[BostonCard]

This figure from the New York Times was useful and a bit scary:



Cases in New York city have followed a trend similar to that in Europe, where they peaked a few weeks ago but have since come down.  Outside New York, however, caseloads still look to be going up, which is not a good trend given that a lot of areas are opening up.

The big caveat here is that this is diagnosed cases, and may reflect increased availability of testing.  Still, you would like to see caseloads going down.

BC

[BostonCard]

This figure from the New York Times was useful and a bit scary:



Cases in New York city have followed a trend similar to that in Europe, where they peaked a few weeks ago but have since come down.  Outside New York, however, caseloads still look to be going up, which is not a good trend given that a lot of areas are opening up.

The big caveat here is that this is diagnosed cases, and may reflect increased availability of testing.  Still, you would like to see caseloads going down.

BC

[82 Card]

It would be interesting to see deaths or hospitalizations broken out the same way. That might help figure out the extent to which the increase outside NY area is caused by testing availability.

[82 Card]

It would be interesting to see deaths or hospitalizations broken out the same way. That might help figure out the extent to which the increase outside NY area is caused by testing availability.

[burger]

It would be interesting to see deaths or hospitalizations broken out the same way. That might help figure out the extent to which the increase outside NY area is caused by testing availability.

Ask and ye shall receive.  This is deaths.  The right graph says entire US, but it's excluding NY. The black line is a 7-day moving average. Note the y-axis scales are different.

[burger]

It would be interesting to see deaths or hospitalizations broken out the same way. That might help figure out the extent to which the increase outside NY area is caused by testing availability.

Ask and ye shall receive.  This is deaths.  The right graph says entire US, but it's excluding NY. The black line is a 7-day moving average. Note the y-axis scales are different.

[magnus]

Thanks burger. 

Are these stats based on recorded date or actual date?

IOW, if hospitalized on 1st, but recorded to the department on the 5th, which date gets credit?

If based on actual date,  that would make the optimistic severe dropoff off over the last few days not as rosy.

[magnus]

Thanks burger. 

Are these stats based on recorded date or actual date?

IOW, if hospitalized on 1st, but recorded to the department on the 5th, which date gets credit?

If based on actual date,  that would make the optimistic severe dropoff off over the last few days not as rosy.

[BostonCard]

I used a loess regression instead of a moving average.

BC

[BostonCard]

I used a loess regression instead of a moving average.

BC

[burger]

Thanks burger. 

Are these stats based on recorded date or actual date?

IOW, if hospitalized on 1st, but recorded to the department on the 5th, which date gets credit?

If based on actual date,  that would make the optimistic severe dropoff off over the last few days not as rosy.

I'm using data from here: https://covidtracking.com/about-data

They get their data from state health departments.  I'm not sure how much lag there is between the date of a death and when it shows up in the stats--it's probably different from state to state.  Hopefully using a moving average helps to smooth over those sorts of discrepancies.

I looked at the hospitalization data, too, but it's a mess. It's probably not reliable outside of places that are doing a good job of reporting those numbers.

[burger]

Thanks burger. 

Are these stats based on recorded date or actual date?

IOW, if hospitalized on 1st, but recorded to the department on the 5th, which date gets credit?

If based on actual date,  that would make the optimistic severe dropoff off over the last few days not as rosy.

I'm using data from here: https://covidtracking.com/about-data

They get their data from state health departments.  I'm not sure how much lag there is between the date of a death and when it shows up in the stats--it's probably different from state to state.  Hopefully using a moving average helps to smooth over those sorts of discrepancies.

I looked at the hospitalization data, too, but it's a mess. It's probably not reliable outside of places that are doing a good job of reporting those numbers.

[82 Card]

The NY death data is remarkably smooth from day to day. In the rest of the country, there's much more up and down. The recent rise in the 7 day average for cases doesn't seem to be reflected in a commensurate rise in the 7 day average for deaths. I guess that could indicate that increased testing is either catching higher proportion of actual cases or may is catching them earlier.

[82 Card]

The NY death data is remarkably smooth from day to day. In the rest of the country, there's much more up and down. The recent rise in the 7 day average for cases doesn't seem to be reflected in a commensurate rise in the 7 day average for deaths. I guess that could indicate that increased testing is either catching higher proportion of actual cases or may is catching them earlier.

[Goose]

The new cases per day statistics of the US are the sum many different areas that are in different phases of the outbreak. It may be the ones that are on the upswing balance out the downswing in New York and keep the curve more constant than one would expect. Eventually these lagging areas will go down as well.

However, that said, the Northern California data is not dropping off very fast. We are in a plateau, and have been for a while. San Mateo county 7 day moving average is stuck at in the 25-27 range. I think the other counties are doing somewhat the same thing. Santa Clara Country has said they are tracking none of the cases in their county, so we have no insight into what is going on with the plateau effect there. It is a bit of a worry.

[Goose]

The new cases per day statistics of the US are the sum many different areas that are in different phases of the outbreak. It may be the ones that are on the upswing balance out the downswing in New York and keep the curve more constant than one would expect. Eventually these lagging areas will go down as well.

However, that said, the Northern California data is not dropping off very fast. We are in a plateau, and have been for a while. San Mateo county 7 day moving average is stuck at in the 25-27 range. I think the other counties are doing somewhat the same thing. Santa Clara Country has said they are tracking none of the cases in their county, so we have no insight into what is going on with the plateau effect there. It is a bit of a worry.

[82 Card]

Contra Costa County new cases have been almost a straight line, average 14.5/day for last 16 day. Even before that, the numbers have been fairly steady since March 23. It's not at a high level bet there's no sign of decline.

[82 Card]

Contra Costa County new cases have been almost a straight line, average 14.5/day for last 16 day. Even before that, the numbers have been fairly steady since March 23. It's not at a high level bet there's no sign of decline.