Definition of a "case" ???

[Boat]

What is the definition of a "case"?

The media are reporting any positive test result as a "case".

Other sources of information define a "case" as a person with a positive test and symptoms.

Why am I seeing conflicting information?

 

[Hippocampus]

There are multiple failures of the lay literature (news outlets, social media etc.) in reporting Covid. They don’t give you the definition of the words they use.

Case fatalities. People with molecular confirmation of the infection (ideally PCR testing), clinical course consistent with the illness and ideally autopsy confirmation of the diagnosis. That’s the gold standard. Case fatality rate . Number of people with confirmed diagnosis who die per number of people. Can include setting so get case facility rate per number of infected or per population of a given geographic area or pre hospitalized patients or per ICU admissions with disease and so forth.
This is distinct from incidence death rate. That uses anyone with a molecular diagnosis who dies per unit time regardless of proximate cause of death. That does capture a guy how was hit by a truck but also the guy driving the truck who was foggy because of mild covid but sought no medical attention so has no molecular confirmation.. It is contaminated by by the guy who died of cancer and caught covid while getting hospice care.
Excess or attributable death. Here you compare a historical death rate with a comparable current death rate. Example. In October of 2019 within the confines of Dallas Texas the death rate was X. In October of 2020 within the confines of Dallas Texas the death rate was
Y. Y minus X is excess death. This is a more accurate assessment of social impact as it includes deaths from food insecurity, poor housing due to unemployment, deaths where testing wasn’t done as well as deaths from deferred care.
Also frequently not mentioned in the lay literature when reporting incidence and prevalence as well as deaths is the nature of the molecular diagnosis.
Need to go through what each type of test means and it’s issues.
Antibody tests. They show you have been exposed to the virus and have been alive long enough to mount a response. They will remain positive after you have cleared the virus and are no longer infectious.
Antigen tests. They require you have been infected long enough that the virus has had the opportunity to replicate enough to be detected on its own. You need a lot of virus for this test to be positive. So it has difficulties with sensitivity.
PCR test. This is the gold standard. Polymerase chain reaction allows you to take a small amount of virus material in your sample and amplify it sufficiently so that with high reliability you know if virus is present or not in the host’s sample.
Physicians judge testing by sensitivity ( if it’s present how likely is it I’ll get a positive test) and specificity (if I get a positive test how likely is it that the person really has the pathogen). PCR has very high sensitivity and specificity. Antigen testing has good specificity but poor sensitivity. Antibody testing is all over the place some good. Some not so much. But antibody testing doesn’t tell you if they still are infectious or even if they still have the disease.
For the lay literature what comes out of Johns Hopkins and Chang school of Columbia U. is good stuff and reliable. They do define their terms. Overall much of the fretting about testing is nonsense. If you look at the crude numbers or even the faulted reporting is quite evident prevalence, incidence and death rates are climbing exponentially. Numbers of false negatives and positives are very unlikely to explain this. Increased testing as the cause is pure utter nonsense. The more testing you do the more likely you will capture positives. But also the more likely you will capture negatives. The more testing you do the more accurate your assessment will be period.

 

[Woodland Hills]

If you have a positive test result on a quality test you have Covid. It does not matter if you are showing any symptoms, you can still have the disease and you can still infect others for a period of time.

 

[boathealer]

For the lay literature what comes out of Johns Hopkins and Chang school of Columbia U. is good stuff and reliable.

They can only collate and report data that is given to them by localities. Inaccurate reported data = bad overall statistics.

Increased testing as the cause is pure utter nonsense. The more testing you do the more likely you will capture positives. But also the more likely you will capture negatives. The more testing you do the more accurate your assessment will be period.

This is only true if you report those negatives right alongside the new positives. This is not being done. Showing only the increase in positives (as the media is doing) is a misrepresentation, and for this media-pressed case, increased testing *IS* partially the cause as the increased negatives are not indicated as a percent of the positives. As a scientist, you should know this.

 

[kthoennes]

Well this thread went off the rails quickly. Time for a mod to zap this one.

 

[Hippocampus]

BH you should try to understand the statistics
Positive rate is number of positives/number of tests. Number of tests is total number. Positives plus negatives.

Rate that positive tests goes up is bad. Doesn’t matter how many tests you do. Illustration.
Do 1000 tests get 2 positive tests. Later do 10000 tests get 2 positive. Rate is going down . Later do 10000 tests get 2000 positives. Rate is going up. Assuming number of total population remains the same you can infer rates from crude numbers. But you don’t have to as they are being reported. Both rate per 100,000 and per cent of positive tests. For instance every day every town and county in my state posts it’s rate of percent of positive tests. Yesterday it was 1.8 for Plymouth. So if rate of positive tests continues to climb over time it’s bad. Yes you’re right if you’re doing a very small sample that sample may not be representative of the population. But once the number of tests per day goes up to a reasonable number what’s called sampling error declines. We’ve been above the number of tests per day where significant sampling error is significant for quite some time. So rate of positive tests is definitely meaningful.

Agree rates are simpler for people to understand. BTW rates are rapidly climbing throughout the country and it isn’t due to increased testing. We haven’t turned the corner. That’s nonsense.

 

[boathealer]

BH you should try to understand the statistics
Positive rate is number of positives/number of tests. Number of tests is total number. Positives plus negatives.

Rate that positive tests goes up is bad. Doesn’t matter how many tests you do. Illustration.
Do 1000 tests get 2 positive tests. Later do 10000 tests get 2 positive. Rate is going down . Later do 100000 tests get 2000 positives. Rate is going up. Assuming number of total population remains the same you can infer rates. But you don’t have to as they are being reported. Both rate per 100,000 and rate of positive tests. For instance every day every town and county in my state posts it’s rate of percent of positive tests. Yesterday it was 1.8 for Plymouth.

Agree rates are simpler for people to understand. BTW rates are rapidly climbing throughout the country and it don’t due to increased testing.

That's exactly what I was saying. Thanks.

Can you point me to your state's accounting for my reference?

 

[KnotYet]

They can only collate and report data that is given to them by localities. Inaccurate reported data = bad overall statistics.
This is only true if you report those negatives right alongside the new positives. This is not being done. Showing only the increase in positives (as the media is doing) is a misrepresentation, and for this media-pressed case, increased testing *IS* partially the cause as the increased negatives are not indicated as a percent of the positives. As a scientist, you should know this.

That's exactly what I was saying. Thanks.
Can you point me to your state's accounting for my reference?

That's not at all what you were saying.

Just sayin...

 

[boathealer]

That's not at all what you were saying.

Just sayin...

No, sir, that is EXACTLY what I was saying about presenting the statistics.

What you are likely referring to is my calling out alarmist reporting of "New Cases" each day by the left-wing media that have no relation to the ACTUAL statistics that Hippocampus outlined in detail so well.

Done.

 

[Hippocampus]

Mass.gov daily dashboard
https://www.mass.gov/info-details/covid-19-response-reporting

 

[Hippocampus]

It’s definitely alarming how rapidly, severely and widespread the rise has been in recent weeks . It clearly bodes poorly as more of our population spends more time indoors due to weather. What particularly concerns me is even in New England, weather has been benign enough most folks can still recreate and socialize outside but rates are rising.
I’m far from “left wing” in my political beliefs. But the numbers are the numbers and we’re doing poorly as is Europe. At least in Europe in e face of that reality and in spite of covid fatigue virtually all governments are mounting an appropriate response . Left, right and center. Once again American exceptionalism comes in to play. For most other countriesthis has nothing to do with political bent and everything to do with competence.

 

[boathealer]

Thank you, H. That MA web site is very good - well documented and statistically accurate.

What are your opinions on why the mainstream media continue to only report "New Cases (per day)" without any of this qualifying information? I think this is a disservice to a (forcibly) dumbed-down American public.

I understand things are going negative. I understand that masks should be worn, precautions are necessary, etc. But what is being done in the name of election-year politics is shameful.

A letter in the Financial Times, https://www.ft.com/content/c3e30eb3-e2ce-482b-bb7c-71ea599ca3bf, makes my point in the second paragraph.

"Reassurance is now just as necessary as supportive fiscal and economic measures. Without an upsurge in confidence, the people whom governments have to date so successfully alarmed are unlikely to resume something close to normal life. Instead of focusing on bare statistics for new “cases”, hospitalisations and deaths from Covid-19, they should include context so people can better judge the extent to which they are at risk, which for most is very small.

For instance, reports should state the percentage as well as the absolute number of tests that are positive (to show the contribution of wider testing to any increase in “cases”), the average age and comorbidities of those hospitalised and dying, and regularly compare the level of “excess deaths” (and deaths from all respiratory diseases, not just Covid) with those in previous years."

"For most other countriesthis has nothing to do with political bent and everything to do with competence."

Most other countries are not in the middle of an election.

 

[KnotYet]

No, sir, that is EXACTLY what I was saying about presenting the statistics.

What you are likely referring to is my calling out alarmist reporting of "New Cases" each day by the left-wing media that have no relation to the ACTUAL statistics that Hippocampus outlined in detail so well.

Done.

Yes, done removing all doubt! Your bias is more than evident.

The Los Angeles Times which is probably on your "left-wing media' list today:
"The county's adjusted rate of cases per 100,000 residents rose to 8 this week, etc..."
It is difficult to find examples that aren't presented scientifically unless you tune to Fox...

The actual increase in cases is, in fact, alarming.

 

[boathealer]

Yes, done removing all doubt! Your bias is more than evident.

The Los Angeles Times which is probably on your "left-wing media' list today:
"The county's adjusted rate of cases per 100,000 residents rose to 8 this week, etc..."
It is difficult to find examples that aren't presented scientifically unless you tune to Fox...

The actual increase in cases is, in fact, alarming.

I didn't say a thing about "left wing media" - you did. Thank you for confirming what everyone knows. I can find thousands of screen grabs that make my point (including Fox, BTW) - I'll spare the bandwidth here.

You might check over the viewership numbers of FOX vs every other media outlet. Might be eye opening.

So a differing opinion is a derogatory "bias" now. Nice. That's very inclusive and accepting of diversity of thought. The hypocrisy is stunning.

I'll say a prayer for you on Tuesday.

Take care,

 

[SteveK]

BH you should try to understand the statistics
Positive rate is number of positives/number of tests. Number of tests is total number. Positives plus negatives.

Rate that positive tests goes up is bad. Doesn’t matter how many tests you do. Illustration.
Do 1000 tests get 2 positive tests. Later do 10000 tests get 2 positive. Rate is going down . Later do 10000 tests get 2000 positives. Rate is going up. Assuming number of total population remains the same you can infer rates from crude numbers. But you don’t have to as they are being reported. Both rate per 100,000 and per cent of positive tests. For instance every day every town and county in my state posts it’s rate of percent of positive tests. Yesterday it was 1.8 for Plymouth. So if rate of positive tests continues to climb over time it’s bad. Yes you’re right if you’re doing a very small sample that sample may not be representative of the population. But once the number of tests per day goes up to a reasonable number what’s called sampling error declines. We’ve been above the number of tests per day where significant sampling error is significant for quite some time. So rate of positive tests is definitely meaningful.

Agree rates are simpler for people to understand. BTW rates are rapidly climbing throughout the country and it isn’t due to increased testing. We haven’t turned the corner. That’s nonsense.

I agree rate of positive going up is BAD
-Looking at the link you provided. cannot see the rate, can you quote it please and is there a graph showing the trend.
-Is it the 7 day weighted @ 133% which actually uses the lowest rate of 0.8 and which is not within the 7 days. Actually I cannot figure where 133% came from. I keep getting 1.06%.
-I see 5.64% 155,660 of all tested were positive from a total tested of 2,758,047, which is below the WHO estimated 10%
-There were 1,139 of 16,724 positive today or 6.8%

 

[KnotYet]

I didn't say a thing about "left wing media" - you did. Thank you for confirming what everyone knows. I can find thousands of screen grabs that make my point (including Fox, BTW) - I'll spare the bandwidth here.

You might check over the viewership numbers of FOX vs every other media outlet. Might be eye opening.

So a differing opinion is a derogatory "bias" now. Nice. That's very inclusive and accepting of diversity of thought. The hypocrisy is stunning.

I'll say a prayer for you on Tuesday.

Take care,

Apparently too stunned to read or remember your own post.
But not surprising. BTW you're entitled to your own opinion, just not your own facts.

 

[SteveK]

Apparently too stunned to read or remember your own post.
But not surprising. BTW you're entitled to your own opinion, just not your own facts.

you should put that line in your sig

 

[KnotYet]

you should put that line in your sig

I am opposed to political sounding statements in signature lines.
Being an engineer and in a technical field I am most comfortable with facts.

 

[SteveK]

I am opposed to political sounding statements in signature lines.
Being an engineer and in a technical field I am most comfortable with facts.

Never thought it as political, it is an all encompassing statement.
I guess I saw a similarity to my sig.

 

[FF]

It was once said ,,,,Figures don't lie, ,,But Liars figure.


After one is vaccinated , wont he test positive for the disease causing figures to soar?

 

[Hippocampus]

No FF.
Antigen testing test for antigen not antibody. So will be unaffected.
PCR testing tests for molecular evidence of the virus and will be unaffected.
Antibody testing may be affected but would need to examine further before making firm statement about a particular antibody test and what’s induced by a particular vaccine (T-cells, particulars of antibody response etc.)

 

[Ski in NC]

This is the CDC chart I look at periodically.

https://www.cdc.gov/nchs/nvss/vsrr/COVID19/index.htm

Gives weekly deaths for all causes, as well as those flagged as Covid.

Interesting that before the pandemic, say in feb 2020, weekly deaths from all causes were about 58,000 per week, with none flagged as Covid.

Look at the last couple months, weekly deaths are actually lower than the typical 58k, yet the % of expected deaths is posted above 100%. How could that be?

Note that not all death records are in for the last month or so, so that skews the numbers lower. Sept numbers probably mostly in. 55k/week from all causes.

If 58k/week died before the pandemic, and 55k/week die now, brings into doubt how many are actually dying from Covid.

 

[soin2la]

If 58k/week died before the pandemic, and 55k/week die now, brings into doubt how many are actually dying from Covid.

Not until you examine all "other" deaths and how many might be down due to COVID. Travel related deaths would be a good place to start.

 

[Hippocampus]

While in training worked for the chief coroner’s office of Manhattan. There’s a significant lag time in certification of death and a further significant lag time as it proceeds up through the various reporting offices of the public health offices. After schooling worked for the Framingham Heart Study. Published a paper on the accuracy of death certification of stroke in that cohort. Was surprised how difficult it was and the reasons for those difficulties.
With this particular event there was a huge bugaboo a few months back when the current administration switched from an in house system utilizing existing federal and state systems to an outside private agency. Recent months have utilized that agency. Some believe that the intent of the administration’s switch was to decrease reporting. Personally believe the prior reporting system needed up dating but this should have been done in house. Now coders needed to learn a new system during a pandemic. Of course this required training and the accumulation of experience and confounds timely and accurate reporting. The numbers you currently see often change. The revisions occur as further information works it way up the reporting chain. So when looking at attributable or excess deaths it’s good to try to compare apples to apples. First wait long enough that the reporting is all in or at least long enough that revisions or additions are expected to be small. Then try to compare like to like. Some years you may have an event causing a outlying result ( flooding event driving it up or new therapy driving it down as examples). You can compute average rate for let’s say February for the last 5 years to try to get around this. But at least you should compare a proscribed past time to a current time. Let’s say June of 2019 to June of 2020. Comparing disparate months- June to October isn’t that helpful as death rates always vary as you go through the seasons.
There’s no doubt people are dying from Covid. There’s no doubt more people are dying from Covid in the last few weeks than recent past. This is true although medical management has improved so death per case has improved.

 

[SteveK]

Not until you examine all "other" deaths and how many might be down due to COVID. Travel related deaths would be a good place to start.

Ski gave a link, where is yours?

 

[SteveK]

While in training worked for the chief coroner’s office of Manhattan. There’s a significant lag time in certification of death and a further significant lag time as it proceeds up through the various reporting offices of the public health offices. After schooling worked for the Framingham Heart Study. Published a paper on the accuracy of death certification of stroke in that cohort. Was surprised how difficult it was and the reasons for those difficulties.
With this particular event there was a huge bugaboo a few months back when the current administration switched from an in house system utilizing existing federal and state systems to an outside private agency. Recent months have utilized that agency. Some believe that the intent of the administration’s switch was to decrease reporting. Personally believe the prior reporting system needed up dating but this should have been done in house. Now coders needed to learn a new system during a pandemic. Of course this required training and the accumulation of experience and confounds timely and accurate reporting. The numbers you currently see often change. The revisions occur as further information works it way up the reporting chain. So when looking at attributable or excess deaths it’s good to try to compare apples to apples. First wait long enough that the reporting is all in or at least long enough that revisions or additions are expected to be small. Then try to compare like to like. Some years you may have an event causing a outlying result ( flooding event driving it up or new therapy driving it down as examples). You can compute average rate for let’s say February for the last 5 years to try to get around this. But at least you should compare a proscribed past time to a current time. Let’s say June of 2019 to June of 2020. Comparing disparate months- June to October isn’t that helpful as death rates always vary as you go through the seasons.
There’s no doubt people are dying from Covid. There’s no doubt more people are dying from Covid in the last few weeks than recent past. This is true although medical management has improved so death per case has improved.

Are you a speechwriter? Sure reads like political double talk.

 

[Ski in NC]

Hippocampus- I rooted around in the CDC site trying to find the same data from 2019 and prior years, so that could be compared to the 2020 link I posted. I could not find it, but may have been me being computer stupid and impatient.

I understand comparing "apples to apples" as you posted, say comparing sept to sept of various years. Just finding it not that easy to do.

I think it is relevant to the discussion of "cases". There is a lot of talk filtering around about the Covid numbers being inflated. Maybe they are inflated, maybe not. I'm no tin foil hat kinda guy, more a "show me the numbers" guy. Comparing total deaths on a weekly basis avoids the coding errors (or malfeasance). Certainly was a spike in the spring, and another around aug, but if sept 2020 weekly total deaths are near the same as sept 2018, 2019 weekly total deaths, then that is telling us how big a problem this really is.

And good point about living in the Covid era that other types of deaths will be different. Maybe less highway deaths, but maybe more from suicide and deferred med treatment, all sorts of numbers will probably change.

And another angle is that for reasons we don't understand, it may be that some humans are, for biological reasons simply immune to severe illness from this bug. Can get it, can pass it on, but it does not bother them, or not bother them much.

If we knew who is vulnerable and who is not, it would make handling this sooo much easier. We know co-morbidities affect outcome, but it seems to be more going on than just that.

The vulnerable being hit hard with this bug in the spring and summer will also affect the analysis of later death numbers, as bad that sounds it is part of sorting numbers.

Just thinking aloud.

 

[Hippocampus]

Steve have given public addresses rarely at medical/scientific meetings. Even including that a rare speechwriter or maker. Have had teaching appointments so have had occasion to lecture. As stated here multiple times I’m neither a Democratic nor Republican and don’t believe that should matter when discussing public health matters. Just trying to share the little I know with you. If that feels like double talk so be it.
Ski share your frustration. Hindsight is 20/20. This pandemic isn’t over and probably won’t be for 2 to 3 years regardless of when a vaccine comes out. Expect the literature will expand for a long while after the acute interval passes. Important to not get lost in the trees at this point but rather see the forest. All indicators point to a rise in number of cases. Recent articles suggest morbidity is higher than first appreciated. The MGH survey of young adults is of interest in this regards.

 

[SteveK]

Steve have given public addresses rarely at medical/scientific meetings. Even including that a rare speechwriter or maker. Have had teaching appointments so have had occasion to lecture. As stated here multiple times I’m neither a Democratic nor Republican and don’t believe that should matter when discussing public health matters. Just trying to share the little I know with you. If that feels like double talk so be it.
Ski share your frustration. Hindsight is 20/20. This pandemic isn’t over and probably won’t be for 2 to 3 years regardless of when a vaccine comes out. Expect the literature will expand for a long while after the acute interval passes. Important to not get lost in the trees at this point but rather see the forest. All indicators point to a rise in number of cases. Recent articles suggest morbidity is higher than first appreciated. The MGH survey of young adults is of interest in this regards.

HC, reading between the lines I can see you have the smarts. But you have a tendency to ramble on, IMO. I see "current administration" can be said but mention the proposed administration and it is deleted.

The CDC has changed their reporting style at least three time since February when they first had a Covid dedicated web page. They gave up having to revise it with the daily changes
I had earlier asked about the rate, you ignored that. post 15. Talk layman talk.

 

[boathealer]

Important to not get lost in the trees at this point but rather see the forest.

Funny you should say that. That is one of my pet peeves about the bashing of the "current administration" on "handling the pandemic".

Everyone cries "He's not following the science! - He's not following the science!"

Guess what? It's not his job to "follow the science"! He'd be a fool to do that, especially since the scientists have changed there minds three or four times in the last 9 months (yes, as more data came in and more was learned...).

His job is to LISTEN to the scientists and then make decisions that are best FOR THE FOREST ---- the country. NOT the trees. These actions are most likely NEVER what the scientists would like. There's economics, unemployment, suicides, schools, increased drug and alchohol use, Antifa burning down cities, impeachment, and a million other things to worry about - all of which have to be considered.

He DID listen to Birx, Fauci and others, and then he made his decision(s). You (and *I*) may not have liked it, but that's what a leader does. Vote them out if they made the wrong decision, but don't yell that they are not "Following the science".

I don't want an administration that just does everything the scientists say.

Same thing with your "we're not turning the corner" comment above. You are focused on the trees. The administration was likely talking about the forest - economy coming back, vaccines are close, 85% reduction in fatalities, better therapeutics...... if you think about all of these, it makes sense.