There is a lot of confusing information emerging about the corona app that has been introduced in the Netherlands.
It was commissioned months ago, and after a false start, a privacy-safe version was produced. Basically, it works like this: if you go for a corona test, you also indicate if you have the corona app working on your telephone. If you test positive, a coded signal is sent to your phone so that your app knows you are positive at this moment.
The app uses the Bluetooth signal to broadcast this information anonymously. Anybody else using the app will pick up that information via their Bluetooth connection if they are within 10 meters of your phone. The strength of the signal is used by the app to determine the approximate distance. If they remain close to you for more than 15 minutes, then they receive a warning on their telephone. They cannot see who the person was that triggered that warning.
Bluetooth is not completely reliable for this purpose. Clothing can mask the strength of the signal, but conversely the signal can sometimes pass through glass or thin walls. The app judges the distance between the phones, not their owners; if two phones are lying close on a table but the users remain well distanced and even masked, the app might react over cautiously.
Not everyone is able to use the app either. On iPhones, for example, a certain minimum operating system release is needed. My slightly older phone cannot upgrade to that operating system, and I would need to replace my phone for the app to work.
Today came the news that around 10.000 people a day receive a warning via their corona app. Some 3.7 million users have downloaded the app; there is no way of knowing how many are actually using it. The national use of the app was launched only a couple of weeks ago.
The advice is: if you receive a warning on the app, go into quarantine. That is a lot of people every day suddenly withdrawing from their jobs because an app told them to. Until they show any symptoms, they are not eligible for a test, although there are signs the government may change that.
Dry runs and anecdotes have shown that a small percentage of the warnings were not justified; the conclusion is that around 70% of the warnings were correct and another 20% were possibly triggered by someone between 1.5 and 3 meters away. That leaves up to a possible 10% where the threshold for the app was not reached. That could mean that several hundred people every day are told by the app to go into quarantine when they do not need to.
This particular app works on the basis of proximity. Earlier apps in other countries were designed to assist tracing. Where an outbreak occurred in a particular place, users of the app who, retrospectively, had been in that place could be traced and tested. Such an app uses GPS data and not Bluetooth technology. Of course, there are privacy issues associated with the storing of an individual’s GPS movements, but it is a different approach.
Today also I heard the first radio commercial encouraging people to use the Dutch app. Fortunately though, the long awaited app is no longer the focus of the battle against corona. Social distancing remains the core message, and I agree with that.
For myself, I am not going to get the app; I have no intention of replacing my iPhone any time soon in order to do so. I will leave others to decide if the app is reliable enough for them to risk using.