Covid Symptom Tracker was developed at King's college and Guy's and St Thomas's hospitals in London in collaboration with the company Zoe Global Ltd.

The app has recently been launched in Sweden and is already used in the UK and the US.

The user states his birth year, his length and weight and his postal code. Thereafter, questions are answered about health and what symptoms you may have had lately.

"Do not make diagnoses"

- The app has been launched in Sweden as a research project. It is not designed at all to provide individual feedback by making diagnoses or being used for contact tracking, says Paul Franks, professor of medical epidemiology at Lund University.

Paul Frank and his colleague Maria Gomez, professor of physiology, would like to see that the Public Health Agency used the information collected through the app.

"Can show where testing should be intensified"

- If you can see how quickly the virus spreads in different parts of the country, for example, you could see where you could intensify the testing, says Maria Gomez.

If the user indicates symptoms without having been tested for covid-19, how do you know that the information is relevant for research on this particular virus?

- Researchers in the UK have been able to identify a number of symptoms that are typical of corona in people who have not become seriously ill, which can be helpful when analyzing the data, Gomez says.

Is there a risk that the app may give an incorrect image of the coronal state?

- In all models based on probability calculations, there is always a certain risk. But in the UK, for example, flu and covid have been quite different.

Tegnell: "Not sure about the added value"

State epidemiologist Anders Tegnell has previously expressed his skepticism about corona apps, and last week the launch of MSB's new corona tool was halted at the request of the Public Health Authority.

Neither does Covid Symptom Tracker seem to interest Tegnell.

- I don't know what added value it could give us for what the spread looks like. We have good data in place, do our own surveys with samples and large studies with randomly selected sections of the population. It is more representative than a study based on test persons who voluntarily participate, says Tegnell.