By Anthony Lee, MD
Faculty, Harvard Medical School
As we make our way through this pandemic, many of us are wondering about its scale and what's in store for the future. Objectivity is lost in the swamp of catastrophic thinking, conflicting opinions, and political agendas. To regain our objectivity, we must examine the numbers.
Many different sources supply us with numbers describing past, present, and future trends in various places around the globe. These numbers tend be focused regionally and are often used in conjunction with info relating to Italy as the de facto standard.
This makes sense, given that data should be relevant to local needs and that Italy has been through the worst of it. Hospitals are using this information to predict how this pandemic will affect their areas locally and allocate resources accordingly. Public policy should also be driven by this data.
Like most authors, I have included Italy in the mix, and I've highlighted the effects of its lockdown on March 9th in terms of cases and deaths. I've added Belgium because it locked down 9 days after Italy, and it would be useful to see if their results resemble Italy's.
The UK's situation is an interesting one; initially it just let things happen so that herd immunity would develop, then later mandated social distancing followed by a relative lockdown. Ireland and Northern Ireland were added separately to represent progressively smaller scale versions of the UK only in terms of when they started social distancing (March 23-24) and eventual relative lockdowns (March 26-27).
New York State was included because it carries approximately 36% of all cases in the US and roughly 10% of all cases in the world. Finally, Sweden is noteworthy because it enacted measures very late (March 29). But when it did, the measures were significantly less restrictive than those of other countries.
The charts used below present data as cases vs. time, or deaths vs. time. With this presentation, it appears that the US has the most cases, and that things must be terrible here. In places such as NY State, it is.
But keep this in mind: the population of the entirety of the US is 331 Million, while the population of NY State and NYC is 19.5 Million and 8.6 Million, respectively. While the US has about 422,800 cases in total as of today, NYC has roughly 78,000 cases.
If these numbers were expressed on a per 100,000 population basis, you'd get a clearer picture of specific geographic severity. For example, the US situation is better expressed as 128 cases per 100,000, and NYC's very concerning 903 cases per 100,000.
By comparison, Italy, a known hot spot for the COVID pandemic, has 139,442 cases today with a population of 60.48 Million. That's 230 cases per 100,000 people. Therefore, for charting both the number of cases and deaths with respect to time, the values will be expressed as per a 100,000 population basis.
Due to time differences, US data reports lag behind the UK and European countries. Therefore, US plots are likely to be limited to the day before a report is published. In some cases, late reporting from European countries and the UK may also limit plots to the day before.
Figure 1. Cases per 100,000 population by location and correlated rate of change.
Looking at Figure 1, we can see the effects of lockdowns and social distancing. It took Italy 10 days to finally reduce the number of daily new cases (basically, the rate of change curve has flattened). This trend has continued for the last 20 days, producing the trailing edge of Italy's pandemic.
Within 9 days Belgium reduced the number of daily new cases and continued on a downward trend for the last 12 days. It took Ireland 6 days of lockdown to reduce their daily new cases, and this trend has continued over the last 6 days. Northern Ireland took 7 days of lockdown to reduce their daily new cases, and this trend has continued over the last 5 days.
Despite loose social distancing, Sweden continues on a low and flat trajectory. Perhaps this can be accounted for by lack of testing.
Due to the setting of social distancing policies, the UK remains on a low and flat trajectory.
Despite a shortage of testing NY State and NYC cases are rising dramatically. Testing has increased in Massachusetts over the last 2 weeks, and this is reflected in the figure. While NY has been about 5-8 days behind Italy over the last several weeks, Massachusetts has been roughly 21-23 days behind.
Although both NY and Massachusetts started their social distancing advisories at the same point in time. Relative to Italy's social distancing, Massachusetts had a roughly 2-week head start on NY. This may have caused the relative flattening of new cases in Massachusetts over the last 6 days. This may change, however.
Rates-of-change plots appear in surges or waves. In locations that are improving, each subsequent wave is smaller than the previous one. This forms the trailing edge of the pandemic in a given location. In locations that are worsening, each subsequent wave is larger than the previous one, forming the leading edge of a worsening pandemic.
Figure 2. Deaths per 100,000 population by location and correlating rate of change.
In Figure 2, we can see the rising deaths by location.
In the rates-of-change chart, the number of daily new cases flattened are delineated by date.
From the point of delineation rightward, we can see when daily deaths begin to decrease. This lag time varies with locations. For Italy, this lag time was 8 days. For Ireland, these reversals were basically at the same time. For Northern Ireland, the lag time was 1 day.
For these countries, the time between lockdown/social distancing and reduction of daily deaths was 18 days for Italy, 6 days for Ireland, and 8 days for Northern Ireland.
I will update and discuss these charts in future installments of this series.
Sources of data: Worldometer.com, Spectrum News NY1, Massachusetts Department of Public Health, HSC Public Health Agency