The Use of Prognostic Models and Stratified Care in Severe Mental Illness: Protocol for a Rapid Scoping Review of Reviews

Severe mental illness (SMI), encompassing conditions like schizophrenia, psychosis and bipolar disorder, significantly impact individuals and healthcare systems globally. Characterised by persistent and often debilitating symptoms, SMI can impede daily functioning, limit quality of life, and increase healthcare utilisation. Traditional, one-size-fits-all treatment approaches often fall short, highlighting the need for more personalised and effective care.

Prognostic research offers a promising avenue for personalised care in SMI. The MRC Progress framework describes four steps for prognosis research, including fundamental prognosis research, prognostic risk factor research, prognostic model research and stratified care research.

Modern prognostic research makes use of statistical methods to analyse data, but the recent growth of machine learning (ML) has led to an explosion of interest in prognostic research. A strength of ML techniques is that they allow us to handle massive, complex datasets with numerous variables, allowing for more accurate and personalised predictions. To identify opportunities for further development of personalised care in SMI using machine learning approaches, a rapid scoping review of reviews is necessary to highlight the strengths, limitations and research recommendations from previous reviews of prognostic research in SMI.