Research student, Yr Ysgol Mathemateg
- M/1.27, 21-23 Ffordd Senghennydd, Cathays, Caerdydd, CF24 4AG
I have keen interests in machine learning, operational research and game theory; particularly their applications to healthcare. Although my thesis and current research projects are more broadly related to data science in general, I am an active member of the Operational Research group.
- BSc in Mathematics (First Class Honours), Cardiff University (2017)
Please find an up-to-date copy of my CV here.
Feel free to get in contact with me in person, via email or on Twitter (@daffidwilde).
Healthcare modelling, machine learning, data analytics, applied statistics, game theory
Utilising machine learning to understand cost variability in the NHS
The health service is facing unprecedented pressure from limited cash growth, increasing demand from an ageing population linked with higher incidences of chronic conditions and a reduction in local community services.
Historically, costing has utilised Health Resource Groups (HRG’s), a currency that seeks to cluster procedures into groups consuming similar resources. As these tend to reflect procedures carried out on patients rather than conditions presented by the patient the focus of the costing community has been on elective procedures which only amount to 35% of (non mental health) admitted patients.
Cwm Taf has sought to widen the use of HRG’s by trying to understand other factors that influence variation in costs, in particular the effect of co-morbidities. The cost of treating patients has been shown to be affected by the hospital site, the type of co-morbidity, treatment ward and the age of the patient. Cwm Taf annually produces over 90,000 coded episodes and has access to a further 19,000,000 coded episodes from other Trusts and Health Boards in England and Wales. The advantage of using this type of analysis is that all this information is routinely gathered and quickly available.
Through the development of a collaborative partnership between the NHS and academia, this piece of work aims to create a mathematical model that best reflects factors affecting the patient pathway.
Cwm Taf Health Board & Cardiff University