hi Harry
yes, I think the untested cases are missing, as you say.
maybe look here:
https://github.com/midas-network/COVID-19/tree/master/parameter_estimates/2019_novel_coronavirus
https://midasnetwork.us/covid-19/#resources
even if these don't include age distribution (I haven't looked) you might
take the US age distribution to see what the CDC age-specific figures
would predict for the overall CFR for the US, and compare that with the
material above.
2) BBC says "the symptoms start with a fever followed by a dry cough,
which can lead to breathing problems. This is a new, continuous cough and
means coughing a lot for more than an hour, or having three or more
coughing episodes in 24 hours (if you usually have a cough, it may be
worse than usual)."
So, the fact that someone coughs a few times does not mean they are
showing symptoms, or that they have COVID. But if it turns out that people
are in fact infectious before showing symptoms (as above), their
occasional cough could transmit without signifying infection. No?
Further, I'm sure I read somewhere (Andrei maybe?) that they think people
are infectious about one day before showing symptoms.
Greg
> Thanks, Greg,
>
> So, if I read that correctly, 'all cases' used as the denominator to
calculate the lower bound, includes only 'cases confirmed by state or
local
> public health laboratories as well as those with a positive test at the
state or local public health laboratories and confirmation at CDC'. That
would appear to exclude the untested so probably a fraction of the total
number of persons infected, resulting in a vastly inflated CFR?
>
> As it happens, I have another naive question about a tangentially
related
> matter. From what I've been reading, the virus is transmitted via
droplets
> spread through coughing and sneezing, and yet, it seems, most of those
infected got the virus from contact with people who have not yet
developed
> symptoms, like coughing. Sneezing is not supposed to be a COVID-19 symptom,
> anyway. So what's the mechanism?
>
> Ta.
>
> In solidarity,
> Harry
>
> On Fri, 20 Mar 2020 at 20:19, gregd <[log in to unmask]> wrote:
>
>> hi Harry, all
>> they explain what they mean by "cases" and why / how they present
ranges, in this para:
>> Data from cases reported from 49 states, the District of Columbia, and
three U.S. territories (5) to CDC during February 12–March 16 were
analyzed. Cases among persons repatriated to the United States from
Wuhan, China and from Japan (including patients repatriated from cruise
ships) were excluded. States and jurisdictions voluntarily reported
data
>> on laboratory-confirmed cases of COVID-19 using previously developed
data collection forms (6). The cases described in this report include
both COVID-19 cases confirmed by state or local public health
>> laboratories as well as those with a positive test at the state or
local
>> public health laboratories and confirmation at CDC. No data on serious
underlying health conditions were available. Data on these cases are
preliminary and are missing for some key characteristics of interest,
including hospitalization status (1,514), ICU admission (2,253), death
(2,001), and age (386). Because of these missing data, the percentages
of hospitalizations, ICU admissions, and deaths (case-fatality
>> percentages) were estimated as a range. The lower bound of these
percentages was estimated by using all cases within each age group as
denominators. The corresponding upper bound of these percentages was
estimated by using only cases with known information on each outcome as
denominators.
>> Greg
>> On 19/03/2020 23:18, Harry Feldman wrote:
>> > I'm still confused about what they're using for denominators? The CSC
document gives rates like '10-27%' for the oldest age group, which
isn't terribly informative. If they are using 'confirmed cases' for a
denominator, then for one thing, shouldn't they be able to pin it
down
>> > closer than a range of 17 points? And for another, won't that result
in an unrealistically high rate?
>> >
>> > On Fri, 20 Mar 2020 at 09:40, Greg Dropkin <[log in to unmask]> wrote:
>> >
>> >> Dear Vassilis
>> >>
>> >> does this do what you want?
>> >> https://www.cdc.gov/mmwr/volumes/69/wr/mm6912e2.htm?s_cid=mm6912e2_w
>> >>
>> >> Greg
>> >>
>> >>> Dear list, (apologies if this has been covered in previous
>> >> discussions)
>> >> Can somebody direct me to a source which gives the age distribution of
>> >> COVID-19 related deaths (preferably by country or sub-nationally; and
>> >> preferably with real(ish)-time updates)?
>> >>>
>> >>> I can find information about (much questionable) death rates by
>> >> age
>> >> group
>> >>> (e.g.,
>> >>>
>> >>
>> >
>> https://www.worldometers.info/coronavirus/coronavirus-age-sex-demographics/
),
>> >> but I have failed to find anywhere any information on the age-group
composition (or age distribution!) of COVID-19 related deaths. The
closest
>> >>> I have come is the data from Statista
>> >>>
>> >>
>> > (
>> https://www.statista.com/topics/5994/the-coronavirus-disease-covid-19-outbreak/
),
>> >> giving counts of _cases_ by age and of _deaths_ by region/country, but
>> >> not
>> >>> of deaths by age.
>> >>>
>> >>> Any suggestions?
>> >>>
>> >>>
>> >>> Thanks,
>> >>> Vassilis
>> >>> ==========================
>> >>> Dr Vassilis Monastiriotis
>> >>> European Institute, LSE
>> >>> ==========================
>> >>>
>> >>> ******************************************************
>> >>> Please note that if you press the 'Reply' button your
>> >>> message will go only to the sender of this message.
>> >>> If you want to reply to the whole list, use your mailer's
>> >>> 'Reply-to-All' button to send your message automatically
>> >>> to [log in to unmask]
>> >>> Disclaimer: The messages sent to this list are the views of the
>> >> sender
>> >> and
>> >>> cannot be assumed to be representative of the range of views held
>> >> by
>> >> subscribers to the Radical Statistics Group. To find out more about
Radical Statistics and its aims and activities and read current and
past
>> >> issues of our newsletter you are invited to visit our web site
>> >>> www.radstats.org.uk [1].
>> >>> *******************************************************
>> >>>
>> >>
>> >> ******************************************************
>> >> Please note that if you press the 'Reply' button your
>> >> message will go only to the sender of this message.
>> >> If you want to reply to the whole list, use your mailer's
>> >> 'Reply-to-All' button to send your message automatically
>> >> to [log in to unmask]
>> >> Disclaimer: The messages sent to this list are the views of the
sender and cannot be assumed to be representative of the range of
views held by subscribers to the Radical Statistics Group. To find
out more about Radical Statistics and its aims and activities and
read current and past issues of our newsletter you are invited to
visit our web site www.radstats.org.uk [1].
>> >> *******************************************************
>> >
>> >
>> > Links:
>> > ------
>> > [1] http://www.radstats.org.uk
>
> ******************************************************
> Please note that if you press the 'Reply' button your
> message will go only to the sender of this message.
> If you want to reply to the whole list, use your mailer's
> 'Reply-to-All' button to send your message automatically
> to [log in to unmask]
> Disclaimer: The messages sent to this list are the views of the sender
and
> cannot be assumed to be representative of the range of views held by
subscribers to the Radical Statistics Group. To find out more about
Radical Statistics and its aims and activities and read current and past
issues of our newsletter you are invited to visit our web site
> www.radstats.org.uk.
> *******************************************************
>
******************************************************
Please note that if you press the 'Reply' button your
message will go only to the sender of this message.
If you want to reply to the whole list, use your mailer's
'Reply-to-All' button to send your message automatically
to [log in to unmask]
Disclaimer: The messages sent to this list are the views of the sender and cannot be assumed to be representative of the range of views held by subscribers to the Radical Statistics Group. To find out more about Radical Statistics and its aims and activities and read current and past issues of our newsletter you are invited to visit our web site www.radstats.org.uk.
*******************************************************
|