...what are the metrics/ranges that would change people's m ind about how much to worry about this even if media kept reporting the most fear mongering click-bait headlines they could possibly draw up? (that is of course the business model of today's media)
I wouldn't worry about the media. Just use your head, and look at past data, and interpret.
Reasons for the discrepancy:
***(?) Indicates speculative; there's not any evidence I'm aware of that would confirm those factors, but I've seen them mentioned.***
1) Increased testing -> finds more cases; CFR will go down. (IFR unchanged all else being equal)
2) Different population infected, older people self-isolating (about 10% of cases are older rather than 20-30%); CFR, IFR go down.
3) Lead-time bias - cases identified earlier, leads to greater lags.
4) Deaths lag. It takes weeks to months to die from this disease, after infection.
5) Hospital staff better at operating equipment and understanding best interventions to use. (CFR, IFR reduced)
6) Treatments improved (remdesivir, steroids at appropriate time) (CFR, IFR reduced)
7) Earlier case identification and treatment in at-risk individuals leads to better outcomes. (CFR, IFR reduced)
8) (?) Inoculum/Initial viral load -> Masks & other precautions have reduced the initial load which may reduce the severity of the disease
9) (?) Vitamin D production might improve outcomes in summer.
With 10% of cases from people over 65, I expect an IFR of about 0.3%. And a CFR of 1-1.5% with
current levels of testing and infection density. In other words, I believe we are locking in 600 deaths a day, minimum, in 2-3 weeks. That's on top of any other long tail deaths from March, April, and May, and you may need to add additional deaths from dramatically increased infections, in the event we're seeing 100k+ cases a day the next couple weeks (we'd be seeing in excess of 1000 deaths a day then in 3-4 weeks I would guess).
(I believe the IFR to date has been around 0.7% nationwide (there's a pretty wide range on this, but unlikely to be less than 0.6%). Exact IFR will vary depending on how many elderly people got infected and other specifics of the situation. So 0.3% would be a huge improvement - and most of that improvement is due to who is infected, so if we see break-in to the elderly population again, it will go up. AZ data suggests more old people are now getting infected, as infection pressure increases (about 13% today, up from 10%).)
Right now covid19-projections predicts nothing of the sort in terms of 1000+ deaths a day, but I believe their model will lag, since I believe it at least partially bases its predictions on deaths, which lag.
COVID-19 Projections Using Machine Learning
I will say their prediction for total deaths has steadily increased from ~185k to 205k now in the last week or so. I expect it to revise up rapidly in the next few days, as it picks up the trend in AZ, TX, FL, and CA.
Hopefully I'm totally wrong.
