Coronavirus Discussion Thread

ffemt8978

Forum Vice-Principal
Community Leader
11,024
1,472
113

SandpitMedic

Crowd pleaser
2,309
1,260
113
Rona is over. Haven’t you heard? We’re moving on in June.
 

FiremanMike

Just a dude
1,129
696
113
Rona is so last week.. It's riot time..

It will be interesting to see if there's a mass 'Rona outbreak with the re-opening of the country combined with the riots..
 

FiremanMike

Just a dude
1,129
696
113
I keep seeing news reports of spikes that are occurring right now.. Nationwide data has been a pretty steady daily increase of ~20,000 +/- ~2000 for the last several weeks with some outliers and no obvious evidence of a spike..

What do you guys make of these reported spikes?
 

Summit

Critical Crazy
2,691
1,312
113
It's regional.
 

mgr22

Forum Deputy Chief
1,656
814
113
So areas seeing a spike are offset by areas seeing a decrease?

I doubt if anyone here could answer that accurately for the U.S. in general. In my state, TN, we're seeing new cases reported at what seems to be an increasing rate.
 

FiremanMike

Just a dude
1,129
696
113
I doubt if anyone here could answer that accurately for the U.S. in general. In my state, TN, we're seeing new cases reported at what seems to be an increasing rate.

Do you think that's due to an increase in testing? I know some states have announced free testing statewide to anyone who wants it..
 

mgr22

Forum Deputy Chief
1,656
814
113
Do you think that's due to an increase in testing? I know some states have announced free testing statewide to anyone who wants it..

Probably some of it, but from where I sit, it seems like we're still dealing with an uncontrolled disease that could get much worse before it gets much better. And as I've said before, I think we need to count long-term disability as a significant consequence of Coronavirus, even though it's too early to have much data on that.

I think you're asking good questions. Maybe someone has better answers than I can offer.
 

DesertMedic66

Forum Troll
11,268
3,450
113
Yesterday was the highest number of new cases in my county since this all started. I’m expecting today to be higher as it has been in a steady climb upwards.

We haven’t opened up any new testing sites and have been testing Monday-Friday 0900-1700 at multiple sites. Some of the local hospitals are only testing if that specific patient is going to be admitted into their facility otherwise no test regardless of their symptoms.
 

Summit

Critical Crazy
2,691
1,312
113
So areas seeing a spike are offset by areas seeing a decrease?
Yes.

Also, this isn't solely or even primarily due to increased testing.

It is blindingly obvious if increased cases is due to increased testing or actual upward inflection of the epidemic curve. If positivity rates change with particular ratio to new cases, then it is a feature of testing revealing true incidence rather than a change in actual incidence. Otherwise changes in new case rates indicates true incidence is changing, as in a spike. Luckily, epidemiologists know all about this, and would not be sounding alarm bells if spikes were merely features of surveillance. Look at AZ hospital capacity. That's not a statistical boo boo.
 

FiremanMike

Just a dude
1,129
696
113
Probably some of it, but from where I sit, it seems like we're still dealing with an uncontrolled disease that could get much worse before it gets much better. And as I've said before, I think we need to count long-term disability as a significant consequence of Coronavirus, even though it's too early to have much data on that.

I think you're asking good questions. Maybe someone has better answers than I can offer.

As a coordinator, I’m a data guy. So many things “feel” like something, but numbers just don’t back it up.. “It feels like we take a ton of mutual aid medic runs” - actually you only take 1ish per shift.. “It feels like” - actually data shows...

This is why I ask so many analytical questions. I come across as an *******, but really I just like to understand. I will say (anecdotally) that in my department, Covid runs have been extremely rare over the last several weeks, my bin of soiled n95 masks has exactly 4 masks to be sent out for cleaning and hasn’t been taken in for 4-5 weeks.. We take about 7000 EMS runs per year, I find that interesting.

I do have some first hand knowledge of the morbidity which is worse than I expected (not me) but unfortunately I don’t think it’s fair to share it due to privacy. I will say that what I’m seeing does tend to confirm the notion that Covid is not just a “recover and done” like the flu..

Yes.

Also, this isn't solely or even primarily due to increased testing.

It is blindingly obvious if increased cases is due to increased testing or actual upward inflection of the epidemic curve. If positivity rates change with particular ratio to new cases, then it is a feature of testing revealing true incidence rather than a change in actual incidence. Otherwise changes in new case rates indicates true incidence is changing, as in a spike. Luckily, epidemiologists know all about this, and would not be sounding alarm bells if spikes were merely features of surveillance. Look at AZ hospital capacity. That's not a statistical boo boo.

I actually don’t understand what you mean in your post, but I’m going to make an effort tomorrow to study how epidemiologists can decipher that data..
 

Summit

Critical Crazy
2,691
1,312
113
I actually don’t understand what you mean in your post, but I’m going to make an effort tomorrow to study how epidemiologists can decipher that data..

Look at:
Sensitivity, particularly how it varies over time and within cohorts (screening/surveillance, asymptomatic, symptomatic sub-acute, hospitalized) being tested (and when these present for testing).
Pretest probability of those being tested with different levels of morbidity.
What groups are tested at what rates and how those rates have varied over time (hint, some groups haven't changed, others have expanded but have naturally low pretest probabilities so you expect low positivity rates).
Test positivity and how it is driven by the above vs testing rates vs disease incidence/prevalence (and know the difference).
Alternate metrics to validate data trends (is it showing what I think its showing?)
 

silver

Forum Asst. Chief
916
125
43
As a coordinator, I’m a data guy. So many things “feel” like something, but numbers just don’t back it up.. “It feels like we take a ton of mutual aid medic runs” - actually you only take 1ish per shift.. “It feels like” - actually data shows...

This is why I ask so many analytical questions. I come across as an *******, but really I just like to understand. I will say (anecdotally) that in my department, Covid runs have been extremely rare over the last several weeks, my bin of soiled n95 masks has exactly 4 masks to be sent out for cleaning and hasn’t been taken in for 4-5 weeks.. We take about 7000 EMS runs per year, I find that interesting.

I do have some first hand knowledge of the morbidity which is worse than I expected (not me) but unfortunately I don’t think it’s fair to share it due to privacy. I will say that what I’m seeing does tend to confirm the notion that Covid is not just a “recover and done” like the flu..



I actually don’t understand what you mean in your post, but I’m going to make an effort tomorrow to study how epidemiologists can decipher that data..

The regional variation in the US is pretty significant and there is a lot of variation in publicly reported data. Just randomly googling looked at AZ DHS. There you can see that over the past few weeks there is an increasing number of PCR tests being performed while also having an increasing % positive rate with increasing utilization of hospital beds and ventilators.

Much of the national data is skewed by NY metro area.
 

Carlos Danger

Forum Deputy Chief
Premium Member
4,510
3,234
113
The regional variation in the US is pretty significant and there is a lot of variation in publicly reported data. Just randomly googling looked at AZ DHS. There you can see that over the past few weeks there is an increasing number of PCR tests being performed while also having an increasing % positive rate with increasing utilization of hospital beds and ventilators.

Much of the national data is skewed by NY metro area.
If you take NYC out of the equation, the US as a whole is one of the least hardest hit of all the nations with the most cases. Ganted this article is like a million years old in 2020 years, but I'd bet the proportions still hold: https://pjmedia.com/news-and-politi...the-united-states-coronavirus-numbers-n385151

As of mid-June, hospitalizations are way up in several states. Some take that as meaning the world is ending AGAIN. The correct interpretation is that any highly communicable virus with any latent period spreads rapidly and that the lockdowns and forced social distancing simply interrupted the natural course of the spread. Also don't forget that if your government incentivizes higher rates of reporting of a thing, that rates of reporting of that thing will increase.

Analyzing all of this is complicated mostly by the fact that our federal and state governments are throughly incompetent, largely by the fact that testing and reporting among the states is very inconsistent, and partly by the daunting logistics of analyzing data from such a massive and heterogeneous nation is ours. The political incentives don't help either, of course.

I follow some very smart and very objective data guys who follow this thing closely and they all say, basically, "the data is so scattered and inconsistent that no one can honestly and objectively say that they have any good understanding of what is happening".
 

silver

Forum Asst. Chief
916
125
43
I follow some very smart and very objective data guys who follow this thing closely and they all say, basically, "the data is so scattered and inconsistent that no one can honestly and objectively say that they have any good understanding of what is happening".

Is it more scattered and inconsistent than other epidemiological data though? I don't do epi, but have read enough to say that the argument that its quite the opposite can be made.

Edit:
That being said, everything comes down to local data at this point as a way of implementing targeted public health interventions. However, that requires competent government.
 
Top