This study of two models of organising stroke care showed that fully centralised services were more likely to deliver clinically effective treatments. This NIHR-funded study looked at stroke care audit data before and after introducing two different models of centralised stroke services in London and Manchester. The audit recorded the proportion of stroke patients in each location who received each of seven effective care interventions, such as brain scans or clot-busting drugs within the recommended time. This study is one output from a large NIHR programme of research on the effects of stroke reconfiguration in London and Manchester. Other outputs include a published paper on the impact on patient outcomes (death and disability) and future analyses of costs and cost-effectiveness of reorganising care.
Why was this study needed?
In the UK, there are about 115 to 150 new strokes per 100,000 people per year which cost society about £8 billion annually. There is good evidence to show that in the hours after a stroke, care from a dedicated, co-ordinated team in a specialist service, using effective interventions, such as brain scans or giving blood clot busting drugs, improves long-term recovery. However, research was lacking about the relationship between different models of specialisation and effective stroke care. Before 2010, people suspected of having a stroke were taken to the nearest hospital that treated acute stroke, preferably a stroke unit. In 2010, London and Manchester specialist stroke services were centralised in different ways. London opted for a fully centralised model, opening specialist hyper-acute stroke units for all people with acute stroke. Manchester adopted a partial centralisation model, restricting specialist care to people within four hours of having a stroke, the time limit for antiplatelet treatment. After the reorganisation, the death rate fell for people treated for acute stroke, but only in London (from NIHR research published in 2014). This study wanted to understand possible reasons for this by looking at differences in the clinical processes undertaken at the two centres.
What did this study do?
This NIHR-funded controlled before-and-after study compared the proportion of acute stroke patients who received each of the following seven effective care interventions:
- Brain scans within 3 hours,
- Being admitted to a stroke unit within 4 hours,
- Brain scans within 24 hours
- Antiplatelet therapy within 48 hours
- Physiotherapy within 72 hours
- Nutrition assessment, and
- Swallowing assessment, both within 72 hours
In 2010, stroke services were centralised in London and Manchester. Two national audits (National Sentinel Stroke Clinical Audit, and the Stroke Improvement National Audit Programme) provided stroke data before (2008) and after (2012), to measure the impact. The researchers also compared the audit recordings from these cities with two urban areas that also completed both audits but did not have centralised services. The study was thorough, but was not able to use national trend data as this was incomplete. In total, data came from 38,623 people with a stroke, in 51 hospitals before centralisation, and 44 hospitals after centralisation.
What did it find?
- Both centralisation approaches led to improvements in stroke care. But overall, post reconfiguration, a higher proportion of people in London received six out of seven of the effective interventions than Manchester patients. For example, after the reconfiguration in London, 72.1% (95% CI 71.4 to 72.8) of people admitted to the centralised service received a brain scan within 3 hours compared with 65.2% (95% CI 64.3 to 66.2) in Manchester. There was no difference between London and Manchester centralised services in the likelihood of receiving antiplatelet drugs.
- However fewer Manchester patients were admitted to centralised services (39%) compared with London (93%). Only two thirds of those eligible to be admitted to hyperacute stroke centres in Manchester were admitted, for reasons not known. The authors thought this was partly because of lack of clarity around admission criteria in a complex partial centralisation model.
- Centralised specialist stroke units were significantly more likely to provide effective stroke interventions than other areas’ non-centralised services. The likelihood of receiving evidence-based care increased over time in all the areas studied.
What does current guidance say on this issue?
The National Stroke Strategy of 2007 set new standards by recommending specialist treatment in the aftermath of stroke and auditing of stroke services. The 2008 NICE stroke guideline and 2012 Royal College of Physicians stroke guideline reinforced this by recommending that people with suspected stroke are admitted to, and treated in, a specialist stroke ward or centre. The seven effective care interventions audited in this study are key parts of that care. Rehabilitation in the days following a stroke should take place on stroke wards, rather than on a general ward.
However, despite the strength of evidence about providing specialist care, there is a lack of evidence and guidance on how centralised specialist stroke services should be organised for acute care.
What are the implications?
The findings add to existing guidance by determining which of two centralised stroke service designs delivered the most effective care interventions to the most people. The study supports offering fully centralised – ‘hyperacute’ – services in large cities and a simple model that does not restrict eligibility to those who had their stroke less than four hours ago.
The researchers speculated that access eligibility criteria and having a complicated, partially centralised model restricted access to services and reduced the likelihood of admitted patients receiving effective care. They also linked their findings of more effective care in the London centres with reduced mortality in London stroke patients, compared to the rest of England.
This NIHR study shows how useful clinical audit data can be to understand what care is actually delivered and how it changes over time. More researchers and service leaders are now making use of clinical audit and other routine data to understand better variation in treatments and services.
Recent stroke service audit results to March 2015 show continuing variation in stroke care throughout the UK. Stroke services in a number of regions need to look at the quality of acute care that they are providing and consider ways of improving stroke care including centralising services.
Ramsay AI, Morris S, Hoffman A, et al. Effects of Centralizing Acute Stroke Services on Stroke Care Provision in Two Large Metropolitan Areas in England. Stroke. 2015;46(8):2244-51.
This project was funded by the National Institute for Health Research Health Services and Delivery Research Programme (project number 10/1009/09)
Intercollegiate Stroke Working Party. National clinical guideline for stroke, 4th edition. London: Royal College of Physicians; 2012.
Morris S, Hunter RM, Ramsay AI, et al. Impact of centralising acute stroke services in English metropolitan areas on mortality and length of hospital stay: difference-in-differences analysis. BMJ. 2014;349:g4757.
NICE. Acute stroke. Care pathway. London: National Institute for Health and Care Excellence; 2015.
NICE. Stroke: diagnosis and initial management of acute stroke and transient ischaemic attack (TIA). CG68. London: National Institute for Health and Care Excellence; 2008.
Royal College of Physicians, Clinical Effectiveness and Evaluation Unit on behalf of the Intercollegiate Stroke Working Party. Sentinel Stroke National Audit Programme (SSNAP) acute organisational audit report: public report for England, Wales and Northern Ireland. London: Royal College of Physicians; 2014.
Royal College of Physicians. SINAP (Stroke Improvement National Audit Programme) [internet]. London: Royal College of Physicians; 2014.
Stroke Association. State of the nation stroke statistics. London: Stroke Association; 2015.
Stroke Unit Triallists’ Collaboration. Organised inpatient (stroke unit) care for stroke. Cochrane Database Syst Rev. 2013;(9): CD000197.
Produced by the University of Southampton and Bazian on behalf of NIHR through the NIHR Dissemination Centre