Sunday, April 12, 2020
I don't celebrate Easter but, for the second time this week, I feel sad by the disruption to traditions and the disappointment of not gathering with family. I'm sure like my family's virtual Seder this week there will be lots of tele- meals and visits today. I know there will be bright spots to these events (to be sure, there was a family member who purposefully chose not to wear pants to our virtual Seder), but the scary disruption of a pandemic sure does suck and these kinds of days drive that point home.
Despite the holiday (and the fact that it is Sunday), there is a lot to review.
----------
Latest from the FT
https://www.ft.com/coronavirus-latest
I received some comments (both via email and on our Q&A sessions) about other ways to slice and dice these numbers. Why the FT graphs? Why not per capita or why not some other measures? Here are my thoughts:
The variability of testing (by US State and by country) is highly variable.
Cause of death is probably incorrect in many, many instances leading to undercounting.
There are confounders with regards to rates of infection based on travel patterns and population density.
The disease is like a wave moving through geographies, so any data has to be normalized to a time course from patient 0 (or some similar anchor point).
Thus, the FT data set may be as good as it gets in a quickly digestible form.
We get a 7-day rolling avg death rate since the day of 3 daily deaths in the country.
We get a 7-day rolling avg death rate by logical geographic region- "NY" or "Lombardy."
We get 7-day rolling avg confirmed cases since 30 daily cases were confirmed
We get a lot of other country-specific rates and relative comparisons, giving us a quick ability to monitor change over time.
Travis sent in an interactive graph of state-level data that allows the user to play with some settings, including linear vs. log scaling.
https://time.com/5814139/coronavirus-state-data-tracker/
Here is another source of interactive country-level data that is very comprehensive, including PER CAPITA deaths
https://ourworldindata.org/coronavirus-data
Per Capita Death data https://ourworldindata.org/grapher/total-covid-deaths-per-million
This site is well done but is not digestible daily.
Here is an analysis and discussion of the logical fallacies you can make by not appreciating confounders in these data sets
https://twitter.com/MaxCRoser/status/1248891496402747392
All of this data does not capture the real impact, which is the saturation of the healthcare system, prolonged illness, and a disturbingly high mortality rate.
https://twitter.com/nikhil_palsingh/status/1248597786360262656
and another article on the missing other illnesses not being seen
https://www.nytimes.com/2020/04/06/well/live/coronavirus-doctors-hospitals-emergency-care-heart-attack-stroke.html
Yesterday, I covered the Remdisivir data from 4/10 in detail. I found information on other treatments in the last 24 hours.
Here is some discussion of a blinded study on HCQ out of Brazil. Note - this is in critically ill patients. BUT we don't know who to treat, when to treat, if this really works, etc.
https://twitter.com/GaetanBurgio/status/1249255758002180096
BCG vaccine as an "immune activator" also popped up in the media in the last few weeks. I am starting to see some critical analysis. Here is a thoughtful review looking at the data behind the media reports. Pay attention to the logical fallacies. https://naturemicrobiologycommunity.nature.com/channels/303-journal-club/posts/64892-universal-bcg-vaccination-and-protection-against-covid-19-critique-of-an-ecological-study
It is not only the public that wants answers, treatments, and end to this. I am starting to see commentary by physicians who are frustrated by hospitals not sharing data and lack of transparency from governments. Here is a thoughtful piece by a physician on KevinMd.com
https://www.kevinmd.com/blog/2020/04/truth-dies-in-silence-sadly-so-do-people.html
However, even doctors can get it wrong. Read the paragraph on his frustration over treatments. He is confounding lack of data sharing/transparency with the lack of good evidence on the indications and efficacy of treatments (and threw in the social shaming as well). As I have said, there is a big difference between any ONE person deciding to try a therapy, like HCQ or convalescent plasma, and making it "the standard of care." The tension is painful for families struggling with critically ill loved ones and differently painful for health care providers who want to do something rather than nothing. But without clinical trials and data, all we have is anecdotal information. And that does not help us in the long-term. There will be plenty more discussions like this.
@compoundchemistry delivers again!
infographics on chocolate
https://twitter.com/compoundchem/status/1248914940871602178/photo/1
---Bonus Round - Capturing Nuance Edition
I love candid portrait photography. It annoys my family and friends, but capturing that look or expression the reflects a person's personality can be so meaningful. Alfred Eisenstaedt is one of the photographers I try to emulate for this kind of work. He's the Life photographer who shot the couple kissing in Times Square on V-J day. I came across the story of one of his other photos, which is an excellent example of context and nuance.
https://petapixel.com/2013/03/31/eyes-of-hate-captured-in-portrait-of-nazi-politician-by-jewish-photographer/
Background: https://en.wikipedia.org/wiki/Alfred_Eisenstaedt
William Manchester was an outstanding historian and biographer. Not only did he write [my personal favorite] the 3-volume biography on Churchill, but a biography on McArthur and his most famous book on the Kennedy assassination. However, I think his most compelling writing is Goodbye, Darkness, a memoir of his time as a Marine in WWII published in 1980. It was the first WWII book that opened up the details of the Pacific campaign for me - capturing the pain, the mundanity, the seeming pointlessness of the violence, and the PTSD often not discussed. It was a very personal and microscopic look at WWII that presupposed the kind of tone set by Band of Brothers and Saving Private Ryan. And, as a bonus to my bonus, I learned Manchester was a mentee of HL Mencken for a time when they worked in Baltimore together. Amazing.
Background: https://en.wikipedia.org/wiki/William_Manchester
https://www.goodreads.com/book/show/19801.Goodbye_Darkness
Clean hands and sharp minds
AW
I don't celebrate Easter but, for the second time this week, I feel sad by the disruption to traditions and the disappointment of not gathering with family. I'm sure like my family's virtual Seder this week there will be lots of tele- meals and visits today. I know there will be bright spots to these events (to be sure, there was a family member who purposefully chose not to wear pants to our virtual Seder), but the scary disruption of a pandemic sure does suck and these kinds of days drive that point home.
Despite the holiday (and the fact that it is Sunday), there is a lot to review.
----------
Latest from the FT
https://www.ft.com/coronavirus-latest
I received some comments (both via email and on our Q&A sessions) about other ways to slice and dice these numbers. Why the FT graphs? Why not per capita or why not some other measures? Here are my thoughts:
The variability of testing (by US State and by country) is highly variable.
Cause of death is probably incorrect in many, many instances leading to undercounting.
There are confounders with regards to rates of infection based on travel patterns and population density.
The disease is like a wave moving through geographies, so any data has to be normalized to a time course from patient 0 (or some similar anchor point).
Thus, the FT data set may be as good as it gets in a quickly digestible form.
We get a 7-day rolling avg death rate since the day of 3 daily deaths in the country.
We get a 7-day rolling avg death rate by logical geographic region- "NY" or "Lombardy."
We get 7-day rolling avg confirmed cases since 30 daily cases were confirmed
We get a lot of other country-specific rates and relative comparisons, giving us a quick ability to monitor change over time.
Travis sent in an interactive graph of state-level data that allows the user to play with some settings, including linear vs. log scaling.
https://time.com/5814139/coronavirus-state-data-tracker/
Here is another source of interactive country-level data that is very comprehensive, including PER CAPITA deaths
https://ourworldindata.org/coronavirus-data
Per Capita Death data https://ourworldindata.org/grapher/total-covid-deaths-per-million
This site is well done but is not digestible daily.
Here is an analysis and discussion of the logical fallacies you can make by not appreciating confounders in these data sets
https://twitter.com/MaxCRoser/status/1248891496402747392
All of this data does not capture the real impact, which is the saturation of the healthcare system, prolonged illness, and a disturbingly high mortality rate.
https://twitter.com/nikhil_palsingh/status/1248597786360262656
and another article on the missing other illnesses not being seen
https://www.nytimes.com/2020/04/06/well/live/coronavirus-doctors-hospitals-emergency-care-heart-attack-stroke.html
Yesterday, I covered the Remdisivir data from 4/10 in detail. I found information on other treatments in the last 24 hours.
Here is some discussion of a blinded study on HCQ out of Brazil. Note - this is in critically ill patients. BUT we don't know who to treat, when to treat, if this really works, etc.
https://twitter.com/GaetanBurgio/status/1249255758002180096
BCG vaccine as an "immune activator" also popped up in the media in the last few weeks. I am starting to see some critical analysis. Here is a thoughtful review looking at the data behind the media reports. Pay attention to the logical fallacies. https://naturemicrobiologycommunity.nature.com/channels/303-journal-club/posts/64892-universal-bcg-vaccination-and-protection-against-covid-19-critique-of-an-ecological-study
It is not only the public that wants answers, treatments, and end to this. I am starting to see commentary by physicians who are frustrated by hospitals not sharing data and lack of transparency from governments. Here is a thoughtful piece by a physician on KevinMd.com
https://www.kevinmd.com/blog/2020/04/truth-dies-in-silence-sadly-so-do-people.html
However, even doctors can get it wrong. Read the paragraph on his frustration over treatments. He is confounding lack of data sharing/transparency with the lack of good evidence on the indications and efficacy of treatments (and threw in the social shaming as well). As I have said, there is a big difference between any ONE person deciding to try a therapy, like HCQ or convalescent plasma, and making it "the standard of care." The tension is painful for families struggling with critically ill loved ones and differently painful for health care providers who want to do something rather than nothing. But without clinical trials and data, all we have is anecdotal information. And that does not help us in the long-term. There will be plenty more discussions like this.
@compoundchemistry delivers again!
infographics on chocolate
https://twitter.com/compoundchem/status/1248914940871602178/photo/1
---Bonus Round - Capturing Nuance Edition
I love candid portrait photography. It annoys my family and friends, but capturing that look or expression the reflects a person's personality can be so meaningful. Alfred Eisenstaedt is one of the photographers I try to emulate for this kind of work. He's the Life photographer who shot the couple kissing in Times Square on V-J day. I came across the story of one of his other photos, which is an excellent example of context and nuance.
https://petapixel.com/2013/03/31/eyes-of-hate-captured-in-portrait-of-nazi-politician-by-jewish-photographer/
Background: https://en.wikipedia.org/wiki/Alfred_Eisenstaedt
William Manchester was an outstanding historian and biographer. Not only did he write [my personal favorite] the 3-volume biography on Churchill, but a biography on McArthur and his most famous book on the Kennedy assassination. However, I think his most compelling writing is Goodbye, Darkness, a memoir of his time as a Marine in WWII published in 1980. It was the first WWII book that opened up the details of the Pacific campaign for me - capturing the pain, the mundanity, the seeming pointlessness of the violence, and the PTSD often not discussed. It was a very personal and microscopic look at WWII that presupposed the kind of tone set by Band of Brothers and Saving Private Ryan. And, as a bonus to my bonus, I learned Manchester was a mentee of HL Mencken for a time when they worked in Baltimore together. Amazing.
Background: https://en.wikipedia.org/wiki/William_Manchester
https://www.goodreads.com/book/show/19801.Goodbye_Darkness
Clean hands and sharp minds
AW
Comments
Post a Comment