The Curve? Week 2 Update

The Curve? Week 2 Update

For 17 days I have tracked the COVID-19 data from Canada in relation to the data from Italy. Since starting this analysis, a lot has changed in the world, but I continue to compare Canada to Italy because I have found their data is instructive. Last week I shared my work because of the surprisingly tight correlation between the Italian and Canadian data. Click the link below to download the current report.

No alt text provided for this image
No alt text provided for this image

 Around April 1st resembles a turning point in Canada’s COVID-19 data relative to Italy’s (Figure 2 in report). That time-frame was the beginning of a sustained multi-day period where Canada’s trend in new cases per 100K was measurably lower than Italy’s at the same point on the ‘curve’. Personally, this turn is what I was hoping to see as a result of Canada’s aggressive public policy interventions.

I am offering this information with a focus to minimize my personal interpretation. I certainly have opinions, but they are best shared over a coffee. With that said, I would like to draw special attention to Figure 3 in the report which shows this data in the context of the forecast published by Public Health Ontario on April 3rd. I will continue to track this information with confidence the current seismic public policy measures will be informed by practical information and common sense.

No alt text provided for this image

'Methodology'

The basic premise of my analysis is that the rate of change in the number of cases could be compared between countries. I made an initial comparison by finding the approximate intersection point where Canada and Italy’s measurements of new cases per 100,000 population were equal. That intersection point became Day 0 in my comparison.

With the benefit of more actual data, I was able to fine-tune the alignment of Canada and Italy’s ‘curves’ by calculating the Mean Absolute Percentage Error (MAPE) between various values for Day 0. From this subsequent test of the fit between the curves, I moved my Day 0 by one day. I continue to calculate the MAPE between the data sets as more actual data is available.

The result is that at this point, the data suggests the maturity of Italy’s spread of the virus is approximately 19 days ahead of Canada. That provides the means of comparing Canada’s cumulative cases by applying Italy’s total cases per 100K to Canada’s population data.

Beata Sejane

I am a facilitator, delivering administrative service excellence. I thrive in a fast-paced and intellectually challenging environment, and I take pride to work for an organization with an important social purpose.

4 年

Looking forward to your updates!

回复
Beata Sejane

I am a facilitator, delivering administrative service excellence. I thrive in a fast-paced and intellectually challenging environment, and I take pride to work for an organization with an important social purpose.

4 年

Terrific job Dave

回复

The Italian population is noticeably older than Canada's, so it makes sense for them to have more cases and more fatalities. To get a comparable baseline you need to age adjust the numbers.

回复
Paul Desjardins

Founder & CEO @ Move Add Change | Board Director @ Ontario Centre of Innovation

4 年

That is a good read, Dave and I like your model. I’ll be hopeful for that chat over coffee soon.

回复

要查看或添加评论,请登录

David H.的更多文章

社区洞察

其他会员也浏览了