4 ms·
If the number of cases are really as high as this paper suggests (5% infected) then we are one or two weeks away from a death wave like nothing else. However le
by whatshisface 7y ago
If the number of cases are really as high as this paper suggests (5% infected) then we are one or two weeks away from a death wave like nothing else. However let's pay attention to the many caveats the authors point out about their calculation, including the assumption that there are no particular individuals who are shedding especially large quantities of virus.
- martinko 7y agoOr the case fatality rate is a lot lower than we think it is.
- mannykannot 7y agoIndeed - there is nothing here to indicate a sudden jump (faster than the already high rate of propagation), while there is good reason to believe that the incidence has been systematically undercounted, on account of the shortage of testing kits. The fatality rate is less likely to be under-counted, if we assume most fatalities are preceded by hospitalization (and often testing.)
- grey-area 7y agoThis seems increasingly likely.
- whatshisface 7y agoThat would be infection fatality rate, case fatality rate is what we think it is by definition. (A case is a known case, an infection is an infection.) Quibbles over wording aside, I think the Icelandic data showed us that the percent asymptomatic is about 50%, insufficient for boundless optimism.
- Symmetry 7y agoAnd IIRC the Icelandic data doesn't tell us if those people were guaranteed to remain asymptomatic or were merely presymptomatic.
- leereeves 7y ago5% is the larger of two estimates given in the paper. The other estimate is 0.1%. These aren't an upper and lower bound, just two very different estimates. As the paper itself details, there are currently too many unknowns for these numbers to guide policy or expectations. But it is an interesting approach that might yield useful conclusions eventually.
- Symmetry 7y agoI'm hoping that, regardless of the absolute number, this method can give us a good handle on the relative number and let us know when infections are going up or down and by roughly how much. Right now the rate of new official infections has pretty much plateaued. My guess would be that actual infections are also plateaued but I can't rule out that just testing has plateaued.
- ianai 7y agoThe problem of course is lack of testing. The same factor censoring observations of the disease among the living censors even more strongly among the dead. AFAIK, no one is trying to test the dead. Ultimately though society became aware of the existence of plagues long before we knew about germ theory. So sticking our heads in the sand won’t save anyone from realizing what’s going on during a pandemic. It’ll just result in potentially many more dead.
- rumanator 7y ago> then we are one or two weeks away from a death wave like nothing else Just to put this sort of hyperbole into perspective, the Spanish Flu killed between 50 and 100 million people. The black plague was even worse. Even if a death wave is coming, the "like nothing else" is a very short-sighted remark.
- troughway 7y agoThere is a substantial amount of fear porn around this virus. It's sad that HN isn't immune (sorry) to the hyperbole.
- whatshisface 7y agoOkay, like nothing else in our lifetimes.
- caseymarquis 7y agoNote/Tangent: A quick search shows the black plague killed an estimated 25 million people over a century. Less than Spanish Flu, but a much higher percentage of the population at the time. Up to 95% of Native Americans died from diseases introduced from Europe. That's the largest percentage of a population to ever die from disease that I've heard of, and is the primary reason Native Americans are not the dominant population in the Americas today. Whether there's potential for an unprecedented amount of deaths really depends on how you measure things.
- rumanator 7y ago> Note/Tangent: A quick search shows the black plague killed an estimated 25 million people over a century. Those estimates sound a bit off by a factor of 5 to 10 regarding deaths, and they were concentrated in a 4 year period. https://en.wikipedia.org/wiki/Black_Death https://en.wikipedia.org/wiki/Black_Death Apparently the black death is estimated to cause the death of between 30 to 60% of Europe's entire population.
- caseymarquis 7y ago
- jeltz 7y agoRandom samples in Stockholm measured 2.5% infected.
- DennisP 7y agoFor any of those studies, look for the specificity of the test, which is basically the false positive rate. A serological test with 2.5% false positives is considered a pretty good test. If the test used in Stockholm were similar, they could have zero cases and get that result.
- gnulinux 7y agoIt's been 15 days since March 25. Most of these people are likely already sick, and if their disease will progress to severe, they're already in hospitals. Massive death wave might be coming, but it'll hardly be unexpected, they're all factored in the models.
- Barrin92 7y ago>we are one or two weeks away from a death wave like nothing else As of now in many countries the average death rate hasn't even gone up at all. Risk is not additive, many people die with covid-19, not of covid-19. So for a realistic estimate of the disease burden one has to compare the total increase in fatalities across some time all other factors held equal. This will be higher in some regions struck particularly hard like Italy, but not by as much as the plain number suggests.