Covid positivity rate

[Hawk_82]

This is the most ridiculous and arbitrary number to track (or base school reopening or closing off of).
There are no standards on who to test and there is no transparency on the people being tested, which basically means the number is worthless. It can easily be manipulated by increasing or decreasing asymptomatic screening among many other variables. Yet our very own governor has decided that this is the number to base schools going online or in person.

I have not seen anyone talk about how stupid this number is, so siap.

 

[H4wkfan4life]

Yep, just overtest, more and more...waste more and more money and we'll just pretend we are doing a great job. Who cares if cases continue to go up.


All numbers are pointless though as we can't believe any of them.

It has been discussed

 

[Pinehawk]

It's better than just using positive case numbers, which is what we have been doing.
The rate at least puts some perspective on if those 20 positive cases were based on 20 tests, or 10,000 tests.

 

[UNIHawkfan]

The worst part, is how Reynolds is calculating the 14 day rolling average. Logically, one would think you would add up all of the positive tests over the last 14 days and divide them by the total number of tests. But nope, we are adding up the 14 daily positivity rates and dividing by 14. This gives way more influence to days that are outliers and gives more weight to tests completed on low testing days. It just doesn't make any sense.

 

[Huey Grey]

The worst part, is how Reynolds is calculating the 14 day rolling average. Logically, one would think you would add up all of the positive tests over the last 14 days and divide them by the total number of tests. But nope, we are adding up the 14 daily positivity rates and dividing by 14. This gives way more influence to days that are outliers and gives more weight to tests completed on low testing days. It just doesn't make any sense.

This is a weird way to calculate it. Her metrics seem really arbitrary. And is it just me, or is the Iowa site where they put the rates god awful? You have to click through numerous pages just to get to the numbers. And am I missing it, or is comparing day to day, week to week, rates information that isn't even on the Iowa site?

 

[Hawk_82]

This is a weird way to calculate it. Her metrics seem really arbitrary. And is it just me, or is the Iowa site where they put the rates god awful? You have to click through numerous pages just to get to the numbers. And am I missing it, or is comparing day to day, week to week, rates information that isn't even on the Iowa site?

I also can't get the numbers from the website on my phone. Which is annoying since everything else works.

 

[Huey Grey]

I also can't get the numbers from the website on my phone. Which is annoying since everything else works.

It seems really simple to me. Click on the website and at the top of the page are the daily numbers. And then if someone wants more details, they can click to other pages.

But right now, you have to know your way around, clicking through mutiple pages, just to find the daily numbers. It's like she is trying to hide the nunbers.

 

[Titanhawk2]

You NEVER hear, who is actually sick. I've heard that 60% of positive tests are asymptomatic. Who gets really sick, how sick to they get? Here in Linn county, active cases have increased 3x since July, but the number of those in the hospital hovers around 10 the whole time.

 

[L. Wade Childress]

This is the most ridiculous and arbitrary number to track (or base school reopening or closing off of).
There are no standards on who to test and there is no transparency on the people being tested, which basically means the number is worthless. It can easily be manipulated by increasing or decreasing asymptomatic screening among many other variables. Yet our very own governor has decided that this is the number to base schools going online or in person.

I have not seen anyone talk about how stupid this number is, so siap.

You are right, it’s illogical. But people are demanding SCIENCE so this metric is simply throwing a bone to the panic/shut down crowd. If it were completely up to her, I’m guessing everyone would be in person because the data say it’s safe.

 

[L. Wade Childress]

You NEVER hear, who is actually sick. I've heard that 60% of positive tests are asymptomatic. Who gets really sick, how sick to they get? Here in Linn county, active cases have increased 3x since July, but the number of those in the hospital hovers around 10 the whole time.

It should not be shocking that once a hospital gets around 75% capacity that magically fewer patients need to stay overnight in the hospital. This was true way before COVID.

 

[Incredible Hawk]

This is the most ridiculous and arbitrary number to track (or base school reopening or closing off of).
There are no standards on who to test and there is no transparency on the people being tested, which basically means the number is worthless. It can easily be manipulated by increasing or decreasing asymptomatic screening among many other variables. Yet our very own governor has decided that this is the number to base schools going online or in person.

I have not seen anyone talk about how stupid this number is, so siap.

I couldn't agree more. I would expect a much higher positivity rate given the recommendations for getting tested - actually it should be closer to 100% based on the recommendation. I don't care how it's calculated it is a foolish number to base opening/closing from.

Also, agree - I'm much more interested in the % of cases that need hospitalization or ICU admittance.

 

[Pinehawk]

Johnson County down to just 4% of positive cases coming from the 0-17 age range. Had previously been 6% I believe.

 

[UNIowaHawk]

Below are just a few issues with the data on hand:

1. Follow-ups-- when someone tests positive, following up to determine if they are clear. If said person chooses not to follow-up (and there are a lot) they perpetually are counted as positive.
2. Initial tests were conducted on site and submitted to labs. This did create dups by onsite and labs submitting numbers for same person. That has since been rectified for the most part. #1 above caused some headache on this as well.
3. Testing positive is Y/N based on markers, but having markers doesn't mean a) you can or ever will get sick, b) can or ever will be able to spread it c) some markers that triggered a positive have since proven false.

 

[Hawk_82]

Below are just a few issues with the data on hand:

1. Follow-ups-- when someone tests positive, following up to determine if they are clear. If said person chooses not to follow-up (and there are a lot) they perpetually are counted as positive.
2. Initial tests were conducted on site and submitted to labs. This did create dups by onsite and labs submitting numbers for same person. That has since been rectified for the most part. #1 above caused some headache on this as well.
3. Testing positive is Y/N based on markers, but having markers doesn't mean a) you can or ever will get sick, b) can or ever will be able to spread it c) some markers that triggered a positive have since proven false.

Does the state count you if you are tested multiple times? Eg. I know of one patient who tested positive on july 24th, he then had a procedure in August so had to be tested again. He tested positive in August but was deemed non infectious since it was 30 days since first positive. He was seen in office without isolation.

Does this patient count as a new positive?

I know there are many situations similar to this. There are also many people who test negative multiple times in a row.

 

[Pinehawk]

Yes, they do count as another case.

 

[Ray Kinsella]

I agree, since it's never clear who's being tested. If it was an entirely random sampling of the population, it would hold more value to me.

What I prefer to look at is the number of those hospitalized with the virus, for two reasons;

1) you're likely to go to the hospital if you need to, regardless of the positivity rate or the numbers tested. If 100 were hospitalized 2 weeks ago, and 80 hospitalized today, then we're improving, regardless of whether we tested 10,000 two weeks ago and 3,000 this week, or vice versa.

2) it's counting those severely affected by the virus. If someone tests positive and it turns out to be the equivalent of a flu-like experience or even less, then it only really matters whether that person has spread it to others. Otherwise, I don't care if X-number tested positive if all of them only experienced a cough and headache for a few days.

 

[hawkeyetraveler]

The worst part, is how Reynolds is calculating the 14 day rolling average. Logically, one would think you would add up all of the positive tests over the last 14 days and divide them by the total number of tests. But nope, we are adding up the 14 daily positivity rates and dividing by 14. This gives way more influence to days that are outliers and gives more weight to tests completed on low testing days. It just doesn't make any sense.

Wait what? That is mathematically incorrect because the denominators will be different by day. They can’t be doing something so stupid can they?

Imagine you have a day with very few tests and a large % of them positive. Then another day with a lot of tests and a small % of them positive. If you simply average the days together you have the possibility to really skew results and make it seem worse than it is (Or in the opposite situation make it seem better). What you know for sure is that the rate they are showing is not a 14 day average...it’s an average of 14 individual days, which is meaningless and exactly opposite of the reason you should consider multi day averages in the first place.

This is a pre-algebra / 6th grade math issue. If they are doing this then someone should be fired for utter incompetence.

 

[Hawk_82]

Yes, they do count as another case.

Wow, that's messed up. I know of 1 patient who was tested 10 + times. He tested positive on every one. The reason was that he was poor and had a language barrier.

She went to the er when she needed treatment, so by protocol had to be tested every time.

 

[Pinehawk]

Wow, that's messed up. I know of 1 patient who was tested 10 + times. He tested positive on every one. The reason was that he was poor and had a language barrier.

She went to the er when she needed treatment, so by protocol had to be tested every time.

Well, situations like that are really pushing our overall positive cases up, which people use to justify business closures and no in person school.

 

[gusto79]

Well, situations like that are really pushing our overall positive cases up, which people use to justify business closures and no in person school.

Are you sure? I just did some quick googling and it looks like positive cases would be unique.

VERIFY: DMV COVID numbers do not count multiple cases for the same person

A viewer asked the Verify team whether two positive tests for the same person would be counted as two cases.

www.wusa9.com

 

[UNIowaHawk]

Does the state count you if you are tested multiple times? Eg. I know of one patient who tested positive on july 24th, he then had a procedure in August so had to be tested again. He tested positive in August but was deemed non infectious since it was 30 days since first positive. He was seen in office without isolation.

Does this patient count as a new positive?

I know there are many situations similar to this. There are also many people who test negative multiple times in a row.

You should only count as 1 positive at any given time. If he tested positive on July 24th he would count as 1+. If he tested positive again in August his designation should still be 1+, not 2+. He would have never tested negative so he would have been a positive carrier that whole time per CDC. The CDC has yet to remove 1+ cases unless they have tested negative or died so they have perpetually been listed as 1+-- 1 week, 1 month, 6 months hasn't matter.

 

[shank hawk]

Whoopsie

•A county in North Carolina incorrectly told nearly 7,000 residents they had tested positive for the coronavirus.

Dr. Cohen praises North Carolina as COVID-19 'standout' among states in the region, citing FEMA report

Here are the latest updates about COVID-19, the disease caused by the novel coronavirus, in North Carolina.

abc11.com

 

[JJ13579]

Johnson County down to just 4% of positive cases coming from the 0-17 age range. Had previously been 6% I believe.

Is this supposed to be meaningful? It's expected given the college kids are the cases

 

[UNIHawkfan]

Wait what? That is mathematically incorrect because the denominators will be different by day. They can’t be doing something so stupid can they?

Imagine you have a day with very few tests and a large % of them positive. Then another day with a lot of tests and a small % of them positive. If you simply average the days together you have the possibility to really skew results and make it seem worse than it is (Or in the opposite situation make it seem better). What you know for sure is that the rate they are showing is not a 14 day average...it’s an average of 14 individual days, which is meaningless and exactly opposite of the reason you should consider multi day averages in the first place.

This is a pre-algebra / 6th grade math issue. If they are doing this then someone should be fired for utter incompetence.

Yes, this is how they are doing it. If they had been actually averaging cases over 14 days polk and dallas county would have both been well above 15% to start the school year. The devil is in the details........

 

[JJ13579]

Below are just a few issues with the data on hand:

1. Follow-ups-- when someone tests positive, following up to determine if they are clear. If said person chooses not to follow-up (and there are a lot) they perpetually are counted as positive.

Not true in Iowa. Patients are automatically considered 'recovered' after 28 days.

 

[BondHawk]

Awaiting tonight's decision from the ICCSD on whether to remain fully remote despite even the flawed positivity metric being down to 11.4% in Johnson County as of this morning. I assume our panicked school board will still be flailing their hands in the air holding the waiver they were granted a week ago when the rate was over 20% due to 5 outlier days when UI students came back to campus. This is all a bit frustrating as the other districts within walking distance of ours seem to be able to miraculously make it through without mass casualties, but maybe cooler heads will prevail this evening.

And for the record, I wear my mask everywhere and wish Trump/Reynolds/Ernst were camping together in a California forest right now.

 

[Pinehawk]

Is this supposed to be meaningful? It's expected given the college kids are the cases

Yes, it goes to the point that according to the data, kids are at very low risk and should be allowed to attend school in person, if they and their family choose that option.

 

[JJ13579]

Yes, it goes to the point that according to the data, kids are at very low risk and should be allowed to attend school in person, if they and their family choose that option.

You really think that's what the change in data showed? In Johnson County? The % of elderly cases went down, too. Are they now safer? ALL ages except 18-40 went down because the college kids were are an outsized share of cases over the past month.

 

[Pinehawk]

You really think that's what the change in data showed? In Johnson County? The % of elderly cases went down, too. Are they now safer? ALL ages except 18-40 went down because the college kids were are an outsized share of cases over the past month.

4%. 6%. I agree, but the point remains. It's not elementary and high school aged kids who are catching the virus.
Kids have been in school, in person, in a number of school districts in Johnson County for 3 weeks now. We are not seeing the cases grow in that demographic.

 

[shank hawk]

Not true in Iowa. Patients are automatically considered 'recovered' after 28 days.

28 days is such an arbitrary, bullshit number. The overwhelming majority of these cases are non-infectious and asymptomatic. The governess/IDPH should reduce that to 7 days if they want to be realistic.