• Follow us on Twitter @buckeyeplanet and @bp_recruiting, like us on Facebook! Enjoy a post or article, recommend it to others! BP is only as strong as its community, and we only promote by word of mouth, so share away!
  • Consider registering! Fewer and higher quality ads, no emails you don't want, access to all the forums, download game torrents, private messages, polls, Sportsbook, etc. Even if you just want to lurk, there are a lot of good reasons to register!

Coronavirus (COVID-19) is too exciting for adults to discuss (CLOSED)

Jump to latest Follow Reply
Status
Not open for further replies.
Sort by date Sort by votes
Onebuckfan said:
My brother and I were talking about testing he lives in Franklin County. He said the positive tests were 13% . Correct me if I’m wrong but aren’t they testing people with symptoms..people expected to have the virus. Wouldn’t you expect more positives. In our rural county up until lately you had to be hospitalized to get the tests. Reporters have been highlighting people in our county with the fever for days, coughs, some difficulty breathing and never been tested. Are tests unavailable or are they being hoarded. The point being without exponentially more testing how can you be confident about anything reported. The denominator is basically unknown.
Click to expand...

SIAP: That's exactly the issue w/r/t the forecasts. The discussion that @jwinslow and I were having regarding the forecasting accuracy. Without a good denominator, incidence rates which are pretty critical for these kinds of models, you'll have bad projections. With a biased positive rate (which this would be), the forecasts would likely overestimate contraction rates. I have to guess that they are trying to adjust the inputs to account for the bias in some of the key parameters, but I've yet to see any discussion regarding what / how they are doing it.
 
Upvote 0
sparcboxbuck said:
SIAP: That's exactly the issue w/r/t the forecasts. The discussion that @jwinslow and I were having regarding the forecasting accuracy. Without a good denominator, incidence rates which are pretty critical for these kinds of models, you'll have bad projections. With a biased positive rate (which this would be), the forecasts would likely overestimate contraction rates. I have to guess that they are trying to adjust the inputs to account for the bias in some of the key parameters, but I've yet to see any discussion regarding what / how they are doing it.
Click to expand...
Ohio has recently started testing vulnerable zones in prisons & nursing homes, so the cases are skyrocketing. So now we know that there are indeed far more cases than we had on record, but now what do we do to the curve? Do we even still have a usable curve?
 
Upvote 0
jwinslow said:
Ohio has recently started testing vulnerable zones in prisons & nursing homes, so the cases are skyrocketing. So now we know that there are indeed far more cases than we had on record, but now what do we do to the curve? Do we even still have a usable curve?
Click to expand...


So funny you asked about that. I mentioned that a part of my team is working on COVID-19 response stuff for our clients. We were _JUST_ reviewing some maps and forecasts. I asked to see Ohio because it's a state that I've been paying attention to. As soon as we filtered to Ohio the guy says, "WHAOOOOOO... what happened here?" I said, "they just released a bunch of tests from the penitentiary system. Marion and Pickaway counties went from pretty clear to hot spots in a day.

To answer your question, based on those tests alone? Assuming that you weren't being tongue in cheek... Not a chance. The penitentiary system is not a good model for establishing base rates in the population due to the living arrangements. It's _just_ a bit worse than some of the housing projects that are being annihilated in big cities.

Again, not peer reviewed, but a study similar to https://abc7news.com/stanford-coron...by-santa-clara-county-total-covid-19/6110894/ would have to be conducted. If memory serves they used FB for recruiting which which has a set of biases itself. But a national stratifed random sample (think similar how they do political polling) would be a good place to start. Good luck tho. Not exactly like you can test over the phone.

A phone based recruiting with a mail in sample may work if the administration of the test is reasonable. I don't know much about the antibody tests that exist, but if it were something as simple as a pin prick to collect a couple drops of blood... maybe that could work.
 
Upvote 0
sparcboxbuck said:
So funny you asked about that. I mentioned that a part of my team is working on COVID-19 response stuff for our clients. We were _JUST_ reviewing some maps and forecasts. I asked to see Ohio because it's a state that I've been paying attention to. As soon as we filtered to Ohio the guy says, "WHAOOOOOO... what happened here?" I said, "they just released a bunch of tests from the penitentiary system. Marion and Pickaway counties went from pretty clear to hot spots in a day.

To answer your question, based on those tests alone? Assuming that you weren't being tongue in cheek... Not a chance. The penitentiary system is not a good model for establishing base rates in the population due to the living arrangements. It's _just_ a bit worse than some of the housing projects that are being annihilated in big cities.
Click to expand...
Absolutely, but those numbers were part of this the whole time, and now are suddenly being introduced. Like the scores of folks who have had it in the last month and were asymptomatic. The lack of testing hides both extremes, and in so doing misleads us about the middle too.

Or we could mock the threat for two months, then sit on our hands for another 1.5 months before beginning to address the supply problem with tests (it's inexcusable that we just took measures to address the swabs). Which means as we rapidly approach a breaking point for societal & financial instability, we still can't take action adequately if the data suggests we need to open soon (whether due to overselling the threat, due to the cure being more devastating, or whatever new revelations tomorrow brings).
Again, not peer reviewed, but a study similar to https://abc7news.com/stanford-coron...by-santa-clara-county-total-covid-19/6110894/ would have to be conducted. If memory serves they used FB for recruiting which which has a set of biases itself. But a national stratifed random sample (think similar how they do political polling) would be a good place to start. Good luck tho. Not exactly like you can test over the phone.

A phone based recruiting with a mail in sample may work if the administration of the test is reasonable. I don't know much about the antibody tests that exist, but if it were something as simple as a pin prick to collect a couple drops of blood... maybe that could work.
Click to expand...
Good stuff. Which leads to this:

"Our findings suggest that there is somewhere between 50- and 80-fold more infections in our county than what's known by the number of cases than are reported by our department of public health," Dr. Eran Bendavid, the Stanford professor who headed the study, told ABC News' Diane Sawyer.


We can choose to arrogantly die in the dark, or we can embrace the other side of the american coin and research, innovate and lead ourselves out of this pit.
 
Upvote 0
If you’ve tested 100,000 with CoVid symptoms and got 13% positives at a minimum shouldn’t you be testing 100,000 in the general population without symptoms to get some starting point. That 100,000 should be less than 13 but greater than 0. To get a starting point to come up with a testing plan that is not some arbitrary guess.
 
Upvote 0
jwinslow said:
Absolutely, but those numbers were part of this the whole time, and now are suddenly being introduced. Like the scores of folks who have had it in the last month and were asymptomatic. The lack of testing hides both extremes, and in so doing misleads us about the middle too.
Click to expand...

Not sure I agree. This is a prefect example for where two sets of forecasts are appropriate. The single model does not adequately represent the separate populations when combined.

Or we could mock the threat for two months, then sit on our hands for another 1.5 months before beginning to address the supply problem with tests (it's inexcusable that we just took measures to address the swabs). Which means as we rapidly approach a breaking point for societal & financial instability, we still can't take action adequately if the data suggests we need to open soon (whether due to overselling the threat, due to the cure being more devastating, or whatever new revelations tomorrow brings).

Good stuff. Which leads to this:

"Our findings suggest that there is somewhere between 50- and 80-fold more infections in our county than what's known by the number of cases than are reported by our department of public health," Dr. Eran Bendavid, the Stanford professor who headed the study, told ABC News' Diane Sawyer.


We can choose to arrogantly die in the dark, or we can embrace the other side of the american coin and research, innovate and lead ourselves out of this pit.
Click to expand...

I think I favor the 'other side of the American coin.'
 
Upvote 0
Onebuckfan said:
My brother and I were talking about testing he lives in Franklin County. He said the positive tests were 13% . Correct me if I’m wrong but aren’t they testing people with symptoms..people expected to have the virus. Wouldn’t you expect more positives. In our rural county up until lately you had to be hospitalized to get the tests. Reporters have been highlighting people in our county with the fever for days, coughs, some difficulty breathing and never been tested. Are tests unavailable or are they being hoarded. The point being without exponentially more testing how can you be confident about anything reported. The denominator is basically unknown.
Click to expand...
I don’t know what the accuracy of the tests are either...I can tell you if you use one of those horseshit “quick” strep tests, they’re about 50% accurate
 
Upvote 0
heisman said:
Perhaps I don’t understand what you mean?

https://www.worldometers.info/coronavirus/
Click to expand...
What I mean is I’ve seen a lot of argument in this thread about how “comorbidities” (why can’t spell check fucking fix that?) are more responsible than the virus for the deaths to minimize the overall effect of the virus itself...now I am seeing deaths “linked to” the flu, seemingly to make the number look greater by comparison
 
Upvote 0
Provisional Death Counts for Coronavirus Disease (COVID-19)
https://www.cdc.gov/nchs/nvss/vsrr/covid19
The provisional counts for coronavirus disease (COVID-19) deaths are based on a current flow of mortality data in the National Vital Statistics System. National provisional counts include deaths occurring within the 50 states and the District of Columbia that have been received and coded as of the date …

Birx says government is classifying all deaths of patients ...
https://www.foxnews.com/politics/birx-says...
12 days ago · The federal government is classifying the deaths of patients infected with the coronavirus as COVID-19 deaths, regardless of any underlying health issues that …
 
Last edited:
Upvote 0
Bucklion said:
What I mean is I’ve seen a lot of argument in this thread about how “comorbidities” (why can’t spell check fucking fix that?) are more responsible than the virus for the deaths to minimize the overall effect of the virus itself...now I am seeing deaths “linked to” the flu, seemingly to make the number look greater by comparison
Click to expand...
If someone can sit in the middle of, “We’re all going to die!” and “It’s a hoax!”, that’s me, or at least I’m trying to be that person. So, many times, I’m not trying to make a definitive point as I am trying to put something in perspective when someone sits a little to close to either of those edges for my taste.
 
Upvote 0
Status
Not open for further replies.
Back
Top