Here’s an update from the Inside Medicine Covid metrics dashboard. Last week, we gave a state-by-state readout focused on hospitalizations, along with some wastewater trends (the best proxy for community virus levels these days). This time around, we are changing up the format and adding a visual component so that you can see how things compare both to recent times (in addition to the previous week, previous peaks, and all-time peaks). The idea is to give you context for these numbers. (
Well done, thank you. I appreciate any time there's "an adult in the room" providing rational statistical data with caveats appropriately addresses. It's really helpful, at least to me.
A few reasons why mass testing may not make sense:
1) Low Cost/Benefit. If a hospital has 1000 employees and has 200 patients pass through a day, that's $12,000/day in rapid tests and $180,000 day in PCR tests.
Considering that non-symptomatic people are extremely unlikely to test positive on either rapid or PCR, it doesn't make sense to spend limited resources (labor, lab capacity, money) without clinical diagnosis. This is how we have always utilized these types of tests in the past - only testing those showing symptoms, whether it is flu, RSV, strep, TB, etc - so what we are seeing is a return back to how it was done pre-2020 for ILI.
2) Inefficient use of laboratory capacity. There is a fixed constraint on # of total types of tests labs can process. Adding on 'n' number of tests almost certain to be negative is at the cost of other tests done based on clinical assessment (HIV, Epstein Barr, Strep, HPV, etc).
3) It often won't change clinical protocols - especially in 2023. If you are not sick but test positive for Covid, that isn't going to change your course of treatment. And again, if you aren't sick, you almost certainly aren't going to test positive for Covid anyway.
4) Both PCR and Antigen tests are notoriously bad detecting infection before symptoms begin, which is why if you aren't sick, you aren't likely to test positive. By the time you are sick, then it can make sense to run a battery of tests to determine if antibiotics, tamiflu, paxlovid, or chicken noodle soup are the best course of treatment. (And PCR tests are also bad at telling when you aren't contagious any longer [1])
5) We tried mass testing and found it didn't work. We stopped mass testing, and paid no price for doing so. To me, the hypothesis that mass testing can reduce spread in communities appears falsified across the world as it failed both conditions (doing it didn't stop things, stopping it didn't make things worse).
I initially thought it would work, buying into the hype from Michael Mina's piece in Time back in November of 2020:
Where he claimed: "Countries like Slovakia and the United Kingdom are currently utilizing mass rapid antigen testing programs and already seeing great success."
But when I started following this claim over the next few months, it was falsified.
Yet it was never walked back (I later learned he quick academia to join a testing company, so perhaps he had vesting interest not walking back that claim?)
And, over a year later, Michael Mina demonstrated in real time on Twitter why mass testing didn't work - they don't catch infection until your symptoms show up, which means the old conventional wisdom of "stay home if sick, go to work/school if healthy" didn't require $20-$150 to validate:
I think this explains why mass testing didn't work in South Korea, Denmark, Germany, or anywhere else in the world either. (there are other possibilities why it doesn't work as well).
Interesting, but if there's no testing, then reinstating masks, preferably respirators makes sense. Contracting a Covid-19 nosocomial infection is a serious issue with high morbidity and mortality. As the pandemic continues to roll on more people, including healthcare workers and patients will get reinfected over and over, which potentially decimates healthcare systems and increases the odds for long covid.
Thank you for the great information. What is the reasoning behind no longer testing hospitalized patients upon admission? This does not seem to make sense in light of the many vulnerable patients or accurate statistics? What is now required in nursing homes?
Can I assume your hospital is still testing if a patient is symptomatic for Covid upon admission?
That is what is still happening here at the Cleveland Clinic, UH, and some hospitals we have been to in Pittsburgh - if patient is admitted with "Influenza like illness" symptoms, they get Flu, Covid, sometimes Strep tests. Otherwise, no tests ILI for broken arms, cancer, in-patient surgery, well visits, etc. Is this how it is in Boston too?
I don't think it's set in stone. If symptomatic, certainly people check. For trauma, not so much. The question I have is for things that viral illnesses could trigger. Heart attacks, congestive heart failure, diabetic crisis...these things may be somewhat asymptomatic from the "classic COVID" standpoint, but be the tip-over cases which I always talk about. And of course they can be contagious in any case...
Left a lengthy comment above trying to "steelman" why mass testing may not be helpful [1], but it comes down to two things (I think):
1) Does testing catch pre-symptomatic people, the answer appears to be largely "No"
My thought process heavily hangs on that assumption - if a hospital was running 10,000 tests a week and only 1 or 2 came back positive on asymptomatic people, I'd have a hard time seeing ROI especially if:
2) Would a positive pre-symptomatic test change course of care? Around here, that answer is also "No", perhaps different protocols in Boston? And if done with PCR, the assumption may be "they cleared a previous Covid infection and
didn't realize it, this is just leftover DNA"
If you have a patient come in the ER with heart attack symptoms, no other respiratory issues, would you prescribe Paxlovid if a PCR test showed positive?
**This might be a good paid-member post? You brainstorm on some clinical situations that don't present as Covid and how your standard of care may be altered if you learned they had a positive test, because it's a topic I'm definitely curious on.
And a lot of my theorizing here is based on our experiences out here in Cleveland where the mandatory quarantining of physicians based on contacts, exposure, tests, etc *appeared* to only make things worse in the hospital (lowering staffed beds, cancelling surgeries due to staffing issues) and when all of this went back to "normal" (or BC) capacity issues were alleviated and Covid cases still declined and stayed that way.
Super data, thank you again. I note the highest hospitalization rate in WEST VIRGINIA [100] . I will check "Notes"
Highest Country WV rates on the Norther border in Harrisson & Pendleton counties and in the south in Mercer & McDowell counties.
let alone that WV is likely not known for over-testing....
Well done, thank you. I appreciate any time there's "an adult in the room" providing rational statistical data with caveats appropriately addresses. It's really helpful, at least to me.
So glad it is useful! And Thank you for your kind words :)
I just don't understand not testing patients or employees. Why would hospitals want to know less information about infection?
A few reasons why mass testing may not make sense:
1) Low Cost/Benefit. If a hospital has 1000 employees and has 200 patients pass through a day, that's $12,000/day in rapid tests and $180,000 day in PCR tests.
Considering that non-symptomatic people are extremely unlikely to test positive on either rapid or PCR, it doesn't make sense to spend limited resources (labor, lab capacity, money) without clinical diagnosis. This is how we have always utilized these types of tests in the past - only testing those showing symptoms, whether it is flu, RSV, strep, TB, etc - so what we are seeing is a return back to how it was done pre-2020 for ILI.
2) Inefficient use of laboratory capacity. There is a fixed constraint on # of total types of tests labs can process. Adding on 'n' number of tests almost certain to be negative is at the cost of other tests done based on clinical assessment (HIV, Epstein Barr, Strep, HPV, etc).
3) It often won't change clinical protocols - especially in 2023. If you are not sick but test positive for Covid, that isn't going to change your course of treatment. And again, if you aren't sick, you almost certainly aren't going to test positive for Covid anyway.
4) Both PCR and Antigen tests are notoriously bad detecting infection before symptoms begin, which is why if you aren't sick, you aren't likely to test positive. By the time you are sick, then it can make sense to run a battery of tests to determine if antibiotics, tamiflu, paxlovid, or chicken noodle soup are the best course of treatment. (And PCR tests are also bad at telling when you aren't contagious any longer [1])
5) We tried mass testing and found it didn't work. We stopped mass testing, and paid no price for doing so. To me, the hypothesis that mass testing can reduce spread in communities appears falsified across the world as it failed both conditions (doing it didn't stop things, stopping it didn't make things worse).
I initially thought it would work, buying into the hype from Michael Mina's piece in Time back in November of 2020:
https://time.com/5912705/covid-19-stop-spread-christmas/
Where he claimed: "Countries like Slovakia and the United Kingdom are currently utilizing mass rapid antigen testing programs and already seeing great success."
But when I started following this claim over the next few months, it was falsified.
https://imgur.com/a/heF9yE9
Yet it was never walked back (I later learned he quick academia to join a testing company, so perhaps he had vesting interest not walking back that claim?)
And, over a year later, Michael Mina demonstrated in real time on Twitter why mass testing didn't work - they don't catch infection until your symptoms show up, which means the old conventional wisdom of "stay home if sick, go to work/school if healthy" didn't require $20-$150 to validate:
https://twitter.com/michaelmina_lab/status/1483116982048329734
I think this explains why mass testing didn't work in South Korea, Denmark, Germany, or anywhere else in the world either. (there are other possibilities why it doesn't work as well).
_____________
[1] https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7587118/
(this section a good read: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7587118/)
Interesting, but if there's no testing, then reinstating masks, preferably respirators makes sense. Contracting a Covid-19 nosocomial infection is a serious issue with high morbidity and mortality. As the pandemic continues to roll on more people, including healthcare workers and patients will get reinfected over and over, which potentially decimates healthcare systems and increases the odds for long covid.
Thank you for the great information. What is the reasoning behind no longer testing hospitalized patients upon admission? This does not seem to make sense in light of the many vulnerable patients or accurate statistics? What is now required in nursing homes?
The reasoning? I don't think there IS reasoning for it. It makes zero sense. It's a head-in-the-sand approach.
Nursing homes highly variable. It's a disgrace.
Can I assume your hospital is still testing if a patient is symptomatic for Covid upon admission?
That is what is still happening here at the Cleveland Clinic, UH, and some hospitals we have been to in Pittsburgh - if patient is admitted with "Influenza like illness" symptoms, they get Flu, Covid, sometimes Strep tests. Otherwise, no tests ILI for broken arms, cancer, in-patient surgery, well visits, etc. Is this how it is in Boston too?
I don't think it's set in stone. If symptomatic, certainly people check. For trauma, not so much. The question I have is for things that viral illnesses could trigger. Heart attacks, congestive heart failure, diabetic crisis...these things may be somewhat asymptomatic from the "classic COVID" standpoint, but be the tip-over cases which I always talk about. And of course they can be contagious in any case...
Left a lengthy comment above trying to "steelman" why mass testing may not be helpful [1], but it comes down to two things (I think):
1) Does testing catch pre-symptomatic people, the answer appears to be largely "No"
My thought process heavily hangs on that assumption - if a hospital was running 10,000 tests a week and only 1 or 2 came back positive on asymptomatic people, I'd have a hard time seeing ROI especially if:
2) Would a positive pre-symptomatic test change course of care? Around here, that answer is also "No", perhaps different protocols in Boston? And if done with PCR, the assumption may be "they cleared a previous Covid infection and
didn't realize it, this is just leftover DNA"
If you have a patient come in the ER with heart attack symptoms, no other respiratory issues, would you prescribe Paxlovid if a PCR test showed positive?
**This might be a good paid-member post? You brainstorm on some clinical situations that don't present as Covid and how your standard of care may be altered if you learned they had a positive test, because it's a topic I'm definitely curious on.
And a lot of my theorizing here is based on our experiences out here in Cleveland where the mandatory quarantining of physicians based on contacts, exposure, tests, etc *appeared* to only make things worse in the hospital (lowering staffed beds, cancelling surgeries due to staffing issues) and when all of this went back to "normal" (or BC) capacity issues were alleviated and Covid cases still declined and stayed that way.
________________
[1] https://insidemedicine.substack.com/p/data-snapshot-covid-hospitalizations/comment/42083433
Any idea whether masks will be reinstated at hospitals?
Why does CDC show 16,700 hospitalizations for the past week?
Oh you know what, you're right. It's a 7 day average we reported. Let me fix that. Sorry!