Stroke and maternity#
Active projects#
SAMueL-2 (Stroke Audit Machine Learning 2)#
Stroke is a common cause of adult disability. Expert opinion is that about one in five patients should receive clot-busting drugs to break up the blood clot that is causing their stroke. At the moment only about one in nine patients actually receive this treatment in the UK. There is a lot of variation between hospitals, which means that the same patient might receive different treatment in different hospitals. In SAMueL-1 we developed the methods for understanding what are the main causes of variation between hospitals: How much difference is due to processes (like how quickly a patient is scanned), how much is due to differences in patient populations, and how much difference is due to different decision-making by doctors. In SAMueL-2 we extend this work to understand which patients different hospitals would treat differently, and we predict the effect these differences have on patient outcomes (death and/or disability on discharge).
What do teams do differently? : link
OPTIMIST (OPTimising IMplementation of Ischaemic Stroke Thrombectomy)#
Stroke is a leading cause of adult disability, with about 1 in 10 patients eligible for emergency thrombectomy, which removes blood clots blocking brain arteries. In England, 24 specialized centres offer this procedure, but most patients first go to local stroke units where clot-busting drugs are given. Thrombectomy is more effective for large artery blockages. Transporting suitable patients to specialized centres delays treatment by 90 minutes. The OPTIMIST project aims to streamline this process by developing methods for ambulance staff to identify patients who benefit most from direct transport to thrombectomy centre. It involves a clinical trial and modelling to assess patient outcomes and the impact on local stroke services.
MUSTER (Modelling mobile stroke Units for Stroke Treatment Equity and Resources)#
Mobile Stroke Units (MSUs) offer an innovative approach to emergency stroke care, delivering treatment in the community rather than hospitals. International research demonstrates their effectiveness in improving patient outcomes and increasing treatment accessibility. In 2022, the European Stroke Organisation (ESO) strongly endorsed MSUs. While MSUs typically incur setup costs of around £1 million and daily running expenses, their integration into the National Health Service lacks clear guidance. Our study aims to provide policymakers with robust computer models of MSUs, assessing resource needs, patient benefits, cost-effectiveness, and implications for treatment disparities based on geography and population density. Through this research, we aim to inform local and national policies regarding the deployment of MSUs in the British healthcare system.
STROKE-PREDICT (Using Explainable Artificial Intelligence to predict future stroke using routine historical investigations)#
Strokes are a major health problem in the UK, with many preventable cases. The aim of this project to use AI to analyze patient data like brain scans, monitoring of electrical activity of the heart, and other medical tests to predict risk of the person having a stroke. By analyzing data from 10,000 past stroke cases, we hope to develop a model that can identify people at high risk. The project will build different AI models for various tests and combine them to obtain the best accuracy. We will also check how reliable the predictions are and explain what factors the AI uses to make those predictions. Ultimately, this initiative aims to revolutionize stroke prevention by identifying both known and novel risk factors, potentially lengthening lives, avoiding disability, and reducing healthcare expenses.
HDR-UK BHF Stroke Catalyst#
When someone has a stroke, they are first taken to hospital for emergency treatment and rehabilitation care. We wish to understand the longer-term consequences of stroke on the patient and to the NHS (especially the length of time patients spend in different hospitals in the year after their stroke). We wish to understand how that length of time varies for different patient characteristics (such as age, stroke severity, the wealth of their home area, COVID history) and what emergency treatments they received. This project will address those questions by accessing broad patient-level data through the new British Heart Foundation Stroke Data Science Catalyst which gives us access to a broad range of patient data from hospitals, GPs, and pharmacies.
BHF Stroke Data Science Catalyst: link
Archived projects#
SAMueL (stroke Audit Machine Learning)#
Stroke is a common cause of adult disability. Expert opinion is that about one in five patients should receive clot-busting drugs to break up the blood clot that is causing their stroke. At the moment only about one in nine patients actually receive this treatment in the UK. There is a lot of variation between hospitals, which means that the same patient might receive different treatment in different hospitals.
In SAMueL-1 we developed the methods for understanding what are the main causes of variation between hospitals: How much difference is due to processes (like how quickly a patient is scanned), how much is due to differences in patient populations, and how much difference is due to different decision-making by doctors. By applying these methods we found that thrombolysis rates could reasonable be doubled, but that each hospital should have its own target, taking into account the characteristics of the local population.
Project WebSite: link