This is a plain English summary of an original research article. The views expressed are those of the author(s) and reviewer(s) at the time of publication.
After some types of stroke, clot-busting drugs can restore blood flow to the brain and prevent disability. But hospitals in England and Wales vary in their use of these drugs (thrombolysis). Researchers assessed how people with stroke are managed in hospital. They used computer modelling to show how changes to stroke management could double the numbers of people who benefit from thrombolysis.
Researchers analysed audit data on 246,600 strokes between 2016 and 2018. They found a 10-fold variation in the proportion of patients who received thrombolysis. It ranged from 1 in 20 (5%) in some hospitals, to 1 in 4 (24%) in others.
The team showed that most of the variation was explained by hospital processes (such as how quickly people can have a brain scan) and in doctors’ decision-making (who they think should or should not receive thrombolysis). Some variation was due to differences between patients.
The researchers say their model could help hospitals identify the aspects of care they could change to increase the numbers of patients who receive thrombolysis. A target of 18% for England and Wales looks possible, the researchers say, but individual hospitals should have their own target.
Information about stroke and thrombolysis can be found on the NHS website.
What’s the issue?
A stroke, when the blood supply to part of the brain is cut off (by a clot, for example), is life-threatening. In England, Wales, and Northern Ireland, 85,000 people are hospitalised with stroke each year; it is a common cause of long-term disability.
Thrombolysis is a clot-busting therapy used to treat strokes caused by blood clots. It is not effective for people who have had another type of stroke caused by a bleed on the brain. The therapy needs to be given within 4 hours of stroke onset.
About 40% of people with a stroke caused by a clot arrive at hospital with known time of stroke, and sufficient time left to receive thrombolysis. Most do not receive it.
Hospitals vary in their use of thrombolysis, possibly because doctors are unsure which people should receive it. Or because of differences in hospital setup, and the length of time it takes for people to have a brain scan prior to treatment.
Researchers analysed data to determine how much of the variation in care is due to: 1) processes (how quickly a patient is scanned, for instance); 2) determination of the time of stroke onset; and 3) decision-making by doctors (which patients they would or would not treat with thrombolysis). They sought to identify ways to help hospitals reach the 20% NHS target for thrombolysis treatment.
What’s new?
The study was carried out between 2016 and 2018 and based on data from the Sentinel Stroke National Audit Programme. It included more than 246,600 people admitted to acute stroke teams in 132 units in England and Wales.
The researchers found that:
- 12% people who had a stroke outside hospital received thrombolysis
- the numbers who received thrombolysis ranged from 5% in some hospitals, to 24% in others
- most of the variation between hospitals was due to differences in hospital process and in doctors’ decision making
- some variation was due to knowledge of the time of stroke.
People are eligible for thrombolysis if they arrive at hospital within 4 hours of stroke onset. Using a computer model, the researchers looked at how to increase the numbers who receive thrombolysis appropriately.
They modelled the potential impact of 3 changes: 1. treatment within 30 minutes of arrival at hospital; 2. time of stroke onset determined in more patients; 3. thrombolysis decisions made in line with hospitals with the highest rates of thrombolysis treatment (upper quarter).
The model suggested that:
- if any single target was met, the use of thrombolysis could increase by 1 – 3% across England and Wales
- if all 3 were met, thrombolysis use could increase from 12% to 18%; this, in combination with other changes (such as faster access to brain scans) could double the numbers in whom clot-busting drugs prevent disability.
The researchers asked 19 doctors about their attitudes towards modelling and machine learning as tools to improve stroke services using audit data. Trust in the model was greatest among doctors from units with higher thrombolysis use. Some doctors wanted to know more about the model before they would trust it.
Doctors at units with lower thrombolysis rates said that differences between patients (whether the time of stroke onset could be determined, for instance) explained their lower use. Those at units with moderate use saw access to specialist resources (such as 24/7 access to brain scans and specialist stroke doctors) as the most important factor in delivering thrombolysis well. Doctors at units with high use believed that establishing a good thrombolysis pathway was the most important factor.
Why is this important?
Computer modelling and machine learning used national audit data to generate insights into how to improve services. The study demonstrated how the use of thrombolysis could be increased; use in 18% or more is achievable in England and Wales, the researchers say.
Hospitals varied in their use of thrombolysis and the model identified actions that could increase use of the treatment. Better methods of determining the time of a stroke would be helpful. This could be doctors asking, “Did they have dinner as usual, or get up in the night to go to the toilet?” Some changes are not within a hospital’s control; however, at rural hospitals, a large proportion of patients may arrive more than 4 hours after a stroke. Hospitals therefore need individual targets for thrombolysis use.
Other factors that affect use of thrombolysis may not have been collected in the Sentinel Stroke National Audit Programme. The model provides information on patterns of thrombolysis use in hospitals and is not suitable for, or intended as, a decision aid for thrombolysis in individual patients.
The data analysed are from 2016 to 2018 and the use of thrombolysis, its variation between hospitals, and how services operate may have changed since then. However, the researchers say the evidence suggests there has been little change.
What’s next?
Hospitals were anonymised in this study. The researchers are now looking at individual hospitals’ data so that they can offer them advice about how to increase thrombolysis use. This will be delivered through the Community of Practice for Thrombolysis group, which is being set up by NHS England.
The researchers have launched a web app to allow hospitals to compare their decision making (which patients they would or would not treat with thrombolysis). The researchers plan to expand the app so that hospitals can explore the likely effect of changing aspects of care. For instance, earlier access to scans for people with suspected stroke.
Methods used in this study to analyse the national stroke audit data may be transferable to other national clinical audits such as maternity care. Future studies could explore how machine learning could predict outcomes of thrombolysis for individual people, including the negative effects. This could be useful as some doctors were concerned that increased thrombolysis use could cause harm. For example, 1 in 50 have a severe brain bleed following treatment.
You may be interested to read
This Alert is based on: Allen M, and others. Using simulation and machine learning to maximise the benefit of intravenous thrombolysis in acute stroke in England and Wales: the SAMueL modelling and qualitative study. Health and Social Care Delivery Research 2022; 10: 1–184.
The researcher’s project website summarising some of their key findings.
The Stroke Association website, which provides patient information and resources.
A YouTube video explaining thrombolysis and pathways patients follow during or after a stroke.
Funding: This study was funded by the NIHR Health and Social Care Delivery Research Programme.
Conflicts of Interest: The study authors declare no conflicts of interest.
Disclaimer: NIHR Alerts are not a substitute for professional medical advice. They provide information about research which is funded or supported by the NIHR. Please note that views expressed in NIHR Alerts are those of the author(s) and reviewer(s) at the time of publication. They do not necessarily reflect the views of the NHS, the NIHR or the Department of Health and Social Care.
