ASA Connect

HI,
So I am noticing a trend in death count reporting that historical death changes are not entered on the day the death would have been reported but on the day it is determined the person died of Covid-19 even if the death was a couple weeks back.  Would there be anyway we can lobby for the deaths to be recorded on the day they occured or one day after even if they are discovered later, or for there to be away to accurately report death counts per day?  I understand case counts will be wrong but seems to me given the nature of this crisis we could at least make sure death counts per day are correct or nearly correct.  The methods they are using mean that spikes up might not be spikes at all.  I know all the states have differential reporting but  a correct count of death counts per day (as in the day the person died) seems like a reasonable ask and I wonder how the ASA could lobby for that data to be available.

For example in my state of Michigan,

https://www.michigan.gov/coronavirus/0,9753,7-406-98163-520743--,00.html

------------------------------
Laura Kapitula
Associate Professor
Grand Valley State University
------------------------------

Just curious, have you ever seen a death certificate filled out? 

When my mother died a few years ago, I was talking with her oncologist a few hours after she passed. There were 3 lines to hand write a (short as in under 20 letters) response. She had lung cancer and she was a smoker. But, she "died" from a heart attack due to complications of cancer. So, he put under cause of death: 1) Heart Attack 2) Cancer 3) Smoking

If this gets digitized, what would be the "cause of death" people look at? 

Plus, as a lecturer of Stats in Flint, I spent about a day talking to my students about this: 

https://www.washingtonpost.com/nation/2019/09/12/there-may-have-been-dozens-more-deaths-linked-flint-water-crisis-than-previously-disclosed/

There is also another issue with death certificates. 

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5692167/

How do we know the death certificate is correct in the first place? 






------------------------------
Andrew Ekstrom

Statistician, Chemist, HPC Abuser;-)
------------------------------

Original Message:
Sent: 04-24-2020 09:10
From: Laura Kapitula
Subject: Covid 19 Death Count Reporting

HI,
So I am noticing a trend in death count reporting that historical death changes are not entered on the day the death would have been reported but on the day it is determined the person died of Covid-19 even if the death was a couple weeks back.  Would there be anyway we can lobby for the deaths to be recorded on the day they occured or one day after even if they are discovered later, or for there to be away to accurately report death counts per day?  I understand case counts will be wrong but seems to me given the nature of this crisis we could at least make sure death counts per day are correct or nearly correct.  The methods they are using mean that spikes up might not be spikes at all.  I know all the states have differential reporting but  a correct count of death counts per day (as in the day the person died) seems like a reasonable ask and I wonder how the ASA could lobby for that data to be available.

For example in my state of Michigan,

https://www.michigan.gov/coronavirus/0,9753,7-406-98163-520743--,00.html

*Note on deaths (04/23/20): Regular reviews of death certificate data maintained in Vital Records reporting systems are conducted by MDHHS staff three times per week. As a part of this process, records that identify COVID-19 infection as a contributing factor to death are compared against all laboratory confirmed cases of COVID-19 in the Michigan Disease Surveillance System (MDSS). If a death certificate is matched to a confirmed COVID-19 case and that record in the MDSS does not indicate the individual died, the MDSS record is updated to indicate the death and the appropriate local health department is notified. These matched deaths are then included with mortality information posted to the Michigan Coronavirus website. As a result of the most recent assessment, today's data includes 55 additional deaths identified by this methodology.

------------------------------
Laura Kapitula
Associate Professor
Grand Valley State University
------------------------------

Interesting question. This might help.

https://www.cdc.gov/nchs/ppt/nchs2012/li-14_minino.pdf

The ICD-10 for Covid-19 is U07.1.  See https://www.cdc.gov/nchs/nvss/vsrr/covid19/index.htm

However, coding accuracy or specificity is a real problem when there are comorbidities. I have resigned myself to waiting for the excess deaths, all causes.

------------------------------
Thomas Messer
------------------------------

Original Message:
Sent: 04-27-2020 08:53
From: Andrew Ekstrom
Subject: Covid 19 Death Count Reporting

Just curious, have you ever seen a death certificate filled out?

When my mother died a few years ago, I was talking with her oncologist a few hours after she passed. There were 3 lines to hand write a (short as in under 20 letters) response. She had lung cancer and she was a smoker. But, she "died" from a heart attack due to complications of cancer. So, he put under cause of death: 1) Heart Attack 2) Cancer 3) Smoking

If this gets digitized, what would be the "cause of death" people look at?

Plus, as a lecturer of Stats in Flint, I spent about a day talking to my students about this:

https://www.washingtonpost.com/nation/2019/09/12/there-may-have-been-dozens-more-deaths-linked-flint-water-crisis-than-previously-disclosed/

There is also another issue with death certificates.

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5692167/

How do we know the death certificate is correct in the first place?






------------------------------
Andrew Ekstrom

Statistician, Chemist, HPC Abuser;-)

Original Message:
Sent: 04-24-2020 09:10
From: Laura Kapitula
Subject: Covid 19 Death Count Reporting

HI,
So I am noticing a trend in death count reporting that historical death changes are not entered on the day the death would have been reported but on the day it is determined the person died of Covid-19 even if the death was a couple weeks back.  Would there be anyway we can lobby for the deaths to be recorded on the day they occured or one day after even if they are discovered later, or for there to be away to accurately report death counts per day?  I understand case counts will be wrong but seems to me given the nature of this crisis we could at least make sure death counts per day are correct or nearly correct.  The methods they are using mean that spikes up might not be spikes at all.  I know all the states have differential reporting but  a correct count of death counts per day (as in the day the person died) seems like a reasonable ask and I wonder how the ASA could lobby for that data to be available.

For example in my state of Michigan,

https://www.michigan.gov/coronavirus/0,9753,7-406-98163-520743--,00.html

*Note on deaths (04/23/20): Regular reviews of death certificate data maintained in Vital Records reporting systems are conducted by MDHHS staff three times per week. As a part of this process, records that identify COVID-19 infection as a contributing factor to death are compared against all laboratory confirmed cases of COVID-19 in the Michigan Disease Surveillance System (MDSS). If a death certificate is matched to a confirmed COVID-19 case and that record in the MDSS does not indicate the individual died, the MDSS record is updated to indicate the death and the appropriate local health department is notified. These matched deaths are then included with mortality information posted to the Michigan Coronavirus website. As a result of the most recent assessment, today's data includes 55 additional deaths identified by this methodology.

------------------------------
Laura Kapitula
Associate Professor
Grand Valley State University
------------------------------

​Take a look at this: (from Washington Post, but this does not need a subscription)

https://www.dailymail.co.uk/health/article-8262903/15-000-excess-deaths-suggest-died-COVID.html

 When this is updated in August, we will get a good estimate of the deaths attributable to Covid-19 in the US.  It is likely double, especially if they factor in the reduction in auto deaths from 3000 per month last year.


If those projections hold, we have over 100,000 excess deaths and counting.

------------------------------
Jon Shuster
------------------------------

Original Message:
Sent: 04-27-2020 08:53
From: Andrew Ekstrom
Subject: Covid 19 Death Count Reporting

Just curious, have you ever seen a death certificate filled out?

When my mother died a few years ago, I was talking with her oncologist a few hours after she passed. There were 3 lines to hand write a (short as in under 20 letters) response. She had lung cancer and she was a smoker. But, she "died" from a heart attack due to complications of cancer. So, he put under cause of death: 1) Heart Attack 2) Cancer 3) Smoking

If this gets digitized, what would be the "cause of death" people look at?

Plus, as a lecturer of Stats in Flint, I spent about a day talking to my students about this:

https://www.washingtonpost.com/nation/2019/09/12/there-may-have-been-dozens-more-deaths-linked-flint-water-crisis-than-previously-disclosed/

There is also another issue with death certificates.

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5692167/

How do we know the death certificate is correct in the first place?






------------------------------
Andrew Ekstrom

Statistician, Chemist, HPC Abuser;-)

Original Message:
Sent: 04-24-2020 09:10
From: Laura Kapitula
Subject: Covid 19 Death Count Reporting

HI,
So I am noticing a trend in death count reporting that historical death changes are not entered on the day the death would have been reported but on the day it is determined the person died of Covid-19 even if the death was a couple weeks back.  Would there be anyway we can lobby for the deaths to be recorded on the day they occured or one day after even if they are discovered later, or for there to be away to accurately report death counts per day?  I understand case counts will be wrong but seems to me given the nature of this crisis we could at least make sure death counts per day are correct or nearly correct.  The methods they are using mean that spikes up might not be spikes at all.  I know all the states have differential reporting but  a correct count of death counts per day (as in the day the person died) seems like a reasonable ask and I wonder how the ASA could lobby for that data to be available.

For example in my state of Michigan,

https://www.michigan.gov/coronavirus/0,9753,7-406-98163-520743--,00.html

*Note on deaths (04/23/20): Regular reviews of death certificate data maintained in Vital Records reporting systems are conducted by MDHHS staff three times per week. As a part of this process, records that identify COVID-19 infection as a contributing factor to death are compared against all laboratory confirmed cases of COVID-19 in the Michigan Disease Surveillance System (MDSS). If a death certificate is matched to a confirmed COVID-19 case and that record in the MDSS does not indicate the individual died, the MDSS record is updated to indicate the death and the appropriate local health department is notified. These matched deaths are then included with mortality information posted to the Michigan Coronavirus website. As a result of the most recent assessment, today's data includes 55 additional deaths identified by this methodology.

------------------------------
Laura Kapitula
Associate Professor
Grand Valley State University
------------------------------

Hi Andrew,
Thanks for your comments.
I have worked with health data and know identifying a unique cause of death is not easy at all.  I was more speaking to the fact of people who were identified as having died from Covid-19 in the past based on autopsies or other diagnistic criteria all being reported on the same day instead of on the day they died.  If you reread my statement I was concerned with date of death being reported in cases who had been identified as having died of Covid-19.

Laura


------------------------------
Laura Kapitula
Associate Professor
Grand Valley State University
------------------------------

Original Message:
Sent: 04-27-2020 08:53
From: Andrew Ekstrom
Subject: Covid 19 Death Count Reporting

Just curious, have you ever seen a death certificate filled out?

When my mother died a few years ago, I was talking with her oncologist a few hours after she passed. There were 3 lines to hand write a (short as in under 20 letters) response. She had lung cancer and she was a smoker. But, she "died" from a heart attack due to complications of cancer. So, he put under cause of death: 1) Heart Attack 2) Cancer 3) Smoking

If this gets digitized, what would be the "cause of death" people look at?

Plus, as a lecturer of Stats in Flint, I spent about a day talking to my students about this:

https://www.washingtonpost.com/nation/2019/09/12/there-may-have-been-dozens-more-deaths-linked-flint-water-crisis-than-previously-disclosed/

There is also another issue with death certificates.

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5692167/

How do we know the death certificate is correct in the first place?






------------------------------
Andrew Ekstrom

Statistician, Chemist, HPC Abuser;-)

Original Message:
Sent: 04-24-2020 09:10
From: Laura Kapitula
Subject: Covid 19 Death Count Reporting

HI,
So I am noticing a trend in death count reporting that historical death changes are not entered on the day the death would have been reported but on the day it is determined the person died of Covid-19 even if the death was a couple weeks back.  Would there be anyway we can lobby for the deaths to be recorded on the day they occured or one day after even if they are discovered later, or for there to be away to accurately report death counts per day?  I understand case counts will be wrong but seems to me given the nature of this crisis we could at least make sure death counts per day are correct or nearly correct.  The methods they are using mean that spikes up might not be spikes at all.  I know all the states have differential reporting but  a correct count of death counts per day (as in the day the person died) seems like a reasonable ask and I wonder how the ASA could lobby for that data to be available.

For example in my state of Michigan,

https://www.michigan.gov/coronavirus/0,9753,7-406-98163-520743--,00.html

*Note on deaths (04/23/20): Regular reviews of death certificate data maintained in Vital Records reporting systems are conducted by MDHHS staff three times per week. As a part of this process, records that identify COVID-19 infection as a contributing factor to death are compared against all laboratory confirmed cases of COVID-19 in the Michigan Disease Surveillance System (MDSS). If a death certificate is matched to a confirmed COVID-19 case and that record in the MDSS does not indicate the individual died, the MDSS record is updated to indicate the death and the appropriate local health department is notified. These matched deaths are then included with mortality information posted to the Michigan Coronavirus website. As a result of the most recent assessment, today's data includes 55 additional deaths identified by this methodology.

------------------------------
Laura Kapitula
Associate Professor
Grand Valley State University
------------------------------