School of Social and Political Science

Content

Dr Van Gunten developed the digital tool to visualise the spread of coronavirus in each regional NHS Scotland health board, making this information as accessible as possible to the public. Although not directly linked to his research, Dr Van Gunten is an affiliate of the Edinburgh Q-Step Centre and has interests in quantitative methods, data visualization and information transparency.

Using daily information from the Scottish Government, the tool shows the data on a simple graph. Users can filter by factors such as location and total coronavirus test results, showing what’s happening in their area. The tool allows users to estimate the number of total cases in an area, based on the data for positive test results.

The number of positive test results, Dr Van Gunten said, “under-estimates the number of real cases for several reasons.”

He added:

“I started compiling the data and initially created the tool primarily for my own information, but realised this might also be useful to others, simply making the information more readily available to the public.

“The pandemic is global but infections are local – they happen as a result of contact in particular places. As a result, knowing how quickly the virus is spreading and whether expansion is slowing in local areas is critical information for an informed public.

“The daily figures shared with the public are not shown in a time series, so it’s difficult for anyone reading it to easily understand how fast the virus is growing. I have seen some good examples of data journalism in Spain and the US, presenting the figures from their countries as graphs to easily illustrate this information. I couldn’t find anything similar for Scotland, so I began tracking the data myself.”

You can access the tool here.

Dr Van Gunten shared a couple of his observations since creating the tool, such as the relationship between the volume of testing and the results, and the pattern shown in Shetland. Please note this is an area of interest, rather than an area of expertise.

 

On the relationship between volume of testing and the results he said:

“What strikes me as most important is the total number of tests, including negative tests, and the percentage of all tests that detect the virus. This percentage grew consistently in the second half of March, indicating the more we test, the more cases we find - or possibly that the speed of expansion of testing has not kept up with the growth of the virus.

“Ideally we hope to see that the growth in testing exceeds the growth of the virus; in this case we would see the growth in the percentage of positive tests level off and begin to fall.”

On using the tool to estimate the true number of positive cases:

“The tool also enables users to adjust the number of cases upwards by a recommended factor in order to get a more realistic estimate of how many Covid-19 cases are really out there.

“Statements of the Chief Medical Officer for Scotland suggest that the real number of cases could be 50 or 60 times the number of detected cases. While this is a very rough estimate, and not intended to replace more rigorous scientific methods, hopefully this will give users a better sense of the scale of the disease – closer to 250,000 cases in Scotland, rather than the measured 4229 cases (as of 8 April).”

On regional differences, he added:

“By visualising the growth of the virus within each of Scotland’s NHS Health Boards, you can get a sense of the substantial regional variations. Shetland saw relatively slow growth in the second half of March, which is a promising sign. Fife, on the other side, has unfortunately seen exponential growth. The Borders area also saw rapid growth, though the most recent data show some hopeful signs. These regional differences underscore the importance of attention to local data as the growth of the virus hopefully slows in the coming weeks.”

You can contact Dr Van Gunten at tvangunten@ed.ac.uk and follow him on Twitter, where he’ll be sharing observations he makes from using the tool.