Improving the outcome of stroke

Abstract

UK needs to reorganise services to follow the example of other countries

The outcome of stroke varies markedly between countries. A study of 12 centres in seven European countries found that mortality varied twofold even when adjusted for case mix and use of healthcare resources.¹ Similar variation was also found in two large international multicentre trials of acute stroke.^{2 3} All three studies found the outcome was worst in the United Kingdom; in one study the differences in the proportion of patients dead or dependent between the UK and the other eight countries were between 150 and 300 events per 1000 patients.²

What underlies this variation and why is outcome so poor in the UK compared with countries with similar economies in western Europe? Residual confounding by case mix is difficult to exclude completely, but differences in the process of care are likely to be important.³ In many European countries stroke care is an integral part of neurology. In contrast, in the UK it has, until recently, been a “Cinderella” subject, often falling between neurology and general and geriatric medicine. It is tempting to conclude that this lack of interest has led to underinvestment and a resulting poor outcome. However, the cost of care of stroke patients seems to be as high, if not higher, in the UK than in European countries with better outcomes.^{1 4} This suggests that organisational and structural problems in delivery of resources are important.

Limited data show that European countries with better outcomes focus resources more heavily on the acute aspects of care.¹ The vast majority of the cost of in-hospital stroke care in the UK is for nursing and hospital overheads, with the cost of investigations and medical care being very low. The higher length of stay in England found in comparative studies suggests that improvements in acute care could not only improve outcome but also lower costs by reducing length of stay.

Organisation of acute stroke care has become even more important now that there are specific treatments for acute stroke. Thrombolysis with alteplase (tPA) improves outcome if given within three hours of ischaemic stroke onset.⁵ Providing thrombolysis is challenging even in countries with well developed services. Essential components include patient education and awareness of stroke symptoms, rapid ambulance assessment and transfer to specialised stroke centres, and rapid brain imaging to exclude haemorrhage before administration of alteplase. Despite these challenges, effective thrombolysis services exist in many countries in Europe, North America, and Australia, in both urban and rural settings, with as many as 20-30% of eligible patients receiving thrombolytic therapy.⁴ Currently less than 1% receive such therapy in the UK.⁴

These deficiencies in stroke care have already been recognised in England in a 2005 National Audit Office report.⁴ The report concluded that if care was better organised, annually £20m (€29m; $41m) could be saved and 550 deaths avoided and 1700 patients would recover fully who would not otherwise do so. In response, England's Department of Health National Stroke Strategy is due to publish its recommendations in autumn 2007.

A major challenge is to change the perception of stroke, among both health professionals and the public, so that stroke is viewed as a condition that requires emergency action. This will require major structural changes at several levels. Despite robust evidence of the efficacy of organised stroke unit care, the 2006 Royal College of Physicians Stroke Sentinel Audit found that only 62% of people admitted for stroke in England, Wales, and Northern Ireland were treated in a stroke unit at any time during their stay, while only 54% spent more than half of their stay in a specialised unit.⁶ The UK has a severe shortage of specialists trained in acute stroke care. This is being remedied by the recently developed stroke subspecialty training programme but will need many years to be fully corrected. Developing acute stroke services and implementing thrombolysis requires not only specialised acute medical teams but also access to computed tomography, and when required magnetic resonance imaging, and brain imaging. Brain computed tomography is the “electrocardiography” of stroke. In many European countries it is performed on admission in the accident and emergency department, while in the UK many units struggle to provide it within 24 hours.⁴ The neolistic response that “it will make no difference to management” must be overcome now that we have effective treatments for acute stroke and research has shown that scanning patients immediately is the most cost effective strategy.⁷ Implementing thrombolysis will need 24 hour availability of specialised expertise, including stroke specialists and imaging support. It is unlikely that every acute hospital will be able to provide such a service, and alternative strategies are needed. These include forming larger regional centres or telemedicine approaches, as successfully implemented in the United States and Germany.⁸

Increasing the proportion of patients receiving thrombolysis will undoubtedly improve outcome, but even in the best units only a minority of patients will be eligible. The benefit of thrombolysis beyond three hours is being examined in international trials, although we know that its efficacy falls dramatically with time from stroke, even within the first three hours.⁸ Furthermore, the risk of intracerebral haemorrhage secondary to alteplase increases with time from onset of stroke. It is hoped that even at later time points newer magnetic resonance imaging and computed tomography techniques will be able to distinguish between patients with potentially reversible damage, who may benefit from thrombolysis, and those with no salvageable tissue in whom alteplase can only cause harm.

In addition, the early risk of recurrent stroke is much higher than previously thought—as high as 10-15% in the first week.⁹ Much of this increased risk is within the first 48 hours. More aggressive antiplatelet regimens in the first days after stroke may prove effective but can only be administered after acute imaging to exclude haemorrhage. The risk of recurrent stroke is particularly high in people with carotid artery stenosis,⁹ who require rapid identification and consideration for carotid endarterectomy.

Probably the most important outcome of reorganisation of services will be a general improvement in acute care of stroke. Improved early diagnosis with imaging, together with improved monitoring and treatment of physiological parameters, will improve outcome independent of administration of thrombolysis. If we can set such acute systems in place they will also facilitate implementation of other new treatments.