NOTE: This post was inspired by a recent post from Jessica Rose:
NOTE: This post has been updated to introduce a 1-week delay for the death count associated with a given number of cases.
The best infection fatality rate (IFR) estimates are the low ones, because there are too many factors which can confound and compound the death seen after infections. If good medicine is withheld, like it was with COVID, then the IFR estimate will be artificially inflated.
But there are hardly any ways that you can “deflate” an IFR estimate.
Because of inherent virulence, harmful micro-organisms do not allow for IFR estimates that are lower than the actual one. To get an IFR estimate that is lower than actual, you would have to undercount deaths and/or over-count infections. The low (good) estimate for the IFR for M-pox comes from 2 Jul 2022, with 3 deaths:
By that time, there had been 7,307 infections:
Using computer simulation to generate plausible IFR values from this many deaths which had occurred from that many infections, you find out that M-pox isn’t worse than flu — using the rough estimate of 0.10% for the IFR for influenza:
[click to enlarge]
The highest bar in the histogram at right shows the most likely IFR value of 0.05% or so. The upper bound of a 99% Credible Interval (bottom-left) let’s you know the highest IFR consistent with actual death data, and for M-pox it was 0.14%, which is similar to a severe flu season (example: 2014/15 and 2017/18 flu seasons).
Reference
Edouard Mathieu, Fiona Spooner, Saloni Dattani, Hannah Ritchie and Max Roser (2022) - “Mpox” Published online at OurWorldInData.org. Retrieved from: 'https://ourworldindata.org/mpox' [Online Resource]