In 2023, a project undertaken as part of the Population Data Science Summer Internship Programme led to a published research methods paper offering critical insights into how the Welsh Index of Multiple Deprivation (WIMD) should be used in health data research. Just months after its publication in BMC Public Health, the Welsh Ambulance Services NHS Trust (WAST) plan to apply these findings operationally – demonstrating how research can directly influence real-world health service delivery.
From Research to Practice
The paper, “Estimating the Effect of Excluding Domains on a Routine Health Data Study Using the Welsh Index of Multiple Deprivation”, was produced by a collaboration of SAIL Databank researchers at Swansea University. It serves as an important resource for researchers using the WIMD in health data studies and is relevant to most research projects using SAIL. The team created a single citable resource for researchers using the WIMD in their research, highlighting the effect of ignoring a commonly overlooked source of endogeneity bias, and offers a methodology to mitigate this bias.
The paper explains, ‘Endogeneity bias can occur when there is a correlation between an outcome and predictor variable. For example, the Welsh Index of Multiple Deprivation (WIMD) 2019 includes seven health indicators’, which the authors suggest leads to bias and skewed health outcomes when used in research.
This nuance had a direct impact on how WAST intend to use the WIMD-health within risk profiling and risk stratification projects, with the hope it will help them investigate how deprivation correlates to ambulance demand in the form of clinical presentations such as stroke and MI.
After reviewing the paper, WAST Senior Analyst Adam Nicholls and Clinical Development Lead Luke Watkins reported the following:
“It’s great to see work that will have such a significant operational impact for the NHS. After reviewing your paper, I’ve made a few adjustments to my data:
- Used the Office for National Statistics (ONS) 2011 data and extracted the Welsh postcodes and Lower-layer Super Output Areas (LSOAs).
- Removed the Health domain from the WIMD and recalculated the weights, ranks, deciles, and quintiles.
- Integrated this with our patient data.
The results showed that removing the health domain led to scores becoming more concentrated around the centre, with a shorter high-deprivation tail. The ranks and quintiles shifted in a similar pattern to those reported in your study. I hope this shows the impact the paper has been having in informing our risk stratification work within WAST!”
Researcher Reflections
Dr Hywel Turner Evans, Senior Data Scientist & User Support Lead at SAIL Databank, and a co-author on the paper, expressed the team’s pride in seeing the research being used in this way:
“This is the kind of impact we hope to have – if we’re lucky – at some point in our careers. We should all be really proud of what was achieved in this project. To see something developed in a summer internship being adopted by operational teams in the NHS within a year is incredible. It shows what rigorous methods and the right partnerships can achieve.”
Building the Next Generation of Data Scientists
This impactful research originated from the 2023 Population Data Science Summer Internship Programme. Intern Shamsudeen Mohammed, who co-authored the paper, has since progressed to become a Research Fellow at the London School of Hygiene and Tropical Medicine.
The internship not only offers hands-on experience with real-world data but also cultivates research that delivers lasting value – something Dr Hywel Turner Evans highlights as central to the programme’s success:
“The ability for interns to contribute meaningfully to high-quality research is what makes this programme truly impactful.”
Prof Ashley Akbari, Professor of Population Data Science at Swansea University, Director of the Population Data Science Internship Programme and Head of the User and Data Support Services Team and Research Innovation at SAIL Databank, added:
“We’re proud to offer our interns real-world experiences that connect data science with public and service impact. This paper is a great demonstration of our interns working on active research projects, technical developments and impactful programmes that contribute directly to the impactful research and delivery.”
The 2025 Summer Internship Programme is now underway. Stay updated on the latest news and intern projects here.
Looking Ahead
With WAST’s Clinical Development Lead expressing interest in deeper collaboration, the research continues to inform conversations about service planning and healthcare access across Wales.
This is the kind of evidence-to-impact pipeline that SAIL Databank hopes to encourage – where research grounded in rigorous data science not only improves methodological standards but also leads to measurable change in public services.
Read the full research paper here.
