Is there a simple plot comparing covid lethality to flu as a function of age?

Question

There is so much arguing over covid-19 vs. influenza. I would like to have some solid data in my hands. Can anyone point to a simple graph of the following? age vs. either CFR or IFR with one line for covid-19 and another for flu.

For covid there is some relevant, possibly-reliable info at Our World in Data, which has data from March (I couldn't find anything more recent, but this is at least something). This is case fatality rate.

For flu, here is a chart on Business Insider from June 23 which cites "Estimated flu cases and deaths from the CDC". I believe the CDC data used is this (for 2018-2019, for example). And the chart shows Deaths/Symptomatic Illnesses. I suppose that ratio corresponds more or less to the infection fatality rate (IFR), or at least the symptomatic infection fatality rate (sIFR?), which is probably the best we can do.

On the covid side of the chart from Business Insider the caption says "Confirmed COVID-19 cases and deaths from CDC". This is kind of vague but it looks like they're comparing apples and oranges -- flu IFR vs. covid CFR -- which is useless.

So my question is, can someone point either to a covid IFR by age or a flu CFR by age so we can make the comparison? IFR vs IFR would be best but in that case I would also like to see at least an attempt to estimate the error bars for IFR for both diseases.

I think such a plot could have a high impact on public understanding of this situation.

Bonus: Is there another plot that compares how transmissible covid is compared to flu? (I'm not sure what the appropriate metric is for transmissibility. R number I guess.)

Answer 1

Marc Bevand presents this graph of IFR versus age for both COVID-19 and flu:

The sources of data are documented on this github page. There is a summary of the sources in the README file.

Unfortunately, the IFR data in the graph does not come with error bars. However, most of the underlying sources do include error bars. For instance, Table 1 of "Estimates of the severity of coronavirus disease 2019: a model-based analysis", by Verity, et. al. states:

| age group | IFR      | confidence interval |
|-----------+----------+---------------------|
| 0–9       | 0.00161% | (0.000185–0.0249)   |
| 10–19     | 0.00695% | (0.00149–0.0502)    |
| 20–29     | 0.0309%  | (0.0138–0.0923)     |
| 30–39     | 0.0844%  | (0.0408–0.185)      |
| 40–49     | 0.161%   | (0.0764–0.323)      |
| 50–59     | 0.595%   | (0.344–1.28)        |
| 60–69     | 1.93%    | (1.11–3.89)         |
| 70–79     | 4.28%    | (2.45–8.44)         |
| ≥80       | 7.80%    | (3.80–13.3)         |

In contrast, the "US CDC" COVID-19 IFR data are presented as "current best estimates" which do not state confidence intervals.

I found the graph by google searching for "ifr flu by age", and clicking the "Images" link. Sifting through the graphs led to tweets which reference the github page.