MikeG.":1nfwdwrx said:
lurker":1nfwdwrx said:
MikeG,
Just a thought, but have yours and your wife’s illnesses been captured by national statistics.
No.
I heard an expert on BBC yesterday say that the number of official cases are thought to represent about 1% of the actual number of cases. They can work back from the number of hospital admissions and arrive at a good estimate of the numbers in the country who have or have had the disease.
This is a strange supposition. I often work with incomplete data, but am not an expert on epidemics. However, we have 60 confirmed cases in my county, and many folks who suggest the same (that the actual number is a hundred times that).
The local university health system here has limited tests,but had tested 900 individuals at random who were asymptomatic or suspected to have another sickness and no family exposure to try to gauge community spread and they had no positive tests. Hard to tell how good that statistic is as it may not age well, but the expanded data that I've seen suggests fatalities will ultimately be something between just less than 1% of cases to slightly higher.
It is, in my opinion, dangerous for people to believe that there are 2-10K cases for each death or some such thing, as it keeps them from taking things seriously. In time, we'll get better numbers if there is widespread antibody testing at random, but that's probably pretty far off.
Our local health system didn't use their study to declare that there's no community spread (vs. family spread, etc), they actually said they believe that there is some community spread, but that it's not significant at this point.
They provide about half of the healthcare in a region with a few million people, and I guess their objective was to plan for capacity. I'm skeptical that the death rate will ultimately end up being a tiny fraction of a percent (between a tenth or a hundredth) as the now-recovered areas suggest otherwise.
As a ruse, I calculated the rough chance that the hospital system here could've done random testing of 900 people in a county with 1.2 million residents where 6000 were infected, and found no infections. The chance of that is about 1 in 100. True randomness in something like this is hard to achieve, of course. Is it possible that the 1 in 100 naysayers are correct? Sure. Is it likely? no.
It's only by chance that the percentage of suspected reported cases (1 in 100) is about the same as the chance of such an assumption being right based on these 900 tests.