Symptom Tracker Apps – The Big Data Mistake
On the 24th of March 2020, Kings College London launched the Covid-19 symptom tracker developed by ZOE, a very valuable tool in trying to understand the novel Corona Virus.
Then, a month and a half later, as endorsed by the British Government, the NHS app for tracking Covid-19 was released on the 4th of May.
Confusingly, it was then announced that the Scottish Government will be endorsing the original Covid-19 symptom tracker app.
Both are still fantastic examples of data being used to saves lives, but it raises the question which app should we use? Why are there two apps? And why is the Government seemingly supporting both?
Having multiple apps is not only confusing but it raises some fundamental data challenges which could have easily been avoided.
The whole point of each app is to gather data in order to learn about the virus and ultimately keep us safe.
Using this ‘big data’, insights can be gathered, which will in turn, enable us to tackle this virus and any future viruses more effectively. Big data should be exactly that… Big. The more data that can be gathered the better, and the faster we will be able to gather valuable insights.
By gathering data in separate apps, we are limiting ourselves to the amount of data available for analysis, weakening our ability to tackle this virus as quickly and effectively as we could.
This problem is magnified by the fact that as the data will be stored in separate databases, structured very differently, it will be extremely difficult to combine and compare.
Furthering this, if the data is to be combined, what about people who have downloaded both apps? Without a common unique identifier, the database will be riddled with duplicates. Trying to find the same person in each app becomes a nightmare.
In an ideal world, we would only use one app and the data would be consolidated in a single database, allowing us to tackle the Corona virus far more effectively.
The greater challenge however, is one of co-ordination. Fighting Corona virus and reducing the impact of future epidemics/pandemics will undoubtedly require mass global collaboration.
What chance do we have in executing a co-ordinated global effort if we can’t even co-ordinate our own efforts to use one database across the relatively tiny border between Scotland and the rest of the UK?
In Conclusion, if we are to use data to fight the virus then we must be sensible in how we gather and handle this data. Furthermore, we absolutely must collaborate globally and co-ordinate our data response in order to effectively beat the virus.