https://ncov.dxy.cn/ncovh5/view/pneumonia?from=groupmessage&isappinstalled=0截至 2020-02-20 17:28 全国数据统计
数据说明
55,051 现存确诊 -1,335 较昨日
5,206 现存疑似 +1,614 较昨日
11,633 现存重症 -231较昨日
75,567 累计确诊 +892 较昨日
2,239 累计死亡 +118 较昨日
18,277 累计治愈 +2,109 较昨日
As of 2020-02-20 17:28 National Statistics
the data shows
54,051 Existing confirmed diagnosis -1,335 since yesterday
5,206 Suspected +1,614 since yesterday
11,633 Existing severe illness -231 since yesterday
75,567 Cumulative diagnoses +892 since yesterday
2,239 Cumulative deaths +118 since yesterday
18,277 Cumulative Healing +2,109 since yesterday
https://gisanddata.maps.arcgis.com/apps/opsdashboard/index.html#/bda7594740fd40299423467b48e9ecf6Johns Hopkins
76,243 Confirmed
2,247 Deaths
18,424 Recovered
https://nextstrain.org/ncov?l=unrooted&m=divNext Strain
103 analyzed strains
We can bound the case fatality rate. We know it takes longer to recover from the disease than to die from it. As a result, the CFR cannot be higher than the total deaths/total recovered and is likely much less. That puts an upper bound on the CFR at 12.2%
A similar naive calculation (one not taking account of the actual dynamics) based on the Deaths/Confirmed and Deaths/Recovered puts those at 3.0% and 24.2%. The total of those should be 100%. Clearly it is far from it, so the dynamics are still very much at work (i.e. it takes time to die or recover). But that does set a lower bound on the CFR at 3.0%.
Do remember though that these bounds are for those who report to hospital. And since we lack any data at all about those who do not report to hospital, and how that population compares in size to the hospitalized portion, we cannot assess what that means about the actual population wide CFR.
Also, the age distribution issue still applies. This disease kills people over 60 and few under 60. So the utility of a simple all-age CFR at all is dubious.
The same cautions still apply. The changing case definition and other issues make it very difficult to compare these numbers day to day or week to week, let alone to use them to infer things about the communities at large.
Oren,
The reports out of the four countries (Japan, South Korea, Iran and Singapore) are indeed concerning, as are the movement of the infected folks off the cruise ship(s). These all greatly increase the chance that this evades control and becomes an international pandemic.
The markets are just now beginning to take note. Goldman Sachs is now cautioning about a potential 10% correction soon. That is likely just the beginning.
Markets and suppliers are also now beginning to realize the implications for supply chain disruptions. I suspect there are a whole lot of corporate heads and corporate board members reconsidering the wisdom of a highly diversified and interdependent supply chain.
This is nothing new. Centralization and decentralization in all its many forms feels an awful lot like a perpetual motion machine.
Sam
Addenda
Something I wonder about from this data.
Note that the existing confirmed diagnosis tally is declining. If we subtract that from the Cumulative Diagnoses, we get 21,516. That is pretty close to the total cumulative tally of deaths and recoveries (2,239 and 18, 277, total = 20,516). I do not like that these numbers are precisely 1,000 apart.
However, if we can assume that the long early tail of diagnoses, deaths and recoveries is large compared to the current count of those ill, then the deaths/diagnosed ratio might set a low bound on the CFR. (2,239/21,516 = 10.9%). That is very close to the reported value for the original SARS at 10%.
It is tempting to think this is meaningful. Not knowing the pedigree of the data, I really can't even speculate on that.
If true, this also suggest that we should expect (or fear) an additional ~6,000 deaths from those already confirmed infected and in treatment.