The presented research projects evaluate digital clinical decision support tools for early detection of psychosis, bipolar disorder, associated metabolic risk, and suicide risk among people living with Psychosis. The Swedish component focuses on external validation, stakeholder exploration, and feasibility testing of these tools using pseudonymised individual-level data from Swedish national healthcare registers, regional Electronic Health Records (EHRs), qualitative interviews, usability testing, and anonymised survey data. The external validation study is designed as a retrospective cohort analysis with longitudinal follow-up. It integrates data from the National Patient Register (NPR), Swedish Prescribed Drug Register (SPDR), Cause of Death Register, Statistics Sweden (SCB; including LISA and the Multi-Generation Register), the Swedish Social Insurance Agency (SSIA), and regional EHR systems. The online survey is designed as a cross-sectional observational study using closed and open-ended questions to gather comprehensive feedback on the practical aspects of implementing digital risk assessment tools in mental health care, including feasibility, clinical utility, and integration challenges within existing healthcare workflows. The interview, focus group and consensus workshop studies are designed as qualitative studies using semi-structured interview guides and discussion schemes to explore stakeholders’ perspectives, experiences, attitudes and recommendations towards the use of clinical prediction models and digital tools for risk assessment in mental health services and risk communication. This application concerns secure storage, management, and analysis of sensitive health-related research data within the BIANCA secure computing environment at UPPMAX.