Until a vaccine is developed and distributed, the only metric likely to bring any comfort is a lowering of the likely Case Fatality Rate (CFR). Since "known cases" come from testing and testing is not universal....what is the best guess as to the "estimated actual cases" and what would be the CFR based upon that?

Not everyone that dies gets tested either so the even with an educated guess on total cases, you would have to turn around and do the same for deaths......

Here's the best calculated guess I can come up with:
Apr 1 deaths: 5102
Assume 0.5% "true" mortality rate and avg. 18 day lag between getting infected and dying.
That gives 1,020,400 "true" cases on mar. 14
Apply R-naught of 1.3 (similar to flu) starting mar. 17 when serious efforts to slow virus down began (R-naught is thought to have been between 2 and 3 at beginning w/ no measures)
That leads to approx. 2.3M "true" cases today, 7.8M cases end of April, 20.3M cases end of May, and 34M "true" cases by end of June.
Apply .5% mortality to +18 days from that, and you get a 170k death count around 3rd week of July, which is in range of 100k-200k that trump cited

Here is my post from another thread attempting to do just this.

I haven't seen this look yet but decided to look into estimating what our actual case load might look like based on various CFRs. We all know we don't know the "denominator" so here are some looks ate a few different scenarios

Methodology: To come up with these estimated case load curves, I assumed the death count trails the case count by 2 weeks, so I took the death total for each day, and use that and the stated CFR to estimate what the total case load was 14 days earlier. For the 2 most current weeks, where 2 week out death numbers are not available, I looked at what the recent trend of the ration of estimated cases to confirmed cases was doing and carried that trend out for 2 weeks and applied that multiplier to the confirmed case count to get the estimated case count. here is a look at that ratio trend for illustration:


A few notes:

• First, 0.02% CFR is currently the absolute lower limit for possibilities, because this means that every single person in the USA currently has been infected by COVID-19. So not the lowest feasible CFR, but the lowest that is physically possible.
• If the CFR is the same as the flu (0.1%), we have already surpassed the worst flu year in the last decade by 44% and just breached the upper bound of the estimated H1N1 case load from the 2009 pandemic at only 6.5 weeks into the pandemic. Also, if not a singled person more gets COVID (unlikely), we will top out at 65k deaths around 2 weeks from now.

1. For most people, this exercise reveals more about your biases than any hard science.

2. I'm not sure it's worth much time following the CFR live day to day. The broad majority of the cases are still active. For example, if some how we had 0 new confirmed cases after today, our CFR would go up day by day as people who are active cases inevitably die.

3. I'm doubly not sure following the CFR day by day is worth it because it seems to heavily influenced by the age distribution of those infected. It seems pretty consistent among all countries the rates at which people die by age. So a low CFR could just mean a younger infected population. A spike in CFR could mean that a lot of old people got infected, in such a case where a nursing home got hit.

4. To gauge how we are doing, it may be better to pay attention to how our hospitals are holding up. Are there enough beds, ventilators, and medicine? We know in certain areas where the virus has overran the hospital capabilities, Required triage of life saving equipment such as ventilators mean that some people may die that could have potentially survived with those treatments. At this point in the US, I don't believe any location is at that point, including NYC. However, it is something to watch as we approach our peak in the coming weeks.

5. If your interest is to gauge how ready we are to approach normalcy, I would probably echo what Scott Gottleib, Trumps former FDA commissioner said were some of the main markers:

• 2 weeks of declining cases/hospitalizatoins due to COVID
• Enough hospital resources to handle the case loads
• Enough testing to test all symptomatic people
• Technology and capability developed to do advanced contact tracing

We are making progress on almost all fronts, although daily test capacity has not really increased for the last week and that is seriously dissappointing.

I applaud your dedication to trying to predict a CFR but I'm not sure this is the best measure of severity or the best way to ballpark surge volume. I have switched to looking at the death trend because it is going to give a more accurate representation of how we are doing bending the curve and then solve to determine likely surge volume and capacity. This is understanding that any social distancing measures we take to today won't show any changes for 1-2 weeks given the incubation time and time for the disease to present.