Dr Samuel Leighton

Affiliate (School of Health & Wellbeing)

email: Samuel.Leighton@glasgow.ac.uk

Clarice Pears Building, 90 Byres Road, Glasgow, G12 8TB

Biography

Dr Sam Leighton is an ST6 General Adult Psychiatrist and Clinical Academic Fellow based in Mental Health & Wellbeing. He is interested in using machine learning to predict outcomes in individuals with a first episode of psychosis.

Dr Leighton graduated from the University of Edinburgh with an MBChB in 2011, having attained an intercalated BMedSci (Hons) in Biochemistry in 2008. Dr Leighton completed his Membership of the Royal College of Physicians in 2014 and his Membership of the Royal College of Psychiatrists in 2016.

During medical school, he taught himself several programming languages (Java, Python, XML, SQLite and R). Later, while working as a junior doctor, he used these skills to help doctors access up-to-date clinical guidance. Dr Leighton developed the NHS Greater Glasgow & Clyde (GGC) Primary Care Antimicrobial Guidelines Prescribing App (in conjunction with the GGC Antimicrobial Utilisation Committee). This app has been used >300,000 times in 195 countries. Dr Leighton organises training in technology for clinicians and has been invited to participate in NHS and Scottish Government advisory groups on information governance and national eHealth policy.

In 2018, Dr Leighton joined the University of Glasgow as a Clinical Lecturer. In this post he began to combine his growing research and bioinformatic skills in the analysis of large clinical datasets. He explored the role of inflammation in psychiatric illness by undertaking a meta-analysis of chemokines in depression, incorporating predictive modelling using R statistical programming.

More recently, Dr Leighton has been focussed on harnessing the power of clinical data to improve patient outcomes in psychosis. This work is only possible through collaboration, including with Dr Pavan Mallikarjun at the University of Birmingham, on several large datasets of first episode psychosis patients with longitudinal follow-up.

Clinical Work

Dr Leighton works as an ST6 General Adult Psychiatrist within NHS Greater Glasgow & Clyde.

Research interests

Dr Leighton is currently undertaking a competitively awarded Chief Scientist Office Clinical Academic Fellowship supervised by Professor Jonathan Cavanagh, Dr Fani Deligianni and Dr Rajeev Krishnadas entitled “Developing biologically relevant prediction models for individual patients with First Episode Psychosis using machine learning”.

Psychosis is a mental illness characterised by hearing voices, muddled thoughts and false beliefs. It is associated with huge morbidity and indeed mortality. Only four in ten reach full functional and symptomatic recovery. We know factors, including premorbid functioning, which are associated with a poor outcome at a group level. However, we currently struggle to predict outcomes at an individual level. Prediction modelling via machine learning has the potential to revolutionalise medicine by allowing us to predict outcomes at an individual level.

It has been successfully employed clinically in other areas of medicine. The QRISK tool predicts cardiovascular risk in individual patients. Based on this score, treatment decisions are made. This is the basis of precision medicine.

Within psychiatry, Dr Leighton has recently published the first externally validated evidence of the ability to predict symptomatic and recovery outcomes in individual patients with first episode psychosis using retrospective data.

Dr Leighton's CSO PhD studentship is focussed on answering three questions:

How do prediction models perform prospectively in first episode psychosis patients in a real-world clinical setting?
Can model performance be improved by the addition of biologically relevant disease markers like peripheral inflammatory bloods or magnetic resonance spectroscopy glutamate?
Can performance be improved by the incorporation of dynamic variables, those which change over time, like response to treatment?

This will be accomplished via the EMPATH platform. EMPATH stands for Electronic Measures in Psychosis – Assessing Trajectory and Health-outcomes and will be used for the recording of routinely collected clinical data from NHS Greater Glasgow & Clyde’s ESTEEM service, the only first episode psychosis service in Scotland.

This work has clear translational benefits. If we can accurately predict at an individual level who will do well and who will do poorly, we can focus resources on those who need it most and improve outcomes for patients with psychosis.