There were some arithmetic errors in my post on COVID-19 risks for young people, which I've now corrected. https://srconstantin.github.io/2020/03/16/COVID-19-young-people-risks.html …
My best estimate for how likely a young person with a case of COVID-19 is to have a severe case is now 3%. Thanks to @LeoGrint for corrections.
It's on my list to look into that in more detail; the standard advice I see is treat it like a bad flu -- rest, fluids, self-quarantine -- plus a bunch of mandatory reporting that the doctors should do if you're confirmed to have COVID-19.