Read on Twitter
Thread on a potential silver lining of #COVID19
My friend @BenMazer is a pathologist and he's noticed huge declines in specimens obtained for cancer diagnoses.
Reason: hospitals have stopped elective procedures, which includes cancer screening procedures
My friend @BenMazer is a pathologist and he's noticed huge declines in specimens obtained for cancer diagnoses.
Reason: hospitals have stopped elective procedures, which includes cancer screening procedures
Ben wrote a column about this natural experiment on @medscape here> https://www.medscape.com/viewarticle/929089
He is great writer and this column is worth a look. I am intrigued by this topic b/c it's not just cancer screening. Lots of screening happens in cardiology too. Think #applewatch
He is great writer and this column is worth a look. I am intrigued by this topic b/c it's not just cancer screening. Lots of screening happens in cardiology too. Think #applewatch
I worry about screening because medicine is most pure when patients ask for our help. Screening is anticipatory medicine in which we tell healthy people they need us to help them. Screening also puts out huge caring signals -- Think Elephant in the Brain. @robinhanson
The problem with screening for one disease is hubris. My gosh, it only takes a few months on in-patient rotation in general medicine or surgery to know there are hundreds if not thousands of diseases that can kill you.
So the best measure of whether cancer screening actually works is not whether it reduces the chance of dying from the screened disease but whether it reduces the risk of dying from anything--or more simply: does it extend life?
My friend @VPrasadMDMPH has written in the BMJ that cancer screening has never been shown to "save lives"
https://www.bmj.com/content/352/bmj.h6080/rapid-responses
A huge lung CA screening trial this year supported his thesis: https://www.nejm.org/doi/full/10.1056/NEJMoa1911793
https://www.bmj.com/content/352/bmj.h6080/rapid-responses
A huge lung CA screening trial this year supported his thesis: https://www.nejm.org/doi/full/10.1056/NEJMoa1911793
One of the most important studies on this matter is a systematic review of all studies on screening by Saquib et al https://www.ncbi.nlm.nih.gov/pubmed/25596211
Though there are exceptions --cerv CA screening in underserved areas-- most cancer screening does not extend life.
Though there are exceptions --cerv CA screening in underserved areas-- most cancer screening does not extend life.
Besides the obvious reason of competing causes of death, there is the nasty problem of overdiagnosis and overtreatment. @BenMazer and I wrote about this and the fact that there is no "overdiagnosis awareness month." https://www.wbur.org/cognoscenti/2017/12/08/does-cancer-screening-save-more-lives-overall-not-necessarily
Now the tie in w #COVID19. First: specificity /false positives. Serology testing has shown that if there is low prevalence of disease say 1 in 1000, using a test w 95% specificity, a pos titer means your chance of having virus is only 2% See @bogdienache https://twitter.com/bogdienache/status/1254122522687279105
Science Friday covered this messy math on mammograms. https://www.sciencefriday.com/articles/math-mammograms/
Most doctors don't understand it --hence there has been so little resistance to anticipatory medicine
Most doctors don't understand it --hence there has been so little resistance to anticipatory medicine
Another tie-in w #COVID19 is the notion of overall death vs COVID19 deaths. Yes, there are hot spots that have seen huge increases of deaths from the virus. And the lockdowns and cancellation of elective procedures are necessary to prevent NYC/Lombardy/Spain like scenarios
Many people are focused on the day-to-day graphs of cases and deaths from #COVID19 -- no doubt we want to be aware of outbreaks and act on them. But the endpoint of this natural experiment is a year or two from now.
Will cities or countries that have looser policies on social distancing, say Sweden, end up better in long-run b/c of less disruption of the economy and less non-covid deaths? The only way to know is counting both covid and non-covid outcomes. Like we should do w Ca screening
