http://www.pref.hokkaido.lg.jp/hf/kth/kak/hasseijoukyou.htm
Table shows quite a bit of data about each case including age, gender, location, any link discovered to other patients and status. The pdf shows additional data on when they had fever and other symptoms, when they went to different medical facilities and when they were confirmed as infected.
- https://data.humdata.org/dataset/novel-coronavirus-2019-ncov...
The GitHub repo with datasets:
- https://github.com/CSSEGISandData/COVID-19
Start there, and reach out to the staff if you’re looking for something more specific. They may be able to point you in the right direction.
You need to be able to figure out infected_at_t_and_dead/(infected_at_t_and_recovered+infected_at_t_and_dead) for some time t sufficiently in the past so that basically everyone falls into these two categories.
And if you want to draw transferable conclusions you also need an interval where treatment conditions were relatively uniform (e.g. already completely overwhelmed health system, or health system that could cope, throughout). As well as an age breakdown of recovered and dead.
The data from Italy does not look promising, they have an old population but seem to have reacted well and still have capacity to treat people. And yet CFR seems 3.8%, without excluding people who are ongoing and will die in the next days and weeks.
It would be very interesting to have a more detailed comparison with the South Korean outbreak, where again reaction has been very good as well and we could look at a subset of people for whom we have no undetected cases, so very good estimates ought to be possible.
E.g. for each case,