It's worth mentioning again that the deaths being reported every day are reporting deaths from up to 2 months ago. In the past, I have posted the Houston Health death reporting every day and in the last week there were several deaths reported that happened in April and early May. I will save you guys the wall of text of showing it again, but just keep in mind that there's an lag not just between infection and death but also between death and reporting death.
Daily Charts
609,452 Views |
2786 Replies |
Last: 2 yr ago by AggieUSMC
I wish they would go back and modify the historical count per day - ie if today they determine a May 1st death was due to Covid then add a death to the May 1st count (and overall of course), not July 1st.HotardAg07 said:
It's worth mentioning again that the deaths being reported every day are reporting deaths from up to 2 months ago. In the past, I have posted the Houston Health death reporting every day and in the last week there were several deaths reported that happened in April and early May. I will save you guys the wall of text of showing it again, but just keep in mind that there's an lag not just between infection and death but also between death and reporting death.
NY does this. I am guessing someone in Texas does. But when you look at the NY data over time it just becomes clear that the picture for what is happening today won't be clearly known until some time (2-4 weeks?) in the future.
HotardAg07 said:
It's worth mentioning again that the deaths being reported every day are reporting deaths from up to 2 months ago. In the past, I have posted the Houston Health death reporting every day and in the last week there were several deaths reported that happened in April and early May. I will save you guys the wall of text of showing it again, but just keep in mind that there's an lag not just between infection and death but also between death and reporting death.
Median lag is 21 days with a pretty wide range. In general sense when people look at the Houston Health numbers being reported they should just assume that's roughly the number of new deaths that occurred 3 weeks ago.
This is from World O Meter. I can't vouch for them, but they seem to be reasonable.Billy Moose said:
Were there over 50 deaths today? The most recent data I can find is 21 deaths yesterday.
https://www.worldometers.info/coronavirus/usa/texas/
I am one of the people who hate the cases over time graphs, as many of you do, because the % of cases detected over time is going up. I have concluded that the best way to unravel this dillemma was just to make my own damn graph, using model data from covid19-projections.com, who does his best to unravel and show what the past infections were with the knowledge we have today. His data is all open sourced and his projections have graded very highly among those who have graded these things.
In this graph, I compared Western Europe to the USA, because the populations and geographical diversity are more apples to apples comparisons, versus comparing USA to UK only. The stacking is sorted by population, largest populated countries at the bottom.
The primary conclusions should have been already fairly obvious:
Also, the stacked graph visualization is quite interesting to see how the virus transferred from country to country.
I did deaths also, since all the data was right there:
In this graph, I compared Western Europe to the USA, because the populations and geographical diversity are more apples to apples comparisons, versus comparing USA to UK only. The stacking is sorted by population, largest populated countries at the bottom.
The primary conclusions should have been already fairly obvious:
- No, the current infection level is not worse than the US/European peaks, despite what raw case numbers state.
- The recent uptick does differentiate us from those states. However, if we split the US between northern states and southern states, I bet you would find that Northern States shape would mirror Western Europe while Southern States would mirror Latin America/India.
Also, the stacked graph visualization is quite interesting to see how the virus transferred from country to country.
I did deaths also, since all the data was right there:
Quote:
The recent uptick does differentiate us from those states. However, if we split the US between northern states and southern states, I bet you would find that Northern States shape would mirror Western Europe while Southern States would mirror Latin America/India.
Exactly.
Latitude and climate matters.
Yep - and 71% of the population is still 1/3 of the peak of the others.
Edit: The third chart he showed tells the story..
Edit: The third chart he showed tells the story..
Would this include deaths resulting from infections around Memorial Day and early protests or will those pop up later? Just curious. Having been a month it seems like they would start to show up...?
It varies by state, but until mid-May there was a pretty clear one-week lag from reported "case" until reported fatality. Then the trendlines decoupled.
No material on this site is intended to be a substitute for professional medical advice, diagnosis or treatment. See full Medical Disclaimer.
In case anyone is interested.....
I looked back at versions of this chart from the TMC
https://www.tmc.edu/coronavirus-updates/current-covid-19-related-patients-through-tmc-system/
I added the deaths and discharged number to get a total of resolved cases. Then by dividing deaths by
resolved cases you get a percentage of hospitalizations resulting in deaths (resolved). Assuming my
back of the envelope (literally) calculations are correct this is what I got for 4 dates:
4/10 14%
5/19 10.5%
6/16 10.3%
7/1 8.5%
Perhaps this is available in their plots somewhere. I didn't see it.
I looked back at versions of this chart from the TMC
https://www.tmc.edu/coronavirus-updates/current-covid-19-related-patients-through-tmc-system/
I added the deaths and discharged number to get a total of resolved cases. Then by dividing deaths by
resolved cases you get a percentage of hospitalizations resulting in deaths (resolved). Assuming my
back of the envelope (literally) calculations are correct this is what I got for 4 dates:
4/10 14%
5/19 10.5%
6/16 10.3%
7/1 8.5%
Perhaps this is available in their plots somewhere. I didn't see it.
Why would this be?Keegan99 said:
It varies by state, but until mid-May there was a pretty clear one-week lag from reported "case" until reported fatality. Then the trendlines decoupled.
1) Different reporting?
2) Different patient demographics?
3) Weaker virus?
What in the wide world of sports is going on here?
Looks like some states figured out how to flatten the curve, and the others have pretty much been through this and its done.Old Buffalo said:
Yep - and 71% of the population is still 1/3 of the peak of the others.
Edit: The third chart he showed tells the story..
BusterAg said:Why would this be?Keegan99 said:
It varies by state, but until mid-May there was a pretty clear one-week lag from reported "case" until reported fatality. Then the trendlines decoupled.
1) Different reporting?
2) Different patient demographics?
3) Weaker virus?
What in the wide world of sports is going on here?
More testing, social changes leading to more younger people getting it, better treatment, and hopefully a weakened virus but I haven't seen much reliable info about that.
Mainly testing. Early on, you could only get a test if you were sick enough to need hospitalization. Now asymptomatic people are being tested for non essential procedures.
So then, case counts as a time series are meaningless, right?GAC06 said:BusterAg said:Why would this be?Keegan99 said:
It varies by state, but until mid-May there was a pretty clear one-week lag from reported "case" until reported fatality. Then the trendlines decoupled.
1) Different reporting?
2) Different patient demographics?
3) Weaker virus?
What in the wide world of sports is going on here?
More testing, social changes leading to more younger people getting it, better treatment, and hopefully a weakened virus but I haven't seen much reliable info about that.
Mainly testing. Early on, you could only get a test if you were sick enough to need hospitalization. Now asymptomatic people are being tested for non essential procedures.
I get hospitalizations and deaths, but if case counts were artificially low 2 months ago, we should be ignoring that data.
Also, has anyone seen any time series data on cases by age group? I keep hearing its younger people getting it now, but I haven't seen a source.
One thought - the blue line on the left side is in the wrong spot, or rather the curve is wrong. There was widespread transmission going on in February, and reportedly some community spread back to December, but there was no testing. The big initial jump in both lines came almost exclusively from NYC and the tri-state area after supposedly ~6 weeks of spread. The curve shape arguably has more to do with a ramp up in testing than the actual population in the cohort.
Deaths lag disease contraction by a month-ish. Delays in testing results and death notices add a whole additional layer of noise to the data.
Deaths lag disease contraction by a month-ish. Delays in testing results and death notices add a whole additional layer of noise to the data.
I agree. Cases don't mean much in my opinion. Hospitalizations, ICU usage, and deaths matter more
Yes, "case counts" are effectively meaningless, for myriad reasons.
Even hospitalizations are impacted by testing, because almost everyone being admitted without a specifically known cause are being tested today - where they weren't necessarily a month ago.
You've got to remember that a high percent of overall hospitalizations are old people who are sick and struggling to breath. That's true with or without covid.
You've got to remember that a high percent of overall hospitalizations are old people who are sick and struggling to breath. That's true with or without covid.
GAC06 said:BusterAg said:Why would this be?Keegan99 said:
RrDIt varies by state, but until mid-May there was a pretty clear one-week lag from reported "case" until reported fatality. Then the trendlines decoupled.
1) Different reporting?
2) Different patient demographics?
3) Weaker virus?
What in the wide world of sports is going on here?
More testing, social changes leading to more younger people getting it, better treatment, and hopefully a weakened virus but I haven't seen much reliable info about that.
Mainly testing. Early on, you could only get a test if you were sick enough to need hospitalization. Now asymptomatic people are being tested for non essential procedures.
Can confirm on the testing, I had 2 tests three weeks apart for out patient procedures I have been waiting to get done. Multiply this times thousands.
Hospitalizations are also going up for some of the same reasons, but bulk of recent moves appear to be Covid. I do think the surge in hospitalizations related to Covid is that they have capacity and need the revenue. I think it will slow as we get to capacity and they revert back to telling people to treat at home unless oxygen levels or symptoms become more severe.
Hospitals need to be nearly full to survive, as long as they have capacity they will admit patients even if not the most critical. If they think they will have a good chance of getting reimbursed, you'll get admitted. Hospitalizations and deaths will likely also start to de-correlate due to this and the reasons mentioned above.
No one is getting admitted because the hospital needs revenue. Not a single ER doc cares about hospital revenue. We are admitting patients because they require admission.
No material on this site is intended to be a substitute for professional medical advice, diagnosis or treatment. See full Medical Disclaimer.
I live in San Antonio. It's all the local news can talk about. 1 new death yesterday. We've had barely over 100 deaths in a city of 1.8 million, and over half of them were at nursing homes early on.Keegan99 said:
Yes, "case counts" are effectively meaningless, for myriad reasons.
I believe that. What's the age category you are seeing right now in hospitalizations? How are we holding up in terms of beds available? DO you think people are less sick now than they were, or are we just better at treating it now?Dr. Not Yet Dr. Ag said:
No one is getting admitted because the hospital needs revenue. Not a single ER doc cares about hospital revenue. We are admitting patients because they require admission.
By the way, thanks for all you do in our community. We are spending a lot of time going back and forth on public policy, but what you do every day is crucial, and I for one appreciate it very much. Maybe you can write a book about it after this is all done.
I appreciate that, bud. Obviously age of hospitalization skew older. I have admitted a handful of patients under 40. I have admitted a couple 20 year olds. Admitted a 31 year old yesterday, all of whom have required oxygen. But the majority are 50 and older. I did see a 76 year old yesterday who I actually discharged who had suspected COVID because her vital signs were pristine other than a fever, hopefully I don't have to see her back in a few days.agforlife97 said:I believe that. What's the age category you are seeing right now in hospitalizations? How are we holding up in terms of beds available? DO you think people are less sick now than they were, or are we just better at treating it now?Dr. Not Yet Dr. Ag said:
No one is getting admitted because the hospital needs revenue. Not a single ER doc cares about hospital revenue. We are admitting patients because they require admission.
By the way, thanks for all you do in our community. We are spending a lot of time going back and forth on public policy, but what you do every day is crucial, and I for one appreciate it very much. Maybe you can write a book about it after this is all done.
I don't think people are less sick now, but I do think they are showing up sooner in the disease process and the average patient showing up is younger than what I was seeing at the beginning of this in March.
EDIT: As far as beds, we are struggling a bit. Lots of bed holds in the ER. Several patients waiting a couple days for beds. We do still have a couple ICU beds available though.
No material on this site is intended to be a substitute for professional medical advice, diagnosis or treatment. See full Medical Disclaimer.
AgreedDr. Not Yet Dr. Ag said:
No one is getting admitted because the hospital needs revenue. Not a single ER doc cares about hospital revenue. We are admitting patients because they require admission.
I'm jealous of people that have never heard of Interqual or Utilization Review.
Let me put it this way: No insurance company, including medicare or medicaid, is going to pay a hospital for admitting someone that could be treated safely at home. It's hard to enough to get them to pay for people that legitimately need to be in the hospital.
No material on this site is intended to be a substitute for professional medical advice, diagnosis or treatment. See full Medical Disclaimer.
I agree.Keegan99 said:
Yes, "case counts" are effectively meaningless, for myriad reasons.
Then why is this the lead story on every news show I watch, local or national? I'm serious. Is it just to scare people? Because that doesn't sound very responsible.
I wish we had a "newsroom insider" on Texags, like we do for almost every other occupation.
Yes, if you've ever had a loved one in the hospital, or wife had a baby, you never think, "Wow, why are they keeping us here so long?" It's usually, "Are you sure we're OK to go home, shouldn't we stay another day?" And it's all because the insurance company wants you to hit the road.ramblin_ag02 said:AgreedDr. Not Yet Dr. Ag said:
No one is getting admitted because the hospital needs revenue. Not a single ER doc cares about hospital revenue. We are admitting patients because they require admission.
I'm jealous of people that have never heard of Interqual or Utilization Review.
Let me put it this way: No insurance company, including medicare or medicaid, is going to pay a hospital for admitting someone that could be treated safely at home. It's hard to enough to get them to pay for people that legitimately need to be in the hospital.
I thought most people want to get out of the hospital as soon as possible.Aust Ag said:Yes, if you've ever had a loved one in the hospital, or wife had a baby, you never think, "Wow, why are they keeping us here so long?" It's usually, "Are you sure we're OK to go home, shouldn't we stay another day?" And it's all because the insurance company wants you to hit the road.ramblin_ag02 said:AgreedDr. Not Yet Dr. Ag said:
No one is getting admitted because the hospital needs revenue. Not a single ER doc cares about hospital revenue. We are admitting patients because they require admission.
I'm jealous of people that have never heard of Interqual or Utilization Review.
Let me put it this way: No insurance company, including medicare or medicaid, is going to pay a hospital for admitting someone that could be treated safely at home. It's hard to enough to get them to pay for people that legitimately need to be in the hospital.
As I understand it, if you go into the hospital for a broken arm, they are going to test you for COVID. If you have COVID, now you have to go to the COVID wing / floor to separate you from the rest of the patient population.
After they put a cast on your arm, if you are not experiencing any major symptoms, you get to go home. It's unclear that each of these types of cases get counted as COVID hospitalizations, but I would be shocked if most of them are not reported as such. Much easier just to count the number of butts in COVID beds and report that number. So, hospital bed usage probably isn't all that much help either as a time series. I'm not sure an uptick in COVID hospital bed usage tells us much.
ICU COVID, ventilator usage, and COVID deaths are probably the only good things to pay attention to as far as time series go.
After they put a cast on your arm, if you are not experiencing any major symptoms, you get to go home. It's unclear that each of these types of cases get counted as COVID hospitalizations, but I would be shocked if most of them are not reported as such. Much easier just to count the number of butts in COVID beds and report that number. So, hospital bed usage probably isn't all that much help either as a time series. I'm not sure an uptick in COVID hospital bed usage tells us much.
ICU COVID, ventilator usage, and COVID deaths are probably the only good things to pay attention to as far as time series go.
Aust Ag said:I agree.Keegan99 said:
Yes, "case counts" are effectively meaningless, for myriad reasons.
Then why is this the lead story on every news show I watch, local or national? I'm serious. Is it just to scare people? Because that doesn't sound very responsible.
I wish we had a "newsroom insider" on Texags, like we do for almost every other occupation.
I hesitate to assign such motives, but I'm hard-pressed to come up with one better than "it draws in viewers and readers".
Well, I guess I knew that....why do they run anything, it's for ratings. I mean, the actual news that you care about is running on the scroll below what you're having to watch. I just wish it wasn't so sensationalistic, the way it's presented.
And I do wish it was more even-handed, showing the flat death numbers...people need to hear some postitivity in this thing.
And I do wish it was more even-handed, showing the flat death numbers...people need to hear some postitivity in this thing.
A hospitalization is defined as a purposeful two midnight stay. So you have to "admit" someone, and then they have to stay 2 nights. You can put someone in for a few hours or overnight, and that doesn't count as a hospitalization. You also have to meet a complicated set of criteria to be considered inpatient even then. Even my tiny hospital has half a dozen people whose job is to make sure we meet these criteria and get insurance to pay us for the hospitalization. Patient admission are tracked to this minute, and every hospital in America knows exactly who is in the facility and for how long. When it comes to tracking patient stays, easy went away a long time ago.BusterAg said:
As I understand it, if you go into the hospital for a broken arm, they are going to test you for COVID. If you have COVID, now you have to go to the COVID wing / floor to separate you from the rest of the patient population.
After they put a cast on your arm, if you are not experiencing any major symptoms, you get to go home. It's unclear that each of these types of cases get counted as COVID hospitalizations, but I would be shocked if most of them are not reported as such. Much easier just to count the number of butts in COVID beds and report that number. So, hospital bed usage probably isn't all that much help either as a time series. I'm not sure an uptick in COVID hospital bed usage tells us much.
ICU COVID, ventilator usage, and COVID deaths are probably the only good things to pay attention to as far as time series go.
No material on this site is intended to be a substitute for professional medical advice, diagnosis or treatment. See full Medical Disclaimer.
Featured Stories
See All
A&M at risk of undercut when undertaking trip to underrated Auburn
by Olin Buchanan
48:02
6h ago
2.4k
5:59
3h ago
862
16:53
7h ago
2.4k