COVID-19 Numerical Observations and Statistics

Dr_SLO

Dr_SLO

Well-known member
This thread has been created to report COVID-19 numerical observations and statistics so they can be discussed separately from the general and specific information threads. Please post all information reporting numerical observations and statistics in this thread and discuss the merits of the data.
 
tuxumino

tuxumino

purrfect
Saw that article yesterday, I'd already been wondering about how the data is presented vs all the things we really don't know. I think the data is often presented by the media as being complete while I think we are years away from complete data on this virus.



GAJ said:
Decent discussion here.

Cheat sheet: How to decipher your county's COVID-19 data

https://www.sfgate.com/coronavirus/...ret-county-COVID-19-data-dashboard-204231.php
Click to expand...

from that article

The raw number of new cases is the metric that often receives the most attention from media outlets — despite the fact that this figure can be misleading. The number of cases is dependent on the number of tests conducted, so any examination of case totals needs proper testing context. There is not much to extract from graphs that only include information on new cases.

later in that article

Hospitalizations are the best "hard metric" one can use when examining the spread of the virus — but there is a long lag time (one to two weeks) between developing symptoms and hospitalization.
The percent-positive rate is a better real-time metric than hospitalizations, but hospitalization figures can be used to confirm upward or downward trends in positivity rates that may indicate that the spread of the virus is accelerating or receding. In Contra Costa County, for example, the percent-positive test rate started to slowly increase from 2.8% on May 29 up to 5% on June 5. The number of hospitalizations first started to noticeably tick upwards on June 5 — exactly one week after the percent positivity rate increased. Percent positivity has since dipped back down to 3.1% over the past week, but hospitalization numbers still have yet to decrease.
 
tuxumino

tuxumino

purrfect
one of the websites I frequent is this

https://www.nytimes.com/interactive/2020/us/states-reopen-map-coronavirus.html

so if you go down to South Carolina and then to Texas they at first glance seem to be increasing at the same rate but if notice the top line for SC is at 1000 while the top line for Texas is at 4000. To me that seems to indicate that Texas is increasing at a faster rate.

another site I go to often is

https://graphics.reuters.com/HEALTH-CORONAVIRUS/USA-TRENDS/dgkvlgkrkpb/

If you scroll down to Weekly reported cases by state you can choose adjusted scale or uniform scale, to me the adjusted one makes it seem that some states are having a greater increase than they may really be having but I'm not a data wonk so I'm unsure which is more accurate.
 
Robert R1

Robert R1

Well-known member
Dr_SLO said:
This thread has been created to report COVID-19 numerical observations and statistics so they can be discussed separately from the general and specific information threads. Please post all information reporting numerical observations and statistics in this thread and discuss the merits of the data.
Click to expand...

Can you please relink that stat site you'd linked before? i lost my bookmark of it somehow.

edit: https://covidtracking.com/

Think i found it again.
 
Last edited:
bosco12

bosco12

Well-known member
The state of California suggests that in Marin county the Spread of COVID-19 is likely increasing rapidly. It's hard to understand from the modeling web site if they are factoring in the San Quentin State Prison population numbers:

https://calcat.covid19.ca.gov/cacovidmodels/

San Quentin State Prison isn't reporting any different affected/recovered employee personnel numbers since yesterday, while news reports over the last few days have shown a rapid rise of incidents within the prison population.

https://www.cdcr.ca.gov/covid19/cdcr-cchcs-covid-19-status/

EDIT: When I reloaded the prison page, the staff incidents has jumped up by 22 from yesterday's number.
 
Last edited:
Archimedes

Archimedes

Fire Watcher
Marcoose said:
NJ Adds Probable Coronavirus Deaths as Total COVID-19 Deaths Near 14,900

1,854 new probable deaths to the tally. Certainly other jurisdictions will revisit their tallies too.
Click to expand...

Of course, anything to add more deaths to the tally. More money for the hospital systems. A material percentage of the deaths already tallied involve either no CV confirming test or have simply been added as excess deaths, assumed to be Covid 19. Why not add some more, call 'em 'probable'. A month from now, we can add 'possible'. Then 'could be'. Then, 'probably not, but we need the money'. Then 'certainly not, but by know nobody's really paying attention any more, so throw them in too.'
 
Marcoose

Marcoose

50-50
Archimedes said:
Of course, anything to add more deaths to the tally. More money for the hospital systems. A material percentage of the deaths already tallied involve either no CV confirming test or have simply been added as excess deaths, assumed to be Covid 19. Why not add some more, call 'em 'probable'. A month from now, we can add 'possible'. Then 'could be'. Then, 'probably not, but we need the money'. Then 'certainly not, but by know nobody's really paying attention any more, so throw them in too.'
Click to expand...

I could've sworn this thread was to report numerical observations, not for the average BARF snide. Go figure.
 
Yakoo752

Yakoo752

USN
tuxumino said:
any of the data wonks play with this site?

https://data.covid.umd.edu/

hard for me to grasp the data and how it's being presented but it's got all kinds parameters and variables you can toggle through.
Click to expand...

I've been in communication with Dr Zhang.

It's interesting stuff but the protests really f-ed their dataset and it's reliability. (My words and thoughts)

The only measure for you to pay attention to is the index (0-100), which includes these weighted values.

The social distancing index is computed from six mobility metrics by this equation: social distancing index = 0.8*[% staying home + 0.01*(100 - %staying home)*(0.1*% reduction of all trips compared to pre-COVID-19 benchmark + 0.2*% reduction of work trips + 0.4*% reduction of non-work trips + 0.3*% reduction of travel distance)] + 0.2*% reduction of out-of-county trips
 
Yakoo752

Yakoo752

USN
Moving to this thread where the data is more appropriate

State level comparative based on hospitalizations, Per capita (100k) Hospitalizations, and hospitalizations compared to beds.

Yakoo752 said:
I am moving my environment to Tableau Public. Easy way to share

This is hospitalizations

This is per capita hospitalizations

This is as a % of total beds

Trended since April 1

The states I am calling out are those that were the worst (bottom of graph) or are trending the worst (top of graph). This will give you a comparative analysis of where we stand today based on what has happened, a scale of severity per se.

I'll be adding more LOD to them through out the day.

If someone wants to build me a web connector, I can set it up as live real time data, other wise I will update them weekly.
Click to expand...
 
Last edited:
tuxumino

tuxumino

purrfect
Yakoo752 said:
Moving to this thread where the data is more appropriate

State level comparative based on hospitalizations, Per capita (100k) Hospitalizations, and hospitalizations compared to beds.
Click to expand...

so looking at this to me it indicates that hospitalizations in CA peaked around April 27th and that's true for the whole country, is that accurate?
 
Yakoo752

Yakoo752

USN
Based on the data, true peak is about the 21st.

We have 100 less hospitalizations now then we did then.

Peak flattened? Seems that way, but we are in an uptick so time will tell.

ah you're talking per capita, yes, for the 27th as peak. (all this does is "normalize" the states to get view on a:a, but test rates and everything else make is not as reliable). I think if you are looking specifically at a state, you don't do a comparative analysis against other states. but it's safe to say 21-27 WAS California's worst week and it's safe to say that this week is on par with it. So theoretically 2 months of social distancing in the shitter.

we're at the same per capita rate today than we were then. and we're increasing so...
 
Last edited:
T

treelogger

Well-known member
I particularly like the modeling these people do: modelingcovid.com. It is quite instructive to see the effect of various quarantine scenarios on counts.
 
bosco12

bosco12

Well-known member
the R metric

http://metrics.covid19-analysis.org/

In this pandemic, R has leapt from the pages of academic journals into regular discussions by politicians and newspapers, framed as a number that will shape everyone’s lives. As Germany’s chancellor, Angela Merkel, explained in a widely viewed video this April, an R above one means an outbreak is growing, and below one means that it is shrinking.

A guide to R — the pandemic’s misunderstood metric
What the reproduction number can and can’t tell us about managing COVID-19.
https://www.nature.com/articles/d41586-020-02009-w
 
tuxumino

tuxumino

purrfect
I've gotten a little way into this article and quite frankly it is way over my head. Maybe you date guys understand it.

This study provides a systematic review and meta-analysis of current RCTs evaluating the use of CHM as an adjuvant to standard care in the treatment of COVID-19.

2.3.1. Bias risk assessment
The Cochrane manual 5.1.0 risk assessment tool was used to quantify the risk of bias.

2.3.2. Heterogeneity assessment
Potential heterogeneity in included RCTs was assessed by using the I2 statistic. When substantial heterogeneity was found (I2 statistic more than 50%), then the sources of such heterogeneity were assessed by rechecking the data or by subgroup analysis based on clinical and methodological variety factors, for instance, participant factors (severity of the condition), outcome measure factors (validation of scoring system) or treatment factors (dose or preparations of interventions), to explain the differences.

https://www.sciencedirect.com/science/article/pii/S2095496420300789
 
You must log in or register to reply here.
Top