COVID-19 detected in United States

[Jacob]

That data does have a 1 to 8 week delay.

I'm not all that familiar with their data to be honest. I wasn't implying anything sinister, just that their death totals now show 37k, which was adjusted down from around 65k the day before I believe. With how this thing is politicized, some groups will jump all over this.

 

[Evan]

Alabama with 317 new cases today. Highest number of new cases in 3 weeks. Also represents 3rd highest increase in new cases since the pandemic began, and 2nd place was only 4 more cases at 321.

Not good news. Next few days will tell us if this was an aberration or a sign of increasing case growth.

 

[gangstonc]

Alabama with 317 new cases today. Highest number of new cases in 3 weeks. Also represents 3rd highest increase in new cases since the pandemic began, and 2nd place was only 4 more cases at 321.

Not good news. Next few days will tell us if this was an aberration or a sign of increasing case growth.

I expect we will still keep ramping up as our ability to test increases.

 

[Jacob]

Alabama with 317 new cases today. Highest number of new cases in 3 weeks. Also represents 3rd highest increase in new cases since the pandemic began, and 2nd place was only 4 more cases at 321.

Not good news. Next few days will tell us if this was an aberration or a sign of increasing case growth.

Do you have the stats for the % that tested positive vs. other days? Just curious if it is related to any increased testing numbers or if the % of positive tests have increased as well.

 

[gangstonc]

Do you have the stats for the % that tested positive vs. other days? Just curious if it is related to any increased testing numbers or if the % of positive tests have increased as well.

You can figure that out from the data posted at Www.bamatracker.com

 

[Jacob]

You can figure that out from the data posted at Www.bamatracker.com

Not sure how I hadn't run across that site yet, thanks!

Looks like they haven't put up the # of tests for yesterday, just the new cases. Seems I'll have to wait another day or two to get my answer. Edit: After looking at the data a little closer, I don't think the two datasets come from the same place, so unfortunately I don't think I can reliable calculate the %.

 

[Jacob]

Apparently there has now been a confirmed case of community transmission in France from before Christmas.

Logically if it was in the community in France in December, it was likely in the US before Christmas as well (maybe as early as around Thanksgiving?).

*I should note this is early "news" from a French journalist on Twitter", so further confirmation is needed.

 

[gangstonc]

Apparently there has now been a confirmed case of community transmission in France from before Christmas.

Logically if it was in the community in France in December, it was likely in the US before Christmas as well (maybe as early as around Thanksgiving?).

*I should note this is early "news" from a French journalist on Twitter", so further confirmation is needed.

I actually think that would be great news.

 

[Evan]

Do you have the stats for the % that tested positive vs. other days? Just curious if it is related to any increased testing numbers or if the % of positive tests have increased as well.

I don't. Closest thing to that in the dataset I'm using would be the total tested trends by day vs growth in case counts.

Here's the dataset I'm using.

You can put total tested right next to number of cases per day, but obviously the Y-axis is a different scale as they're trying to capture trends. From the analysis I've done, which is crude and not very granular, there's obviously a correlation between total tested and the growth in case counts (as one would expect), but it's not clear how well that correlation maps overall to the growth in cases.

I suspect that our initial peaks around April 8th - April 13th reflect a number of backlogged cases as testing was really just starting to ramp up and scale. I base this conclusion on the fact that testing growth rate was nearly flat between April 8 - April 12th while cases started peaking out. Although not proven, that seems to indicate to me that the case count peaks were more a reflection of a backlog of pending tests being cleared than a growth in testing. Otherwise, you'd have to assume an excessively high positive test rate which I don't think is supported by the data. I presume the % positive test rate was rather high during the time, however, as testing wasn't growing that quickly, and those initially tested were more likely to have the virus as the screening criteria in place (because of testing scarcity) was more stringent.

So, I would tend to agree that part of why we're seeing a new peak in cases is because of recent growth in test capacity. However, if you look at the slope of the trend for both tests completed and % cases, you'll see a fairly consistent slope for case counts whereas tests completed is much less consistent and has plateaus/sharp increases.

My ultimate conclusion is that we weren't testing enough previously and were probably underestimating case counts, and now we're about to see a more realistic view of positive cases because testing capacity has grown enough that the heavy rationing that was in place is no longer necessary.

It also shows why developing ACCURATE antibody tests (obviously this test type isn't currently reflected in the dataset I'm using) that have much better sensitivity and specificity are necessary. Improving the antibody tests is necessary to achieve lower rates of false positives and negatives otherwise they're just not trustworthy enough to use to guide decision-making. Actually, even the PCR tests still have problems with false negatives/positives.

Bottom line: we are not where we need to be yet as it pertains to testing. We're making good progress on overall testing growth and availability, but we also need to simultaneously increase the accuracy of both the PCR and antibody tests to be able to have a clear understanding of how the virus is spreading throughout the community. We'll get there -- but improving accuracy requires time and experience and definitely involves some trial and error as we've seen.

Ultimately, I don't know that we can conclude that the recent peak is an ominous sign or an aberration until we see several more days of data (or even a week). Which is why I mentioned that as a caveat in my original post about the new peak we experienced yesterday. There's a lot of variables involved, so there's a number of things that could explain the increase. Until more data is available, I'm not at all confident any conclusion can be made other than we need to watch the data closely in the coming days to see if case counts continue to rise compared to the previous average per day, and how that correlates with total tests completed.

 

[skelly]

Apparently there has now been a confirmed case of community transmission in France from before Christmas.

Logically if it was in the community in France in December, it was likely in the US before Christmas as well (maybe as early as around Thanksgiving?).

*I should note this is early "news" from a French journalist on Twitter", so further confirmation is needed.

You know Stanford try to put the results of their study out there pointing to it was most likely in California before Christmas and apparently Fauci poo pooed it and the media has kind of ignored and booed the idea that heard immunity may already have been building up in California. I hate this being in a limbo of what not to believe or think....

 

[gangstonc]

You know Stanford try to put the results of their study out there pointing to it was most likely in California before Christmas and apparently Fauci poo pooed it and the media has kind of ignored and booed the idea that heard immunity may already have been building up in California. I hate this being in a limbo of what not to believe or think....

Yes. It’s a fog of war scenario. We are all in a tough position and evidence isn’t readily available or strong.
We desperately need millions more tests of both the active virus and the antibodies so we can figure out where we stand.

 

[Evan]

One other thing to keep in mind...looking at data on the state level isn't exactly the best way to understand how the virus is truly affecting us. A more granular dataset at the county/community level is necessary as we've seen repeatedly that community spread is occurring in clusters in specific hotspots spread throughout the stage.

What's happening in Bibb County may be starkly different from what is occurring in Mobile County. And, the healthcare infrastructure, level of care, testing ability, demographics, population density, cultural attitudes, etc all can vary significantly by county.

I'm not sure how granular the data needs to be, but just looking at the state as a whole isn't going to be the best approach when it comes to making decisions on re-opening. We may be ready to open up some areas right now whereas other areas might need to have STRICTER regulations than what is currently in place.

A one-size fits all approach isn't going to cut it. However, we also have to take into account that people live and work in different counties/communities, and visit/travel to others for various reasons (family, church, work, etc). That makes it tough to take the more granular approach of setting regulations at the county/community level as we have a highly mobile population.

I'm at the point at which I think the restrictions should be drawn out on a map with circular layers of higher to lower restrictions emanating out on a geographic radius from a known hotspot. Think of a bullseye target that would be overlaid on hotspots/clusters or an entire, county, metro area, or other designated population.

For instance, you might draw an inner bullseye out from Jefferson County and decrease restrictions as you approach areas with fewer cases although the restrictions would still be tighter than places where there are no known hotspots or cases. Doesn't have to be a perfect circle. Think of a severe weather outlook issued by the SPC or the gradient lines we see emanating out in an oval like fashion on various models for a given indice.

I know it sounds insanely complicated, but this is a very rough concept not a finished idea. It allows you to be much more targeted and local with restrictions without needlessly locking down the entire state or unimpacted areas well-isolated from hotspots. Obviously an aggressive track, trace, and test approach has to be layered in with any such system.

We need to prevent random parts of somewhere like Blount County being locked down in the same way as Jefferson County. By the same token, it shouldn't be that once you leave the confines of Jefferson your risk is suddenly eliminated. It's a way to locally target risk and taper it off as you approach areas that have been identified as being at a lower risk. And Blount County shouldn't be treated the same way as Shelby County. Both are similarly situated in geographic relation to Jefferson County, but CLEARLY there's a huge difference in population, demographics, pop. density, and transit/travel/commerce flow.

I see something like 4-5 color-coded levels of regulations/restrictions that increase from a base level that has no businesses closed, but recommends social distancing, improved hygiene, increased cleaning, and suggests the usage of masks in public to increasing restrictions/requirements. You start at a level where everything is just a suggestion or recommendation, and increase from there. Level 2 might see language that STRONGLY URGES masks and social-distancing and recommends, but doesn't require, some businesses to close and that retail stores limit in-store capacity.

Level 4 might include mandated business closings and the kind of restrictions we saw seen statewide with Ivey's statewide stay at home order. Level 5 is where you get curfews, mandatory masks in public, essential businesses only, and stringent requirements across the board that we haven't yet seen in Alabama.

If this kind of layered approach was applied in a common sense fashion, and was based on levels of decreasing restriction out from a radius of known hotspots, I think the public would be more accepting of it and feel less like their rights are being trampled on because it would be customized to risk and would help get a lot of people back to work. You've got people in rural areas of Blount County right now wondering why the hell such a big deal is being made about the virus and becoming increasingly resentful and hostile toward efforts to control the pandemic. Likewise, our economy and people's livelihoods are taking a massive hit.

Don't want a lockdown or mandates? Voluntarily engage in social distancing, better hygiene, and be smart. It makes individual areas and communities have a much better feeling of control of their lives and makes them more apt to do what's needed voluntarily. Because no one wants the government to mandate or require things. Here's a way they won't have to if your community takes this seriously and does what it needs to. If you do end up with a cluster or nearby hotspot locally, you can get back to the lower restrictions quickly by doing the things that have been recommended for months and complying with any mandated restrictions.

This is all just an extremely crude rough draft of a back of an envelope thought. I've no idea how it would work in reality, if it would even be feasible, or if there's even an appetite to try something like this. Logically, however, I think it makes sense but I'd love to see people poke holes in it, improve it, explain why it is insane, etc.

 

[Evan]

Speak of the devil...local mayors ask Ivey to ease restrictions and give them more latitude than larger cities like Birmingham.

 

[StormStalker]

Franklin County seems to be an emerging hotspot in Northwest Alabama. That will be something to watch over the next few days. We also have to note that two weeks ago was Easter. Did families go ahead with Easter gatherings and help spread the virus? I know Easter Sunday was nasty weather wise, but the Saturday before was really nice.

 

[Jacob]

Franklin County seems to be an emerging hotspot in Northwest Alabama. That will be something to watch over the next few days. We also have to note that two weeks ago was Easter. Did families go ahead with Easter gatherings and help spread the virus? I know Easter Sunday was nasty weather wise, but the Saturday before was really nice.

Looks to me more like increased testing in the area. There had been ~435 tests run for Franklin county from March 20 through April 28, and in the last 4 days there's been 300 tests processed. Albeit, a higher percentage of those most recent tests have come back positive than the earlier tests.

 

[Jacob]

So the CDC/White House now says that by re-opening things, we'll go from 25k cases a day like we are currently seeing, to 200k cases a day by June 1st.

Consider me skeptical.

 

[Evan]

Looks to me more like increased testing in the area. There had been ~435 tests run for Franklin county from March 20 through April 28, and in the last 4 days there's been 300 tests processed. Albeit, a higher percentage of those most recent tests have come back positive than the earlier tests.

I think we have to remember that more testing also occurs when more people have symptoms and either they or their physician request a test. We don't know if that's what is happening in Franklin or if it simply represents an increase in testing capacity.

That's one of the issues I mentioned previously when I said it is tough to use the current data to make conclusions about trends. There's so many variables that without having access to additional data on those other variables, or waiting to see how other indicators change in the coming days/weeks, it's very difficult to make any firm conclusions. Right now, we'll probably have to watch hospitalizations in Franklin County if we want to see if the increase in tests is because it is becoming a hotspot, or if it is simply due to increased testing. I'm not even sure that data is being released for every county.

Another issue is that we don't know what kind of lag exists between the growth in total tests (or tests per day) and positive cases. The increase in positive cases we saw for May 2nd/3rd compared to the 10 days previous to those days may be because there's substantial lag time from when total test numbers are published and positive lab results come back.

It's extremely frustrating because it seems like there should be more coordination on data releases so we find out not only the positive case counts and total test counts, but on what date those new positive cases were tested. I haven't even seen a metric for current waiting time for lab results recently, and it varies depending upon which lab does the test (whether one of the various private labs, state lab, or a University lab).

I have to think that is being tracked internally by the state and Ivey's team, but if it is, why isn't it being released to all of us so we can see whether cases are really growing or we're simply seeing increases due to more testing availability and capacity (or even less stringent testing guidelines).

 

[StormStalker]

The spike in cases in Franklin County seem to be tied to a poultry plant. I doubt the Hospital is Russellville would be able to accommodate an influx of critical patients. Those folks would probably have to be moved to Helen Keller Hospital or North Alabama Medical Center. Let’s hope it doesn’t come to that.

 

[Evan]

The spike in cases in Franklin County seem to be tied to a poultry plant. I doubt the Hospital is Russellville would be able to accommodate an influx of critical patients. Those folks would probably have to be moved to Helen Keller Hospital or North Alabama Medical Center. Let’s hope it doesn’t come to that.

Makes sense. Major concern would be transfer of the growth in cases to the wider community. With how contagious this virus appears to be, it's something that public health officials have to attack immediately lest it get out of control.

I agree that it's unlikely that Russellville would be able to handle a sudden cluster hotspot and would need assistance from larger facilities. Like you said, let's hope it doesn't come to that.

 

[Jacob]

I think we have to remember that more testing also occurs when more people have symptoms and either they or their physician request a test. We don't know if that's what is happening in Franklin or if it simply represents an increase in testing capacity.

That's one of the issues I mentioned previously when I said it is tough to use the current data to make conclusions about trends. There's so many variables that without having access to additional data on those other variables, or waiting to see how other indicators change in the coming days/weeks, it's very difficult to make any firm conclusions. Right now, we'll probably have to watch hospitalizations in Franklin County if we want to see if the increase in tests is because it is becoming a hotspot, or if it is simply due to increased testing. I'm not even sure that data is being released for every county.

Another issue is that we don't know what kind of lag exists between the growth in total tests (or tests per day) and positive cases. The increase in positive cases we saw for May 2nd/3rd compared to the 10 days previous to those days may be because there's substantial lag time from when total test numbers are published and positive lab results come back.

It's extremely frustrating because it seems like there should be more coordination on data releases so we find out not only the positive case counts and total test counts, but on what date those new positive cases were tested. I haven't even seen a metric for current waiting time for lab results recently, and it varies depending upon which lab does the test (whether one of the various private labs, state lab, or a University lab).

I have to think that is being tracked internally by the state and Ivey's team, but if it is, why isn't it being released to all of us so we can see whether cases are really growing or we're simply seeing increases due to more testing availability and capacity (or even less stringent testing guidelines).

I don't disagree with anything that you say here, the coarse-ness of the data is frustrating.