[burger]

I bemoaned not having some stats on age versus serious cases of Covid-19.  So I am deriving some from table 1 of this NEJM article from China (data ended Jan 29) of 1099 patients.  NOTE: I do not know how these patients were chosen.  That might affect the validity of any conclusions one might draw.

Percentages are number of severe cases / total number of cases in age group
Age 0-14:  11.1%  1 of 9 cases were severe.
Age 15-49: 12%   67 of 557 cases were severe.
Age 50-64: 17.5%  51 of 292 cases were severe.
Age 65+:  28.8%  44 of 153 cases were severe.
NOTE: These numbers are minimums.  More of the cases could have become severe after the data was taken.

Percentages are number of severe cases in the age group / total number of cases (ie, 163)
Age 0-14:  0.6%
Age 15-49: 41.1%
Age 50-64: 31.3%
Age 65+:  27.0%
My conclusion:  Those in the 15-49 group may be at 12% risk of a serious case, but the age group's serious cases will be a major impact on the health system.

Note that ages are only given for 1011 of the patients.  I do not know why 88 are not counted.  (Death?)

Note that this study seemed to use patients that were caught very early. 
Of the 280 or so, for which exposure was known, they were admitted 4 days after exposure.
Only 44% had a fever on admission. 68% had a cough.
To me, this means that the selection criteria was simply a positive test prior to or concurrent to first seeking medical attention. 

I am neither a physician nor a statistician.

As with all these studies, the results are only as accurate as the testing is complete.  So if they were not testing everyone or a random sample, the denominators in all of those percentages are uncertain, and so are the results.  And it's not clear from the paper how they decided whom to test.  The best you can say that these are the percentages of serious cases for those that were tested and positive and hospitalized.  Generalizing to all CV cases is impossible.

That said, the increase with age is completely consistent with a bunch of other results.  

The missing 88 people is probably just cases where the age or some other data point was not recorded.  Easy to imagine that happening in a crisis.

[burger]

I bemoaned not having some stats on age versus serious cases of Covid-19.  So I am deriving some from table 1 of this NEJM article from China (data ended Jan 29) of 1099 patients.  NOTE: I do not know how these patients were chosen.  That might affect the validity of any conclusions one might draw.

Percentages are number of severe cases / total number of cases in age group
Age 0-14:  11.1%  1 of 9 cases were severe.
Age 15-49: 12%   67 of 557 cases were severe.
Age 50-64: 17.5%  51 of 292 cases were severe.
Age 65+:  28.8%  44 of 153 cases were severe.
NOTE: These numbers are minimums.  More of the cases could have become severe after the data was taken.

Percentages are number of severe cases in the age group / total number of cases (ie, 163)
Age 0-14:  0.6%
Age 15-49: 41.1%
Age 50-64: 31.3%
Age 65+:  27.0%
My conclusion:  Those in the 15-49 group may be at 12% risk of a serious case, but the age group's serious cases will be a major impact on the health system.

Note that ages are only given for 1011 of the patients.  I do not know why 88 are not counted.  (Death?)

Note that this study seemed to use patients that were caught very early. 
Of the 280 or so, for which exposure was known, they were admitted 4 days after exposure.
Only 44% had a fever on admission. 68% had a cough.
To me, this means that the selection criteria was simply a positive test prior to or concurrent to first seeking medical attention. 

I am neither a physician nor a statistician.

As with all these studies, the results are only as accurate as the testing is complete.  So if they were not testing everyone or a random sample, the denominators in all of those percentages are uncertain, and so are the results.  And it's not clear from the paper how they decided whom to test.  The best you can say that these are the percentages of serious cases for those that were tested and positive and hospitalized.  Generalizing to all CV cases is impossible.

That said, the increase with age is completely consistent with a bunch of other results.  

The missing 88 people is probably just cases where the age or some other data point was not recorded.  Easy to imagine that happening in a crisis.