Machine Learning Predictions Identify High-Risk Patients for Emergency Hospital Admission in Scotland
The SPARRA project, initially developed by NHS Scotland's Information Services Division, is set to revolutionise healthcare in Scotland with its innovative approach to predicting the risk of emergency hospital admission for a significant portion of the population.
Recently, the Alan Turing Institute, a leading centre for data science and artificial intelligence, has collaborated with NHS Scotland to update and extend the SPARRA project. This collaboration aims to leverage advanced machine learning and AI methods, enhancing the project's predictive capabilities.
At the heart of this project is the Department of Mathematical Sciences, a renowned institution offering postgraduate and undergraduate programmes. The department, ranked 4th in the UK in The Complete University Guide 2023, plays a crucial role in the SPARRA project, providing statistical analysis by experts in the field.
Once deployed, the SPARRA project will assess the risk of emergency admission to hospital for approximately 3.6 million patients every month, covering around 80% of the Scottish population. The new model in the SPARRA project offers substantial improvements to both precision and calibration, ensuring more patients at risk are correctly flagged and the level of risk is more accurately judged.
The analytical workflow of the SPARRA project is being improved for better predictive performance by statisticians from the project. The new model in the SPARRA project utilizes a super learner of multiple machine learning models, further enhancing its predictive capabilities.
The SPARRA project score, once calculated, is automatically delivered to GP surgeries, potentially informing primary care interventions. This tool for addressing real-world problems is a testament to the groundbreaking work being done by researchers at the forefront of global innovation.
Moreover, the Department of Mathematical Sciences offers a unique blend of high-quality teaching and research across a wide range of disciplines, providing practical experience to support future careers and employment prospects. For those interested in learning more about the work of Dr Louis Aslett and Dr James Liley, as well as the SPARRA project, the department's webpages offer valuable insights.
The work on the new SPARRA model was made possible by multiple funding sources, including the AI for Science and Government programme (Turing), Health Data Research UK, The Health Foundation, and EPSRC. The goal of the SPARRA project, while not explicitly stated in the provided search results, is undoubtedly to improve the healthcare system in Scotland by providing a powerful predictive tool for emergency hospital admissions.
