The entire PRONIA consortium is happy to announce that the first project paper has been just published online in JAMA Psychiatry. After five years the consortium managed to show in a multinational study across 7 European countries that machine learning algorithms perform better than doctors at predicting mental health outcomes, up to 80 per cent or 82 per cent of accuracy. Machine learning can be used to prognosticate social and vocational ability in adolescents and young adults at risk of developing psychiatric illnesses (psychosis or depression). Clinicians could use this tool to obtain decision support before starting treatment in a given patient. The creation of the diagnostic algorithms was so far the most important research outcome of the PRONIA consortium, funded by the European Commission and could have major implications not only for patients, but also for how we organise the healthcare system in general.

Just one day after publication, the paper received a huge press echo in radio and print.

A podcast of a radio interview with the coordinator of PRONIA, Prof. Koutsouleris and project partner Stephen Wood can be found here.

An online article on this paper can be found here as well as a nice editorial of Aristotle Voinekos in JAMA Psychiatry.