The COVID-19 pandemic

[tomh009]

The national stats say that 12,069 people were tested in Canada yesterday. 201 people were positive (note that the timing of testing and test results is a bit different) or 1.69% of the completed tests.

One additional case (20s female) in Waterloo Region today, resulting from a close contact with a previously-known infected person. The total is 16 now.

[tomh009]

A lot of new cases -- 15 -- reported by the Waterloo Region since Friday, or five per day. However, 13 of the 15 are presumptive, so not yet confirmed by positive tests.

Of the 15 cases:

• Five from travel
  
• Two from close contact
  
• Six from community
  
• Two TBD
  

Currently eight of the 30 total cases are hospitalized.

Ontario reported a total of 78 cases this morning. Vast majority are either travel or close contact -- or still TBD.

Nationally 201 positives of the 9618 test results for the 24h until Sunday evening, or 2.09%.

[panamaniac]

That's a startling jump in numbers for the Region, no? Especially the number of community transmissions.

[danbrotherston]

That's a startling jump in numbers for the Region, no?  Especially the number of community transmissions.

This is something that is frustrating me, they changed the methodology for reporting, almost all of the new cases (like what 15 of 16) are "presumptive"...which means they haven't tested positive, but are believed to have the disease, which means we cannot compare this with preivious results where only positive tests are reported.

The government needs to make clear why this change is made (it makes sense, but it should be clarified why now), and how this number can be taken in context of the previous.

Right now, the reporting is useless, and the number is useless, and it has shock value...frankly, it's probably the worst thing I've seen on the COVID stuff locally since Ford's March break fiasco.

[panamaniac]

If I had to guess, the increase in "presumptives" reflects the slowness in getting test results. Virtually all "presumptives" are subsequently confirmed by the second test, afaik.

[tomh009]

That's a startling jump in numbers for the Region, no?  Especially the number of community transmissions.

This is something that is frustrating me, they changed the methodology for reporting, almost all of the new cases (like what 15 of 16) are "presumptive"...which means they haven't tested positive, but are believed to have the disease, which means we cannot compare this with preivious results where only positive tests are reported.

The government needs to make clear why this change is made (it makes sense, but it should be clarified why now), and how this number can be taken in context of the previous.

Right now, the reporting is useless, and the number is useless, and it has shock value...frankly, it's probably the worst thing I've seen on the COVID stuff locally since Ford's March break fiasco.

That's right. There are only two positive tests in the latest batch. I expect some of the 13 presumed positives to be confirmed, but whether it's 1 of 13 or 12 of 13 is anyone's guess.

We can maybe compare this (five total presumed+confirmed cases per day) to the future reports, but, as Dan says, it's not comparable to the previous reports -- or to the Ontario reports, which include only confirmed cases.

[tomh009]

If I had to guess, the increase in "presumptives" reflects the slowness in getting test results.  Virtually all "presumptives" are subsequently confirmed by the second test, afaik.

They might be, or they might not be. As I posted above, only about 2% of the latest 9000 or so tests came back positive.

[panamaniac]

This is something that is frustrating me, they changed the methodology for reporting, almost all of the new cases (like what 15 of 16) are "presumptive"...which means they haven't tested positive, but are believed to have the disease, which means we cannot compare this with preivious results where only positive tests are reported.

The government needs to make clear why this change is made (it makes sense, but it should be clarified why now), and how this number can be taken in context of the previous.

Right now, the reporting is useless, and the number is useless, and it has shock value...frankly, it's probably the worst thing I've seen on the COVID stuff locally since Ford's March break fiasco.

That's right. There are only two positive tests in the latest batch. I expect some of the 13 presumed positives to be confirmed, but whether it's 1 of 13 or 12 of 13 is anyone's guess.

We can maybe compare this (five total presumed+confirmed cases per day) to the future reports, but, as Dan says, it's not comparable to the previous reports -- or to the Ontario reports, which include only confirmed cases.

A "presumed" case means that they have tested positive but have not yet had the results of the second, confirming test, no?

[danbrotherston]

That's right. There are only two positive tests in the latest batch. I expect some of the 13 presumed positives to be confirmed, but whether it's 1 of 13 or 12 of 13 is anyone's guess.

We can maybe compare this (five total presumed+confirmed cases per day) to the future reports, but, as Dan says, it's not comparable to the previous reports -- or to the Ontario reports, which include only confirmed cases.

A "presumed" case means that they have tested positive but have not yet had the results of the second, confirming test, no?

see, this is the kind of confusion this is causing. It is different from the previous policy on reporting cases, yet we have a newspaper reporting ‘confirmed cases skyrocketing’, which is entirely wrong and likely to be panic inducing. 

https://t.co/2FyfIndYp4?amp=1

[jamincan]

This is one of those situations where any statistic you report has to be understood in context. Using the early policy of reporting only confirmed cases would bias results down comparatively from earlier when the testing lab wasn't as far behind. Of tests that have tested positive once and are presumptive positive cases, how many then tested negative after a second test?

[tomh009]

Top 15 countries by 5-day rate of growth of confirmed cases, using yesterday's data (countries with 500+ cases only):

• Turkey: 92% (1236 cases)
  
• Ecuador: 69% (789 cases)
  
• Luxembourg: 42% (798 cases)
  
• USA: 39% (32662 cases)
  
• Brazil: 38% (1593 cases)
  
• Ireland: 32% (906 cases)
  
• Portugal: 29% (1600 cases)
  
• Thailand: 28% (599 cases)
  
• Pakistan: 27% (776 cases)
  
• Israel: 26% (1071 cases)
  
• Chile: 26% (632 cases)
  
• Canada: 25% (1327 cases)
  
• Indonesia: 24% (514 cases)
  
• Saudi Arabia: 24% (511 cases)
  
• United Kingdom: 24% (5683 cases)
  

[tomh009]

This is one of those situations where any statistic you report has to be understood in context. Using the early policy of reporting only confirmed cases would bias results down comparatively from earlier when the testing lab wasn't as far behind. Of tests that have tested positive once and are presumptive positive cases, how many then tested negative after a second test?

This is the question. The first test is less accurate and results in some false positives -- but how many?

The issue I have with the change is that it invalidates the previous data.

[danbrotherston]

This is one of those situations where any statistic you report has to be understood in context. Using the early policy of reporting only confirmed cases would bias results down comparatively from earlier when the testing lab wasn't as far behind. Of tests that have tested positive once and are presumptive positive cases, how many then tested negative after a second test?

This is the question. The first test is less accurate and results in some false positives -- but how many?

The issue I have with the change is that it invalidates the previous data.

It means the context of previous data is lost, the previous data is still valid, just in a different context...

The biggest problem is this change in context was not explained, and as a result, you get newspaper headlines like the ones I posted, which has the potential to increase panic.

[tomh009]

409 new cases in Quebec ... that actually does merit a newspaper headline. Except that it's 409 confirmed and "probable" cases.

[jamincan]

This is the question. The first test is less accurate and results in some false positives -- but how many?

The issue I have with the change is that it invalidates the previous data.

It means the context of previous data is lost, the previous data is still valid, just in a different context...

The biggest problem is this change in context was not explained, and as a result, you get newspaper headlines like the ones I posted, which has the potential to increase panic.

The problem with continuing to report with the same methodology is that the context of that data has changed as well, so you can't interpret the results under the old methodology the same either. In fact, continuing to use that methodology could result in the public understanding that the situation is far less severe than it is. In that situation, then, you have to ask if waiting for confirmation of results from the National Laboratory better reflects the situation, or if using the preliminary results does.