Martin Juhl Profile picture
PhD, Drug Development, Diploma in Pharmaceutical Medicine (ECPM). Studied @ Uni of Copenhagen, UIUC, DTU, Harvard, @FulbrightPrgrm. Opinions my own

Dec 29, 2020, 28 tweets

Thorough comparison of SARS-CoV-2 and the seasonal influenza

Comparisons
➡️ CFR & IFR
➡️ Case-to-case
➡️ R0 & Herd theshold
➡️ Direct deaths to deaths
➡️ Relevant articles

1/x

Will use US and DK as examples

US had 12,000 - 61,000 deaths pr year (330 mill ppl)
And 9.3 - 45 mill estimated symptomatic infected

Both influenza deaths & symptomatic individuals are only estimates, based on statistical analysis, excess deaths, & sentinel monitoring
2/x

This paper states ~3-11% symptomatic flu cases pr year
onlinelibrary.wiley.com/doi/abs/10.111…

Below refs show that there are an additional 50-75% asymptomatic cases, thus indicating 6-44% flu cases (incl asymptomatic) in total pr year
3/x
wwwnc.cdc.gov/eid/article/22…

ecdc.europa.eu/en/seasonal-in…

Now we have enough data to calculate both CFR and IFR for influenza.

➡️Influenza US
The CFR is ~0.14%
Taking into account 50-75% asymptomatic cases
IFR (US)=0.034-0.068%

➡️Covid-19 US
CFR = 1.7%
IFR = 0.65%

➡️ Ratio Covid-19/flu
CFR ratio is 12
IFR ratio is 9.8-19.7
4/x

Age-adjusted comparison looks like this

IFR of COVID-19 vs seasonal influenza: covid is significantly more fatal than the flu at all ages above 30:

• 2.9× more fatal at age 30
• 5.0× at 40
• 8.6× at 50
• 13.6× at 60
• 14.4× at 70

5/x
Source: github.com/mbevand/covid1…… .

4,674 deaths in the US by Covid-19 in 0-39 year olds in Nov 2020
Using age-adjusted CDC data from 2014, one can calculate how many flu deaths happened in 0-39 year olds, that was 515 flu deaths

That is actually 9x worse even in younger adults?

6/x

According to CDC (2017) only 6,515 coded flu deaths, while 49,157 pneumonia deaths, which this graph highlights

Only 10-12% of flu&pneumonia cases are coded as flu deaths, that same is seen in the UK

7/x

As influenza deaths are estimates based on excess deaths, one could argue, that a fair comparison would be to use excess deaths for Covid-19 as well.

Excess deaths are a factor of 1.6x the Covid-19 deaths

That would lead to a Covid-19 being 15-30x worse than the flu
8x

Denmark had an estimated 0 - 2,822 flu deaths per year (5.8 mill ppl), based on excess deaths and sentinel monitoring.

No data available for total estimated influenza cases in Denmark

One can zoom in on the worst season 2017/2018 and try and compare
9/x

The DK flu season 2017/2018 have the following data available

2018:
192 deaths with influenza as primary cause (Danmarks Statistik)
603 deaths with a positive flu test within 30 days
2,822 estimated deaths based on excess deaths & sentinel
Excess deaths in 2018 = 1,666
10/x

Furthermore, 7,667 hospitalizations and 549 in ICU.

➡️ 7% of hospitalizations which ended up in ICU

With Covid-19 it is 12-25% of hospitalizations which requires ICU care
ecdc.europa.eu/en/publication

11/x

In DK we have had 3-5% infected with Covid-19 in total & 1,204 deaths, 8,185 hospitalizations & 805 ICU patients

If one expects 60% infected in total to compare with a seasonal flu:
15,000-24,000 deaths
98,000-164,000 hospitalizations
9,700-16,000 ICU

12/x

Denmark comparison Covid-19 vs 2018 flu

C19/flu
➡️deaths 5-9x
➡️hospitalizations 13-21x
➡️ICU 18-29x

13/x

A strange caveat to the DK 2018 flu data, is looking at age-adjusted IFR for that flu-season

0-4=0.00787%
5-14=0.00077%
15-64=0.00627%
65+ =0.22103%

If one takes the age-adjusted IFR & multiply with DK age pyramide, one gets exactly 2,822 deaths at 100% infected?

14/x

Case-to-case comparison

The JAMA paper comes in at a factor 20.5 (CI 16.3 - 27.7)

"Assessment of Deaths From COVID-19 and From Seasonal Influenza"

jamanetwork.com/journals/jamai…

15/x

Looking at case-to-case comparisons instead from France:
The mortality rate was significantly higher in COVID-19 group than in influenza group (20% vs. 5%; p = 0.002).
Factor of 4

journalofinfection.com/article/S0163-…

16/x

Danish study find factor of ~6, but as the testing for flu and Covid-19 are markedly different, experts judge that it is problably 10-20x, if one corrects for testing
“Det er nok mere reelt, at covid-19 er 10 eller 20 gange så dødelig som influenza” siger Christian Wejse.

17/x

Links to danish study and newspaper reference

"Compared to hospitalized patients with influenza A/B, new-onset ischemic stroke, diabetes and nephropathy occurred more frequently in patients with COVID-19 (all p < 0.05)."

frontiersin.org/articles/10.33…

berlingske.dk/samfund/en-del…

18/x

The median R value for seasonal influenza was 1.28

bmcinfectdis.biomedcentral.com/articles/10.11…

R0 value of SARS-CoV-2 is estimated at 2.5

thelancet.com/pdfs/journals/…

Herd immunity threshold (HIT) = 1 - 1/R0

Flu HIT = 22%
SARS-CoV-2 HIT = 60%

19/x

The markedly lower R0 value of influenza, is also observed indirectly as influenza has disappeared in all parts of the world
Very low to no flu observed in Europe, US, Australia, Argentina, and so on.

This means restrictions work, but SARS-CoV-2 is just very infectious

20/x

Direct death to death comparison

US
Common flu: 12-61,000
Covid-19: 340,000 (excess deaths >400,000)

C19/Flu
5.6-28x worse (without excess death)

If one takes into account that only 15-20% of US has been infected in total & 60% is needed for Herd Immunity

➡️ 17-22x
21/x

Good article about Covid-19 and Flu comparison
denverpost.com/2020/08/06/jus…

Suggest factor of 10

"... help us better understand how Covid-19 stacks up, which is to say that it’s definitely the worst such pandemic to come along in 60 years and probably the worst in a century."
22/x

De fleste analyser peger på, at COVID-19 er værre, hvor meget vides endnu ikke. Hvis vi ser bort fra de tre outliers i både den lave og høje ende, som alle er behæftet med store usikkerheder, får man, at COVID-19 er 4-40 gange værre.
23/x

videnskab.dk/forskerzonen/k…

Telling visualization from @jburnmurdoch on the difference in hospitalizations & ICU between between a seasonal flu season and COVID-19 in England
24/x

Comparison of flu to COVID-19 in younger adults. Might be skewed a bit due to different R0 value, but quite significant difference
25/x

As there are very limited flu serology studies and thus poor estimates of IFR, one can look at population fatality rates.

Currently Covid-19 vs avg flu season➡️2-10 x worse

Adj for % infected with Covid-19➡️21-30 x worse

Only compare to worst flu season➡️7-23 x worse
26/x

Comparing the Covid-19 deaths to influenza deaths in Sweden
27/x

Population Fatality Rate (PFR) comparison of COVID-19 vs seasonal influenza (Average & worst flu season)

PFR is how many deaths out of the entire country population

Covid-19 data updated (15/2-21)
28/x

Share this Scrolly Tale with your friends.

A Scrolly Tale is a new way to read Twitter threads with a more visually immersive experience.
Discover more beautiful Scrolly Tales like this.

Keep scrolling