For providing appropriate and personalized care in mental health care (GGZ), it is necessary to evaluate the quality of delivered care. Currently, we lack sufficient information on how well treatments work in practice due to insufficient available data. It is possible to evaluate using data from medical records instead of additional questionnaires. However, it is crucial to protect patient privacy and comply with GDPR regulations.

A new data technique, federated learning (FL), learns from data in records without moving this data. With FL, what is learned can be shared and aggregated among institutions. In our project, we aim to: 1) promote collaboration among GGZ institutions by bringing together existing FL initiatives, 2) evaluate the effect of an intervention using FL with three institutions, and 3) draft a national advisory on Healthcare Evaluation using FL based on lessons learned.