You’re right. That should be done. Sort of like ‘polling’ the population.
That should be done. Sort of like polling the population.
This has been unfolding so fast and the sick take so relatively long to recover that little solid data has been available. Some of the attempts so far were in the sample size of 12 to 30 infected range. If we could trust China’s figures, we are looking at about 4-5% fatalities. That said, China still has about 8k listed as still sick despite few new cases within the past 3 weeks, and the number dying in these end cases has been over the 4.0%.
For the rest of the world, we’re still in the early expansion range where fatalities are overstated because it takes longer to recover than to die. It’s been hanging in there at an illusory 30-35% for more than a week.
In the US we’ve begun to move from the first death lag phase where people are sick, but haven’t begun to die yet (WA nursing home excepted), where the fatality rate is understated, into the recovery-lag phase, where the fatality rate is greatly over-stated.
South Korea has been testing everyone, so their baseline numbers will likely be pretty close at the end. Currently they are running about a 5.5% fatality rate for resolved cases, with only 18.3% resolved.
Because of these lags, anything less than cohort studies is going to be off, especially for a while.