Text Size:    +   -

Acute Stroke Unit Care

5th Edition
October, 2015

The Canadian Stroke Best Practice Recommendations for Acute Inpatient Stroke Care, 5th Edition (2015) is published in the International Journal of Stroke (IJS) and available freely online. To access the specific recommendations for Acute Stroke Unit Care and all other sections of the Acute Inpatient Stroke Care recommendations, please click on this URL which will take you to the recommendations online in the IJS: http://journals.sagepub.com/doi/pdf/10.1177/1747493015622461

For the French version of these recommendations, open the appendix at this link : http://wso.sagepub.com/content/11/2/239/suppl/DC1

All other supporting information, including performance measures, implementation resources, evidence summaries and references, remain available through www.strokebestpractices.ca, and not through the IJS. Please click on the appropriate sections on our website below for this additional content.

Box One: Optimal Acute Inpatient Stroke Care


A stroke unit is a specialized, geographically defined hospital unit dedicated to the management of stroke patients and staffed by an experienced interprofessional stroke team. Refer to the resource Taking Action Towards Optimal Stroke Care for detailed information about stroke unit criteria.

Alternate Stroke Care Models: It is recognized that many models of acute stroke care exist across Canada. Many organizations do not have the official administrative designation as an ‘acute stroke unit’; however they have most or all of the stroke unit criteria in place and should be recognized as attempting to meet optimal stroke care in the face of administrative/structural resource challenges. These models are sometimes referred to as clustered acute stroke care, or purposeful grouping of stroke patients.

Core Elements of Comprehensive Stroke and Neurovascular Care

(Based on Stroke Unit Trialists Collaboration 2007)  

  • It is recognized that not all hospitals are able to deliver all of the stroke unit elements, and every hospital should be Taking Action to establish protocols and processes of care to implement as many elements as possible to achieve optimal stroke care delivery within their geographic location, hospital volumes and resource availability (human, equipment, funding). Refer to Figure 1 in the Introduction and Overview section of this module, and in the Taking Action Towards Optimal Stroke Care resource kit.
  • Specialized care for patients with ischemic stroke, intracerebral hemorrhage (ICH), and transient ischemic attack (TIA) (care may be expanded in some institutions to include patients with subarachnoid hemorrhage [SAH] and other neurovascular conditions);
  • Dedicated stroke team with broad expertise – including neurology, nursing, neurosurgery, physiatry, rehabilitation professionals, pharmacists, and others;
  • Consistent clustered model where all stroke patients are cared for on the same hospital ward with dedicated stroke beds by trained and experienced staff, including rehabilitation professionals;
  • Access to 24/7 imaging and interventional neuroradiology expertise;
  • Emergent neurovascular surgery access;
  • Protocols in place for hyperacute and acute stroke management, and seamless transitions between stages of care (including pre-hospital, emergency department and inpatient care);
  • Dysphagia screening protocols in place to assess all stroke patients without prolonged time delays prior to commencing oral nutrition and oral medications;
  • Access to post-acute rehabilitation services, including inpatient, community-based, and/or early supported discharge (ESD) therapy;
  • Discharge planning starting as soon as possible after admission, and anticipating discharge needs to facilitate smooth transitions;
  • Daily/bi-weekly patient care rounds with interprofessional stroke team to conduct case reviews, discuss patient management issues, family concerns or needs, and discharge planning (discharge or transition to the next step in their care, timing, transition requirements);
  • Patient and family education that is formal, coordinated, and addresses learning needs and responds to patient and family readiness;
  • Provision of palliative care when required, ideally by a specialized palliative care team;
  • Ongoing professional development for all staff – stroke knowledge, evidence-based best practices, skill building, orientation of trainees;
  • Involvement in clinical research for stroke care.










Stroke unit care reduces the likelihood of death and disability by as much as 30 percent for men and women of any age with mild, moderate, or severe stroke. Stroke unit care is characterized by a coordinated interprofessional team approach for preventing stroke complications, preventing stroke recurrence, accelerating mobilization, and providing early rehabilitation therapy. Evidence suggests that stroke patients treated on acute stroke units have fewer complications, earlier mobilization, and pneumonia is recognized earlier. Patients should be treated in a geographically defined unit, as care through stroke pathways and by roving stroke teams do not provide the same benefit as stroke units.  Access to early rehabilitation is a key aspect of stroke unit care.  For patients with stroke, rehabilitation should start as early as possible and rehabilitation should be considered an intervention that can occur in any and all settings across the continuum of stroke care.

System Implications

  • Organized systems of stroke care including stroke units with a critical mass of trained staff (interprofessional team). If not feasible, then mechanisms for coordinating the care of stroke patients to ensure use of best practices and optimal outcomes.
  • Protocols and mechanisms to enable the rapid transfer of stroke patients from the emergency department to a specialized stroke unit as soon as possible after arrival in hospital, ideally within the first six hours.
  • Comprehensive and advanced stroke care centres should have leadership roles within their geographic regions to ensure specialized stroke care access is available to patients who may first appear at general healthcare facilities (usually remote or rural centres) and facilities with basic stroke services only.
  • Telestroke services should be optimized to ensure access to specialized stroke care across the continuum to meet individual needs (including access to rehabilitation and stroke specialists).
  • Information on geographic location of stroke units and other specialized stroke care models available to community service providers, to facilitate navigation to appropriate resources and to strengthen relationships between each sector along the stroke continuum of care.

Performance Measures

  1. Number of stroke patients who are admitted to hospital and treated on a specialized stroke unit at any time during their inpatient hospital stay for an acute stroke event (numerator) as a percentage of total number of stroke patients admitted to hospital (core).
  2. Percentage of patients discharged to their home or place of residence following an inpatient admission for stroke (core).
  3. Proportion of stroke patients who die in hospital within 7 days and within 30 days of hospital admission for an index stroke (reported by stroke type) (core).
  4. Proportion of total time in hospital for an acute stroke event spent on a stroke unit.
  5. Proportion of patients admitted to a stroke unit, who arrive in the stroke unit within 24 hours of emergency department arrival.
  6. Proportion of designated stroke unit beds that are filled with stroke patients (weekly average).
  7. Percentage increase in telehealth or telestroke coverage to remote communities to support organized stroke care across the continuum.

Refer to Canadian Stroke Quality and Performance Measurement Manual for detailed indicator definitions and calculation formulas. www.strokebestpractices.ca/

Measurement Notes

  • Performance measure 1: calculate for all cases, and then stratify by type of stroke.
  • Definition of stroke unit varies widely from institution to institution. Where stroke units do not meet the criteria defined in the recommendation, then a hierarchy of other stroke care models could be considered: a) dedicated stroke unit; (b) designated area within a general nursing unit or neuro-unit where stroke patients are clustered; (c) mobile stroke team care; (d) managed on a general nursing unit by staff using stroke guidelines and protocols.
  • Institutions collecting this data must note their operational definition of “stroke unit” to ensure standardization and validity when data is reported across institutions.
  • Performance measure 5 – start time for assessing stroke unit admission within 24 hours should be emergency department triage time.
  • Patient and family experience surveys should be in place to monitor care quality during inpatient stroke admissions

Summary of the Evidence

Evidence Table and Reference List

Stroke Unit Care

It is now well-established that patients who receive stroke unit care are more likely to survive, return home, and regain independence compared to patients who receive less organized forms of care. Stroke unit care is characterized by an experienced interprofessional stroke team, including physicians, nurses, physiotherapists, occupational therapists, speech therapists, among others, dedicated to the management of stroke patients, often located within a geographically defined space. Other features of stroke units include staff members who have an interest in stroke, routine team meetings, continuing education/training, and involvement of caregivers in the rehabilitation process. In an updated Cochrane Review, the Stroke Unit Trialists’ Collaboration (2013) identified 28 randomized and quasi-randomized trials (n=5,855) comparing stroke unit care with alternative, less organized care (e.g., an acute medical ward). Compared to less organized forms of care, stroke unit care was associated with a significant reduction in the odds of death (OR= 0.81, 95% CI 0.69 to 0.94, p = 0.005), death or institutionalization (OR=0.78, 95% CI 0.68 to 0.89, p = 0.0003), and death or dependency (OR= 0.79, 95% CI 0.68 to 0.90, p = 0.0007) at a median follow-up period of one year. Based on the results from a small number of trials, the authors also reported that the benefits of stroke unit care are maintained for periods up to 5 and 10 years post stroke. Moreover, subgroup analyses demonstrated benefits of stroke unit care regardless of sex, age, or stroke severity. Saposnik et al. (2011) investigated the differential impact of stroke unit care on four subtypes of ischemic stroke (cardioembolic, large artery disease, small vessel disease, or other) and reported that stroke unit care was associated with reduced 30-day mortality across all subtypes.

To determine if the benefits of stroke unit care demonstrated in clinical trials can be replicated in routine clinical practice, Seenan et al. (2007) conducted a systematic review of 25 observational studies (n=42,236) comparing stroke unit care to non-stroke unit care. Stroke unit care was associated with a reduction in the risk of death (OR=0.79, 95% CI 0.73 to 0.86, p<0.001) and of death or poor outcome (OR=0.87, 95% CI=0.80 to 0.95; p=0.002) within one year of stroke. Similar findings were reported for the outcome of death at one year in a secondary analysis limited to multi-centered trials (OR=0.82, 95% CI 0.77 to 0.87, p<0.001).

In-hospital Stroke

Estimates of persons who experience a stroke while already hospitalized for other conditions range from 4% to 17% (as cited by Cumbler et al. 2014). Many of these patients have pre-existing stroke risk including hypertension, diabetes, cardiac diseases, and dyslipidemia (Vera et al. 2011). These in-hospital strokes often occur following cardiac and orthopedic procedures, usually within 7 days of surgery. There is evidence to suggest that, compared with persons who suffer a stroke in the community, patients who experience an in-hospital stroke have more severe strokes, worse outcomes and do not receive care in as timely a fashion. Of 15,815 consecutive patients included in the J-MUSIC registry, (Kimura et al 2006), 694 (4.4%) experienced an in-hospital ischemic stroke. The mean admission NIHSS score was significantly higher for patients with in-hospital stroke (14.6 vs. 8.1, p<0.0001). In-hospital stroke was an independent predictor of severe stroke, defined as NIHSS score ≥11 (OR=3.27, 95% CI 2.7-3.88, p<0.0001). Significantly more in-hospital stroke patients died both in hospital (19.2% vs. 6.8%, p<0.0001) and within 28 days (12.1% vs. 4.8%, p<0.0001). Farooq et al. (2008) compared the outcomes of 177 patients who experienced an in-hospital stroke and 2,566 who were admitted from the community to 15 hospitals in a single state over a 6-month period. In-hospital case fatality was significantly higher among in-hospital patients (14.6% vs. 6.9%, p=0.04). The distribution of mRS scores was shifter towards poorer outcomes for the in-hospital group (p<0.001) and fewer in-hospital stroke patients were discharged home (22.9% vs. 52.2%, p<0.01).

One of the largest studies to examine quality of care received and stroke outcome included 21,349 patients who experienced an in-hospital ischemic stroke and were admitted to 1,280 hospitals participating in the Get with the Guideline Stroke registry from 2006-2012, and 928,885 patients admitted to hospitals from the community during the same time frame (Cumbler et al. 2014). In-hospital stroke patients were significantly less likely to meet 7 achievement standards (t-PA within 3 hours, early antithrombotics, DVT prophylaxis, antithrombotics/anticoagulants on discharge, statin meds), and were less likely to receive a dysphagia screen or receive t-PA within 3.5-4.5 hours, but were more likely to receive a referral for rehabilitation and to receive intensive statin therapy. When quality/achievement measures were combined, in-hospital stroke patients were less likely to receive investigations/care for which they were eligible (82.6% vs. 92.8%, p<0.0001). In-hospital stroke patients also experienced worse outcomes. They were less likely to be independent in ambulation at discharge (adj OR=0.42, 95% CI 0.39-0.45, p<0.001), to be discharged home (adj OR=0.37, 95% CI 0.35-0.39, p<0.001) and the odds of in-hospital mortality were significantly higher (adj OR=2.72, 95% CI 2.57-2.88, p<0.001). Although a higher percentage of patients with in-hospital stroke received thrombolytic therapy with t-PA (11% vs. 6.6%), fewer received the treatment within 3-hours (31.6% vs. 73.4%, p<0.0001).