Second wave of the pandemic is at our door and we do not know much about how bad it already could be. We have data coming through various sources but most of it give little about the trends, where we can see if we are doing bad, and if so, how bad it is. We have started to follow the infection trends since March and those graphs gave an accurate picture in March-April but but now in mid-August, numbers have started to make less sense and things are getting fuzzy when we see cumulative graphs which take all the data into account since March. An example is below:
There is not much information above. Actually, only meaningful information there is the number of the deceased. Even 146 infections in last one week is not correct (When you check THL (Finnish Institute for Health and Welfare) web page regarding the infection data, as of today 14.8.2020, summing up all the data for sampling days from “Päiväkohtaiset tautitapaukset (näytteenottopäivän mukaan)” , the number is different depending on where you take the “start date”, but non of the sums for 7 or 8 days make 146.
There are other news platforms with similar data but let’s check THL, the authority for coordinating the efforts for fighting with COVID19. THL publishes two kind of regular updates:
- In their website, updates are published 3 times a week.
- Confirmed coronavirus cases page for detailed statistics, updated every day.
THL posts daily updates to the detailed statistics page. It is also possible to see how many infections have been recorded to their system for each city till that day. Note that, those daily updates do not tell how many people are diagnosed on any particular day per city. It just shows the snapshots of the situation. There could be some positive results which are not yet in their system and those stats will be recorded for a later day. Another interesting thing is that sum of all cities do not make the total sum given at the page. For such reasons, it is not possible to extract exact number of diagnoses in any particular city, any particular day. Considering these, there is somewhat useful information in THL data, but it is not easy to make sense of it. Even if you make sense of it, it is not possible to draw a complete picture.
So what am I after in this post? In the ideal world, best is to know how many people are still sick with COVID19 today, if possible city by city. In the beginning of the pandemic, YLE’s graph was telling exactly that for the entire country, number of sick people with the disease. If the curve is upwards, then the virus is spreading faster among us so we need to slow it down and “flatten” the curve. However, now after five months, many people in the graph is no longer sick and the flat graph is meaningless due to the scale. However, since there is no clear info about who is no longer sick other than the deceased, we do not have an absolute way of showing that information in a graph. So what can we do?
I have been collecting daily snapshots from THL’s confirmed coronavirus cases page for my personal use about the daily status of the infection for quite some time now. I previously did some work to extract graphs for each city, similar to the graph of YLE. In the beginning of the pandemic, cumulative data for each city was useful for the reasons I have mentioned above, especially if I was planning to visit those places. You can never be too careful, right? Then I took it one step further by calculating the moving sum for each date. We cannot know how many people are sick but if we have a number for each day that holds the total number of reported cases in the last few weeks, this could at least tell in which direction the infection is going. So for example, the graph below shows how we did in July. The numbers are not daily reported infections for Helsinki, it is the cumulative number up to each date for last 21 days.
Note that this is not a scientific work. I am just sharing my personal observation about the situation.
To explain with an example, the number on 22nd of July is 22. That is the total number of reported cases in THL website between 2nd and 22nd of July (21 days) for Helsinki. In the beginning of the month, it shows numbers 51 and 53, which tells us that the infection rates dropped during July.
Above graph shows the situation from 28th of July until today. It shows that on 28th of July, we had 20 people diagnosed with #COVID19 in Helsinki during the last 3 weeks. Comparing that to 14th of August, there are 89 people. This shows that confirmed COVID19 cases are raising fast and potentially we have more infected people in Helsinki today, than late June or early July. Until July 13, gatherings more than 50 people were prohibited, there were restrictions to restaurants and bars. There is no restrictions at the moment despite of having more infections in the community.
I will add the graphs of other big cities below. Thanks for reading.
DISCLAIMER: Graphs above does not show the exact number of sick people. They are just a different representations of the snapshots taken from THL website on each date.