Data Science

Oxford

Combining routine and clinical data will lead to the personalised care of individual patients with mental health disorders, improving their outcomes and reducing NHS costs. This Theme builds on the work of the Informatics & Digital Health Theme of the previous BRC.

Mental illness is currently treated in a one-size-fits-all, trial and error approach.  Better outcomes are likely if existing therapies are augmented by available clinical trial data combined with technical advances in data synthesis, remote monitoring and using long-term outcomes from real-world datasets. This Theme will develop and refine processes, pipelines, clinical models and algorithms to aid data analysis to help customise treatment, improve outcomes and reduce NHS costs in healthcare systems.

Thed main focus will be on three interconnected research projects with close interaction with other BRC Themes, and a commitment to national mental health data efforts, e.g. HDRUK’s DATAMIND

WP1-Using multi-modal data will enable translational precision mental health. Diverse datasets across defined patient populations will be used to create a large, research-ready cohort of patients participating in diagnostic and therapeutic research.

WP2– This will focus on the capture of remote data from electronic devices, termed digital phenotyping.

WP3-Risk stratification in general and forensic mental health settings, including prisons.

Builds on four developed and validated prediction models which help assess the nature and size of risks as part of clinical decision-making in high-risk populations (eg prisoners).

WP1– clinical records, imaging, outcome measurements, digital devices/apps and other research datasets will be integrated using an advanced digital infrastructure to support patient recruitment. Targeted sub-groups of patients will be invited to undergo more detailed observations (e.g. via apps, sensors, imaging) or to participate in clinical trials. This work partners with world-class academic institutions (Harvard University, University of Toronto, Centre for Addiction and Mental Health) and industry collaborators, including Janssen and Akrivia Health.

WP2– This builds on current expertise e.g. True Colours, to utilise clinical, mathematics (AI, natural language processing), engineering/informatics and ethics to understand mental health problems and collect outcomes.  It makes use of existing rich data from patient cohorts including the Bipolar Disorder Research Network and OXTEXT.

WP3– Current free risk calculators help assess adverse outcomes in schizophrenia and bipolar disorder and have focused on suicide (OxMIS), general violence perpetration (OxMIV), forensic psychiatric patients (FoVOx) and released prisoners (OxRec). Risk stratification models will be expanded to improve their links to treatments and reinforce safety planning.

Current Theme work has involved the lived experience of people who have mental health problems in a diversity of contexts.  PPIEP work is integral to the proposed new work and will involve co-production and the representatives of local and national charities including MIND, McPin Foundation and MQ.

For more information on PPIEP please contact: Thomas.Kabir@psych.ox.ac.uk

The Theme is committed to enhancing capacity in this increasingly important area. It will continue to offer bespoke training (eg currently offered online course on systematic reviews and meta-analyses) and offer specialisation opportunities in conjunction with national and international collaborators.  It will link to the Alan Turning Institute and support Fellowships for outstanding researchers.