Key points

Most improvements in stroke care to date have been driven by research.

Immediate access to advanced imaging allows fast decision making, is cost-effective, and improves outcome.

Hyperacute interventions for acute ischaemic and haemorrhagic stroke can prevent permanent brain damage and reduce disability.

Strokes and stroke complications do not just happen during working hours: 24/7 working is essential for effective stroke management.

High quality nursing care is essential and has been shown to have a major impact on survival.

Pneumonia is the most common post-stroke complication, and can be prevented by early swallow assessment.

Urinary catheters are associated with infections and should be avoided.

Foot pumps reduce thromboembolism and save lives.

1 Introduction

It is well established that stroke units save lives, reduce disability, and allow more stroke survivors to return to their own homes (1). Evidence suggesting a benefit from treatment in dedicated stroke units first emerged in the 1980s with single unit trials, but has only been widely accepted after statistical significance of the survival advantage for patients treated in stroke units was shown in a systematic review (2). It has nevertheless been argued that stroke units tested in RCTs are not representative and that benefits may be overestimated (3). This concern has been put to rest by a recent review of observational studies of stroke units in their natural habitat which showed that the benefit associated with stroke-unit care in routine clinical practice is comparable to that seen in RCTs (4). Furthermore, evidence is now emerging that benefit from stroke-unit care is not restricted to the largely Anglo-American and Scandinavian ‘homelands’ of stroke-unit trials, but can also be shown in other health care systems throughout the world (5, 6). What is less clear is which aspects of organized stroke care determine better outcome.

Evidence for stroke-unit care, different types of stroke unit, which patients benefit, and current practice is reviewed by Kalra (Chapter 13). The aims of this chapter will be to describe key aspects of the content of the stroke unit ‘black box’ and how various interventions have been shaped by evidence or the lack thereof. The chapter will end with a brief exploration of future developments in stroke care and research.

2 Structural elements of specialist stroke services

Stroke units in the UK can be classified either by the stage of stroke recovery of the patient: hyperacute (the first 72 h after stroke), acute, rehabilitation, and combined units, or by the services that the unit provides: comprehensive stroke centres, primary stroke centres, and acute stroke-ready hospitals. The US Brain Attack Coalition defines a comprehensive stroke centre as a facility that provides high-intensity medical and surgical care, specialized diagnostics, and interventional therapies (7). Key elements of primary stroke centres include acute stroke teams, stroke units, written care protocols, an integrated emergency response system, availability and interpretation of computed tomography scans 24 hours a day, rapid laboratory testing, administrative support, strong leadership, and continuing education (8). In 2011 the coalition authors refined the recommendations (9) into seven key requirements:

acute stroke teams

telemetry monitoring

brain imaging with MRI

assessment of cerebral vasculature with magnetic resonance (MR) or computerized tomography (CT) angiography

cardiac imaging

early initiation of rehabilitation

certification by an independent body, including site visits and disease performance measures.

A third type of facility, the acute stroke-ready hospital, has fewer capabilities and more limited resources, but is organized to diagnose, stabilize, treat, and transfer most patients with stroke. In 2013 the European Stroke Organization published criteria for stroke units and stroke centres. These specify 37 and 65 requirements for certification of the two types of unit respectively, and cover, inter alia, departments, staff, facilities for investigation, hyperacute interventions, protocols, pathways, monitoring, assessment, and multiprofessional teams (10). In the UK requirements for stroke units are summarized in the National Stroke Guidelines published by the Royal College of Physicians (2012), with more detail in the Stroke Service Standards (11) and the Consultant Workforce Requirements (12) by the British Association of Stroke Physicians.

While new acute interventions such as thrombolysis and endovascular treatments for stroke emphasize the need for hyperacute or comprehensive stroke centres, the strongest evidence for the effectiveness of stroke units is based on trials conducted later in the stroke journey in combined and rehabilitation stroke units (13), and relates to interventions aimed at prevention of complications rather than the brain injury.

3 Key aspects of stroke-unit practice and treatment associated with good outcome

3.1 Hyperacute stroke treatments

Over the past two decades a number of acute stroke treatments have been developed and shown to be effective in clinical trials. This research has changed the face of stroke care and introduced the concept of stroke medicine. Ischaemic stroke is caused by blockage of an artery supplying part of the brain, causing cell death and loss of function. The key concept underlying hyperacute stroke care is that symptoms can be reversed and the stroke aborted or reduced in severity if this artery can be unblocked before cell death occurs. Thrombolysis with alteplase was first shown to be effective for patients under the age of 80 years if given within three hours of symptom onset by the National Institutes of Neurological Disorders stroke trial in 1995 (14). The time frame was subsequently extended to 4.5 h with the results of the ECASS-3 trial (15), and inclusion widened to patients over the age of 80 following the IST-3 study (16). The results of IST-3 also confirm that ‘time is brain’, with the potential for full recovery highest when treatment is started within 90 minutes, and little or no benefit after 4.5 hours.

Intravenous thrombolysis is contraindicated in patients with increased bleeding risk. Timely recanalization with intravenous thrombolysis is considerably less likely if the thrombus burden is large, such as in occlusions of the carotid, proximal middle cerebral, and basilar arteries. Mechanical thrombectomy, using a device introduced via the femoral artery into the occluded brain vessel to extract the clot, can be used where thrombolysis is contraindicated or has failed. A systematic review of case series suggests that this can be performed safely (mean mortality 17%, range 4–44%) and is at least as effective, if not more so, than intravenous thrombolysis (independent recovery in 42%, range 15–54%) (17). The first RCTs of intra-arterial thrombolysis and/or thrombectomy (IMS-3, MRRESCUE, SYNTHESIS) have not shown significant benefits over conventional treatment. Those studies, however, had significant limitations: outdated devices, low treatment rate, and very late interventions (18). Several studies using newer devices and tighter inclusion criteria are ongoing (MRCLEAN, THRACE, PISTE).

Decompressive hemicraniectomy significantly improves survival in young patients with a large ischaemic stroke and malignant middle cerebral oedema, previously an almost invariably fatal condition. The number of patients who need to be treated with this procedure to save one life (NNT) is as low as two (19). More recently decompressive hemicraniectomy has also been shown to be effective in older people up to the age of 82 years with a NNT of 4 (20).

Arguably one of the most important reasons for recent improvements in stroke outcomes is the availability of acute and hyperacute treatments. While these are still applicable only to a very small proportion of stroke patients, the idea that something can be done has changed attitudes of medical staff from seeing stroke as a fate to be managed to a disease to be treated.

3.2 Better management of intracerebral haemorrhage

Stroke-unit care does not only benefit patients with cerebral infarcts. It also benefits those with intracerebral haemorrhage, with a 38% reduction in 30-day mortality (p = 0.007) (21). There is no robust evidence to support any specific therapeutic strategy, but there is good evidence that some interventions are not beneficial. These include the administration of corticosteroids (22), early surgery for intracerebral haemorrhage (23), and haemostatic therapy in patients with normal coagulation (22). Following common sense, but not formally tested, approaches such as rapid correction of coagulation disorders, control of very high blood pressure, stabilization of haemodynamic parameters, early mobilization to prevent thromboembolism, adequate nutrition, airway hygiene, and good nursing care are all likely to contribute to better outcomes.

The negative results of the STICH and STICH-2 (23, 24) trials comparing early surgery with delayed surgery are often falsely interpreted to mean that surgery for intracerebral haemorrhage is not effective. But patients were included only when the surgeon was uncertain about benefits of ‘immediate’ (e.g. within 12 hours) intervention. Neither trial addressed the question of effectiveness of surgery as such, but rather the timing of the intervention. In the management of cerebral infarct, ‘early’ relates to the first 4.5 hours after symptom onset, but in the STICH studies ‘early’ was defined as up to 60 hours after the ictus. This may have been too late for prevention of secondary brain damage. Nevertheless, both trials showed a non-significant reduction in death and disability. While this result does not give a clear indication for surgery, it definitely shows that patients were not harmed by early intervention. Surgery should therefore still be considered in selected patients fit for the procedure.

Potential indications for surgery are large cerebellar haemorrhages and cortical haematomas >30 ml and less than 1 cm from the surface (25). A trial of tight blood pressure control (INTERACT-2) with rapid reduction of the systolic pressure to 140 mmHg has also shown a trend towards better outcomes (26).

Conversely, early withdrawal of active management is associated with adverse outcomes. In particular, early institution of do-not- resuscitate (DNR) orders is associated with an increased risk of death, even after correction for case mix and co-morbidities. The American Heart Association guidelines for management of intracerebral haemorrhage therefore state that DNR orders should not be made within the first 48 hours, and that patients with DNR orders should continue to be given active treatment (25, 27).

It becomes increasingly clear that active management results in better outcomes. Ongoing studies will give us more answers. These include the TICH-2 study examining the effect of tranexamic acid as a haemostatic agent and the CLEAR III study of clot lysis for intraventricular haemorrhage.

3.3 Immediate access to advanced imaging and diagnostics

National stroke guidelines in the UK recommend that brain imaging should be performed within one hour of arrival under the following conditions: if there are indications for thrombolysis or early anticoagulation; if there is a known reversible bleeding risk (to exclude a haemorrhage); or if a non-vascular cause is suspected. These criteria are fulfilled in approximately 50% of acute strokes, with CT scanning recommended in the remainder as soon as possible within 24 hours (28). Immediate access to CT scanning for all patients presenting with acute stroke is practised in many units in Europe and the US. Early universal scanning has been established as a more effective and cost-effective diagnostic strategy than early scanning only for patients with specific indications (29). Further, more advanced imaging techniques, such as CT angiogram, CT perfusion, and magnetic resonance imaging, allow rapid and accurate diagnosis of the site of the occlusion and viability of the ischaemic penumbra, and therefore facilitate therapeutic decision making. This is supported by evidence from a meta-analysis of stroke-unit trials which showed that 24/7 availability of MRI scanning was associated with better outcomes (30).