A stroke patient’s chance of receiving a clot-removing treatment depends heavily on which hospital they are taken to, not just on their medical condition. This project tackles a known gap in NHS stroke care: thrombolysis and thrombectomy are proven to reduce disability, yet their use varies widely between hospitals and falls short of national targets. The researchers will combine machine learning, clinical pathway simulation, and causal inference to identify why some stroke teams deliver these treatments faster and more often than others, and whether that variation is causing worse outcomes for patients from deprived or ethnic minority backgrounds. If successful, the work will give every stroke team in England and Wales a realistic, data-driven target for thrombolysis and thrombectomy rates based on their own patient population, along with the process changes needed to reach it. The models will be released as a web application for regional stroke networks to explore their own systems. Results will feed directly into the national stroke audit, clinical guidelines, and NHS improvement programmes—meaning the findings could reshape how emergency stroke care is organised across the country.
View original technical description
RESEARCH QUESTIONS What causes hospital-level variation in the use and speed of proven medical (thrombolysis) and mechanical (thrombectomy) clot-removing treatments in emergency stroke care, and how does that variation affect inpatient lengths of stay, bed use, and outcomes across the population and across key population subgroups (especially deprived and ethnic minority populations). BACKGROUND Thrombolysis and thrombectomy are proven medical and surgical treatments that reduce disability after an acute stroke. Use of these treatments is significantly below NHS ambitions, and varies considerably between hospitals. This variation is likely to be leading to significant variation in outcomes after stroke, depending on where people live. AIMS AND OBJECTIVES We will identify variation in use/speed of clot-removal treatment by stroke team and analyse how that variation affects demographic subgroups such as ethnicity, socio-economic status, or rural vs. urban. Modelling will also investigate how variation in treatment is affecting inpatient lengths of stay and bed requirements. We will address which effects are likely to be causal and which are likely to be co-incidental. We investigate these causal relationships through both qualitative and quantitative methods. We will use co-production workshops to identify barriers and facilitators to promoting change through our work. We will help teams and regional stroke networks identify how best to reduce variation, and improve use, of clot-removing treatments. All models will be made available as a web application for stakeholders to use to understand their own emergency stroke system. This significantly extends our previous work, extending the work to investigate thrombectomy, investigate how variation in thrombolysis and thrombectomy use affects in-patient lengths of stay, and the employment of causal inference work to further strengthen conclusions on the source and effect of variation in clot-removing treatment. METHODS We will combine mixed methods: qualitative research, clinical pathway simulation, machine learning, causal inference analysis, and health economics to understand the causes and consequences of variation in emergency stroke pathways across England and Wales. TIMELINES FOR DELIVERY The project will run from Jan 2025 to Dec 2026. Output and from the project will occur throughout the project. ANTICIPATED IMPACT AND DISSEMINATION Implementation and impact will be delivered by multiple routes (necessary partnerships are already established) * Incorporation of summary results for each stroke team in the national stroke audit yearly result. This will include a realistic target for thrombolysis and thrombectomy based on a stroke team’s own patient population, and what process targets would be required to reach that overall target. * Inclusion of key results in national stroke clinical policy guidelines * Working with national stroke improvement projects (as we are with the NHS-England sponsored Thrombolysis in Acute Stroke Collaborative, TASC). * Publication of work in high impact stroke journals, and presentation at national and international stroke conferences.
Plain English summaries and category classifications on this site are generated by AI and may not perfectly reflect the original research.
Is something wrong? Let us know