Fellowship Bio

Paediatric registrar interested in developing technical solutions to common clinical problems. Inventor and co-developer of the Magpie mobile clinical guideline application. Industry recognised professional certification in data science, with proficiency in Python and SQL for data science applications and Web App development.

Fellowship Project

Safe Machine Assisted Real Time Transfers (SMARTT) Critical care pathways

University Hospitals Bristol NHS Foundation Trust

The SMARTT system is a platform housing a decision support tool and a multitask de-escalation checklist. The decision support tool highlights patients in intensive care in real time who will benefit from a clinical review as they may be fit for de-escalation to ward based care. The multitask de-escalation checklist then alerts clinicians to and keeps track of essential tasks that need to be completed before a safe transfer. Together, these should improve unit efficiency, reducing delays to care across the system. This will enable patients to receive the care at the level they need, at the right time for them. I have been involved in planning the clinical implementation of the SMARTT system and preparing the Clinical Safety Case for both the dashboard and decision support tool. I have also been establishing the regulatory framework into which SMARTT will sit and mapping a route to future commercialisation. The SMARTT dashboard Clinical Safety Case and program will be completed and implemented in the Summer of 2024, with the decision support tool following and running in ‘shadow’ mode by autumn 2024. This functionality will be made available after establishing it’s clinical safety in production. I plan to continue working on the SMARTT project’s implementation, having been appointed to the Topol Digital Fellowship. Additionally, I will be investigating the feasibility of using the SMARTT platform to provide novel decision support applications in Neonatal Intensive Care using similar machine learning techniques. Alongside my main project, I’ve been able to co-lead the local validation of a commercial outpatient surgical prioritisation solution by analysing 1,000,000 hospitalisation episodes over a 5 year period.

Fellowship Testimonial

The Fellowship in Clinical AI has been an invaluable resource, bringing us perspectives across clinical, management, regulatory, and industry sectors involved in the delivery of novel AI applications being implemented in the NHS today. We have benefited from the collective experience of fellows assisting in the development, implementation, or validation of AI programmes originating within the NHS or from commercial suppliers. Personally, I have had the time and resources to further develop my abilities and experience in Python for Data Science, achieving an industry-recognised certification and developing two novel web applications for implementation in 2024. I have now enrolled in the FastAI Practical Deep Learning course, aimed at domain experts across industries applying deep learning and machine learning to practical problems. I plan to continue to work on implementation of the SMARTT system through the Topol Digital Fellowship. If successful in establishing its feasibility, I will apply for further grant funding to develop its application in Neonatal Intensive Care together with the University of Bristol. I will complete my paediatric training in the next two years and will apply for consultant posts in Cornwall or Devon, where I have knowledge of local advanced IT provision and clinical links, having worked across multiple specialties. Recent advances in digital health have currently left the South-West underserved; however, the unique geography and limited number of agencies delivering care mean these areas could implement technologies effectively and benefit most from these advances. With the unique skills that I will continue to develop after the fellowship, I aim to become a Trust-level AI transformation lead.