mike Greig says:

        Mark, Clive The CEBM have today updated their excess deaths calculation today to include linear as well as harmonic trends and to look at just the COVID period as well as the year to date. When looking at just the COVID period the results of all 3 approaches are closer. Although the article is written in academic speak, they seem to favour the year to date approach which now has 31,417 excess deaths. They point out that the deaths and excess deaths are concentrated in the over 85s – who account for c.40% of all deaths and also have the greatest variation from year to year – i.e. years which have lower counts are then followed by more deaths in the next year – as in 2014/15 and 2019/20. There were also a negative excess deaths for weeks 1 – 12 of 2020.
        https://www.cebm.net/covid-19/thoughts-on-estimating-excess-mortality-from-covid-19/

        • MarkR says:

          Looks to me like their implied covid numbers are similar to mine (~55k) but other cause deaths are lower because the flu season wasn’t as deadly as it sometimes is. I don’t read their article as “favouring” the year to date approach for estimating covid.

          I still think my statement is roughly right: “Clive’s IFR implies almost f*17% infection rate in the UK population”, sometime before 5th June. Assuming ~3 weeks from infection to case resolution then it would be ~mid-March.

          Maybe many of these victims would have died next year or the year after anyway. I’ll leave all those calcs to the QALY folks.

  3. MarkR says:

    Have you tried using the ONS excess deaths plus serology test infection rates to get an IFR range?

    Good news is it should go down in future, assuming the Oxford steroid results hold up. Improved treatment as we learn more always seemed like a nice extra benefit of how lockdowns delay infection, even if they ultimately fail.

Leave a Reply