The Elements of Stroke Rehabilitation

EBRSR [Evidence-Based Review of Stroke Rehabilitation] 6 The Elements of Stroke Rehabilitation Robert Teasell MD, Norine Foley MSc, Norhayati Hussein MBBS, Mark Speechley PhD Last updated November 2013 Abstract The primary goals of stroke rehabilitation are to encourage and foster functional improvement and neurological recovery. Organised stroke care, processes of care, early timing of rehabilitation and high intensity of rehabilitation therapies are important factors which have been identified as promoting better overall outcomes for individuals with stroke. This chapter examines the evidence for those elements which have been proven to be important in the effectiveness of stroke rehabilitation. 6. The Elements of Stroke Rehabilitation pg. 1 of 54

Key Points Individuals with stroke who received organised inpatient stroke units care are more likely to survive, return home and regain independence. Many elements contribute to the success of stroke rehabilitation units. It is unclear if patients who have suffered a hemorrhagic stroke have worse outcomes compared with those who have had an ischemic stroke. There is insufficient evidence to support routine implementation of care pathways for acute stroke management or stroke rehabilitation as care pathways do not improve stroke rehabilitation outcomes or reduce costs. However, compliance with stroke rehabilitation guidelines does improve outcomes. Stroke patients who are appropriate candidates should be admitted to a rehabilitation unit or facility as soon as possible. Very early mobilization may help to prevent medical complications post stroke. Greater intensities of physiotherapy and occupational therapy result in improved functional outcome. It is uncertain whether more-intensive language therapy is better than less-intensive therapy. Greater functional improvements made on interdisciplinary stroke rehabilitation units are maintained over the long-term. Dr. Robert Teasell 801 Commissioners Road East, London, Ontario, Canada, N6C 5J1 Phone: 519.685.4000 Web: Email: Robert.Teasell@sjhc.london.on.ca 6. The Elements of Stroke Rehabilitation pg. 2 of 54

Table of Contents Abstract... 1 Key Points... 2 Table of Contents... 3 6. The Elements of Stroke Rehabilitation... 4 6.1 Functional Improvements and Neurological Recovery... 4 6.2 Hemorrhagic vs. Ischemic Stroke... 5 6.3 Elements of a Stroke Unit Associated with Improved Outcome... 7 6.3.1 What Form of Stroke Unit is Best?... 7 6.4 What Elements of Care are Associated with Improved Outcomes?... 9 6.5 Impact of Care Pathways and Guidelines... 12 6.6 Timing of Stroke Rehabilitation... 15 6.7 Intensity of Therapy... 21 6.7.1 Intensity of Physiotherapy and Occupational Therapy... 21 6.7.2 Previous Reviews and Meta-Analyses... 22 6.7.3 Intensity of Aphasia Therapy Post Stroke... 32 6.8 Durability of Rehabilitation Gains... 36 6.8.1 Previous Reviews... 36 6.9 Summary... 42 References... 44 6. The Elements of Stroke Rehabilitation pg. 3 of 54

6. The Elements of Stroke Rehabilitation 6.1 Functional Improvements and Neurological Recovery Some authors have suggested that the majority of functional recovery after stroke is simply related to spontaneous natural recovery from neurological impairment (Lind 1982, Dobkin 1989). The fact that specialized stroke rehabilitation units and greater intensities of rehabilitation (see later discussion) are associated with improved functional outcomes suggests that neurological recovery alone cannot account for the degree of functional improvements seen in stroke rehabilitation. A more basic science approach to the issue of the brain recovery following stroke is found in section 2. In terms of the expected time course of recovery following stroke, Yagura et al. (2003) suggested that functional recovery is thought to reach 80% of maximum by three months, 95%, by six months and 100% by 12 months. Table 6.1: Studies on Functional Improvement and Neurological Recovery Author, Year Methods Country Pedro Score Roth et al. 1998 USA No Score Kwakkel et al. 2006 The Netherlands No Score A prospective cohort of 402 patients consecutively admitted for rehabilitation within three months of stroke was studied. Impairment was quantified with the National Institutes of Health Stroke Scale (NIHSS) while the Functional Independence Measure (FIM) was used to measure disability, in particular the motor and cognitive subscales of the FIM. 101 patients with first-ever ischemic strokes were studied during the first 16 weeks post stroke. Progress of time was categorized into 8 bi-weekly time intervals and was used as the independent covariate in a first-order longitudinal regression model. The bi-weekly time change (progress of time) was related to improvement in upper and lower limb motor recovery assessed with Fugl- Meyer (FM) score and Motricity Index (MI), reduction in visuospatial inattention based on the letter cancellation task, and improvement in walking ability, dexterity, and activities of daily living measured with the Functional Ambulation Categories (FAC), Action Research Arm test, and Barthel Index. Outcomes 342 patients experienced no substantial reduction while 60 patients had a significant reduction in their impairment level. Both groups experienced significant reductions in disability during rehabilitation, including those patients who did not experience a substantial reduction in impairment. Time explained a significant change of 8.4 (42%) measurement units on the Barthel Index for the first 10 weeks poststroke, 1.1 (22%) measurement units on FAC, and 19% on the Action Research Arm test for the first 6 and 8 weeks poststroke. Approximately 25% (for FM-arm) to 26% (for MI-arm) of the significant change in measurements units was explained by time alone for the upper limb compared with 33% for FM-leg and 39% for MI-leg of the lower limb. Time accounted for a reduction of 16% in the letter cancellation task. Observed associations did not change after controlling for covariates such as age, gender, hemisphere of stroke, type of stroke, or intervention. In their assessment of the relationship between impairment and disability after a stroke, Roth et al. (1998) concluded that, although stroke-related impairment and disability are significantly correlated with each other, reduced impairment level alone does not fully explain the reduced disability that occurs during rehabilitation. Even patients without substantial impairment reduction demonstrate disability reduction during rehabilitation, suggesting that rehabilitation has an independent role in improving function beyond that explained by neurological recovery alone. Kwakkel et al. (2006) using a modified form of regression analysis estimated that time alone could explain between 16% and 42% of the observed improvement in many parameters of recovery during the first 6 to 10 weeks post stroke. 6. The Elements of Stroke Rehabilitation pg. 4 of 54

However, the authors speculated that the effects of time on recovery were likely overestimated given that patients in the study also received therapeutic interventions. Conclusions Regarding Functional Improvement and Neurological Recovery There is limited (Level 2) evidence that the improvement in disability seen in stroke rehabilitation cannot be explained on the basis of natural recovery of the neurological impairment alone. 6.2 Hemorrhagic vs. Ischemic Stroke Approximately 10% of all strokes are the result of intracerebral hemorrhage (ICH) (Kelly et al. 2003, Paolucci et al. 2003). The proportion of patients presenting with hemorrhagic stroke tends to be higher in Eastern European and Asian countries where the prevalence of untreated hypertension may be higher. (Kalra & Langhorne, 2007). Although primary ICH has been associated with more severe neurologic impairment and higher mortality in the acute phase (up to one half of patients with primary ICH die within the first month), it is generally believed that patients with ICH have a better recovery compared to patients with ischemic strokes. (Table 6.2). Table 6.2 Functional Outcomes of Patients with Hemorrhagic vs. Ischemic Stroke Author, Year Country Pedro Score Jørgensen et al. 1995 Denmark No Score Methods 1,000 unselected patients with acute stroke of a verified type in the Copenhagen Stoke Study were included. Logistic regression was used to examine the influence of stroke type on outcome, adjusting for age, sex, initial stroke severity, comorbidity and recurrent stroke during hospital stay. Neurological deficits and functional disabilities were evaluated weekly from the time of acute admission throughout the rehabilitation period. Outcomes Eighty-eight (9%) of the sample had intracerebral hemorrhage. The relative frequency of intracerebral hemorrhage rose exponentially with increasing stroke severity. In multivariate analyses, stroke type had no influence on mortality, neurological outcome, functional outcome, or the time course of recovery. Initial stroke severity was the all-important prognostic factor. The relative importance of hypertension and blood pressure on admission was not greater for intracerebral hemorrhage than for infarction. Kelly et al. 2003 USA No Score Paolucci et al. 2003 Italy No Score Data from 1064 patients (871 with cerebral infarction, 193 with ICH) admitted for inpatient rehabilitation was examined. FIM admission, discharge and change scores were compared between stroke types. A case-control study of 270 inpatients admitted for inpatient rehabilitation. Patients with sequelae of first stroke were matched for stroke severity, basal disability, age (within 1 year), sex, and onset admission interval (within 3 days). The 2 groups' length of stay, Barthel Index (BI), Rivermead Mobility Index (RMI) and Canadian Neurological Scale (CNS) scores were compared, as well as Total admission FIM score were higher in patients with infarction compared with ICH (59 vs. 51, P=.0001). There was no difference in total discharge FIM score between groups (82.3 ischemic vs. 79.1 ICH, p=0.200). Patients with ICH gained more FIM points during rehabilitation (28 vs. 23.3; P=.002). The ICH patients with the most severely disabling strokes had significantly greater recovery than cerebral infarction patients with stroke of similar severity. Compared with ischemic patients, hemorrhagic patients had significantly higher CNS (7.3 vs. 6.7) and RMI scores (6.6 vs. 5.4) at discharge and higher BI (50 vs. 40) and RMI (38 vs. 29) efficiency scores). Hemorrhagic patients showed a probability of a high therapeutic response on the BI that was approximately 2.5 times greater than that of ischemic patients. 6. The Elements of Stroke Rehabilitation pg. 5 of 54

Andersen et al. 2009 Denmark No Score Katrak et al. 2009 Australia No Score Chiu et al. 2010 USA No Score efficiency (gain in score/los) of BI and RMI. Stroke patients with hemorrhagic (ICH) and ischemic strokes were compared with regard to stroke severity and mortality using patients in a national registry of all hospitalized stroke patients (n=39,484) in Denmark, which collected data from 2001-2007. The patients underwent an evaluation including stroke severity using the Scandinavian Stroke Scale (SSS). Patients were followed-up from admission until death or censoring in 2007. Independent predictors of death were identified by means of a survival model based on 25,123 individuals with a complete data set. 718 consecutive stroke admissions (589 CI and 129 ICH) admitted for inpatient rehabilitation over 9.5 years was studied. The following outcomes were compared between groups: FIM gain, FIM efficiency, Motor Assessment Scale (MAS) change, gait velocity, and discharge destination. Data from the Glycine Antagonist in Neuroprotection (GAIN) Americas trial were used to assess the relationship between stroke type (ischemic vs. hemorrhage) and long-term outcome. 1604 non-obtunded patients with acute stroke were treated within 6 hours of symptom onset irrespective of hemorrhagic (N = 237) versus ischemic (N = 1367) subtype. Multiple logistic regression analysis was performed to evaluate predictors of mortality and neurologic outcome (modified Rankin scale [mrs] score of 0-1 vs. 2-6 at 3 months) adjusting for baseline National Institutes of Health Stroke Scale score, stroke risk factors, clinical and demographic characteristics, and treatment group. 3,993 (10.1%) of the sample had ICH. Stroke severity was almost linearly related to the probability of having ICH (2% in patients with the mildest stroke and 30% in those with the most severe strokes). Compared with ischemic strokes, ICH was associated with an overall higher mortality risk (Hazard Ratio: 1.564; 95% CI, 1.4-1.7). The increased risk was, however, time-dependent; initially, risk was 4-fold, after 1 week it was 2.5-fold, and after 3 weeks it was 1.5-fold. After 3 months stroke type did not correlate to mortality. Patients with ICH were more severely disabled on admission (mean FIM 74.3 vs. 88, p<0.001). ICH patients also had longer median LOS (35 vs. 26 days, p<0.001). By discharge there was no difference in the mean FIM scores between groups (106 vs. 102.8, p=n/s); however ICH patients had gained significantly more FIM points and their FIM efficiency was higher (0.80 vs. 0.61, p<0.001). Stroke type remained a significant explanatory factor for FIM gain, after adjusting for admission FIM, length of stay, age, and days from stroke onset to rehabilitation admission. The majority of patients in both groups went home at discharge. ICH significantly increased the odds of a poor neurological outcome (odds ratio 1.94, 95% CI: 1.23-3.06) and was independently associated with a mean 0.25-point increase in the 3-month mrs score (P =.04). ICH had no effect on mortality compared with ischemic stroke (odds ratio 1.01, 95% CI: 0.68-1.49) after adjusting for initial stroke severity and other baseline characteristics. Discussion Jørgensen et al. (1995) found that, after controlling for potential confounders that stroke type (ischemic vs. hemorrhagic) did not influence mortality, the time course of neurological recovery, neurological outcome or the time course of recovery from disability. The apparent effect of poorer outcome among patients with hemorrhagic stroke was due to initial greater stroke severity. Andersen et al. (2009) also reported an initial increased risk of mortality associated with ICH, although the effect was time dependent, disappearing after 3 months. Using data from the Glycine Antagonist in Neuroprotection (GAIN) Americans, Chiu et al. (2010) also examined whether the poor long-term prognosis associated with hemorrhagic stroke could be explained by initial stroke severity. They reported that ICH was an independent predictor of poor neurologic outcome, nearly doubling the odds of long-term disability. 6. The Elements of Stroke Rehabilitation pg. 6 of 54

However, ICH was not associated with higher mortality compared with ischemic stroke after adjusting for initial stroke severity and other baseline characteristics. Paolucci et al. (2003) matched patients on the basis of initial stroke severity, age, sex and onset to admission time and reported that ICH patients had superior rehabilitation outcomes and demonstrated a higher therapeutic response on ADL. ICH patients had higher Canadian Neurological Scale scores, Rivermead Mobility scores as well as greater efficiencies in scores. Lengths of hospital stays were similar between the groups. The authors attribute the greater gains, relative to patients with ischemic strokes to better neurological recovery associated with resolving brain compression. Kelly et al. (2003) reported similar results. Although patients with ICH admitted to a rehabilitation hospital had significantly lower FIM scores compared to patients with ischemic stroke, there were no differences in discharge FIM scores between the groups. Patients with ICH had higher FIM change scores. The same pattern, whereby patients with ICH had lower admission FIM scores but gained the same number of FIM points was also reported by Katrak et al. (2009). Although initial disability did not significantly predict the amount of recovery during rehabilitation, initial stroke severity was a strong predictor of functional status at discharge. Lipson et al. (2005) also found no significant differences in discharge FIM scores between ICH and ischemic stroke patients; however, there was no difference in the mean FIM scores on admission despite the fact that ICH patients were admitted for rehabilitation significantly later compared with ischemic stroke patients. Conclusions Regarding Hemorrhagic versus Ischemic Stroke There is conflicting (Level 4) evidence that patients suffering from hemorrhagic stroke have worse long-term outcomes compared with patients with ischemic stroke 6.3 Elements of a Stroke Unit Associated with Improved Outcome 6.3.1 What Form of Stroke Unit is Best? Specialized stroke rehabilitation units are associated with better outcomes, compared with mixed rehabilitation wards, general medicine, and mobile stroke teams.: The Stroke Unit Trialists Collaboration (SUTC) systematic review (2013) has described the hierarchical service organization in stroke care, moving along a continuum from more organised to less organised care: 1. Stroke ward: Wards where a multidisciplinary team including specialist nursing staff based in a discrete ward cares exclusively for stroke patients. This category included the following subdivisions: a) Acute stroke units These units accept patients acutely but discharge early (usually within seven days). These units are further subcategorised into: i) intensive model of care with continuous monitoring, high nurse staffing levels and the potential for life support ii) semi-intensive with continuous monitoring high nurse staffing but no life support facilities 6. The Elements of Stroke Rehabilitation pg. 7 of 54

iii) non-intensive with no high nurse staffing or life support facilities b) Rehabilitation stroke units Patients are accepted after a period of usually of five to seven days or more, and focus on rehabilitation c) Comprehensive (ie combined acute and rehabilitation) These are stroke units that accept patients acutely but also provide rehabilitation for at least several weeks if necessary. Both the rehabilitation unit and comprehensive unit models offer prolonged periods of rehabilitation. 2. Mixed rehabilitation ward: where a multidisciplinary team including specialist nursing staff in a ward provides a generic rehabilitation service but not exclusively caring for stroke patients. 3. Mobile stroke team: where a peripatetic multidisciplinary team (excluding specialist nursing staff ) provides care in a variety of settings. 4. General medical ward: where care is provided in an acute medical or neurology ward without routine multidisciplinary input The SUTC (2013) reported that more-organised care was consistently associated with improved outcomes, with decreased mortality, institutionalised care and dependency. Based on 21 trials, stroke unit care showed reductions in the odds of death recorded at final (median one year) follow-up (OR 0.87, 95% CI 0.69 to 0.94; P = 0.005), the odds of death or institutionalised care (OR 0.78, 95% CI 0.68 to 0.89; P = 0.0003) and the odds of death or dependency (OR 0.79, 95% CI 0.68 to 0.90; P = 0.0007) compared to care provided on a general medical ward. Outcomes were independent of patient age, sex, initial stroke severity or stroke type, and appeared to be better in stroke units based in a discrete ward. There was no indication that organised stroke unit care resulted in a longer hospital stay. Subgroup analyses in the SUTC (2013) indicate that the observed benefits of organised stroke unit care are not limited to any one models of stroke unit organisation that were examined. Comprehensive units and mixed assessment/ rehabilitation units, tended to be more effective than care in a general medical ward. There were also trends towards better outcomes within the dedicated stroke rehabilitation ward setting as opposed to the mixed rehabilitation ward, and within the acute (semi-intensive) ward as opposed to the comprehensive ward. Foley et al (2007) compared three models of stroke care (acute, rehabilitation and comprehensive units) and found that all models stroke units were associated with significant reductions in mortality, combined death and dependency and length of stay. However not every model was associated with equal benefit. (Table 6.3) Table 6.3: Mortality and dependency rates for different models of stroke Models of stroke care Mortality OR Death/dependency OR (95% CI) (95% CI) Acute stroke care 0.80 (0.61-1.03) 0.70 (0.56-86) Combined acute and rehabilitation 0.71 (0.54-0.94) 0.50 (0.39-0.65) 6. The Elements of Stroke Rehabilitation pg. 8 of 54

Post acute rehabilitation 0.60 (0.44-0.81) 0.63 (0.48-0.83) Overall 0.71 (0.60-0.83) 0.62 (0.53-0.71) Further analyses in the SUTC (2013) indicate that the observed benefits of organised stroke unit care are not limited to any of the subgroup of patients. Apparent benefits were seen in people of both sexes, aged under and over 75 years, with ischemic or hemorrhagic stroke and across a range of stroke severities. Similarly, Langhorne et al (2013) noted that there whilst stroke unit care reduced death or dependency (RR 0.81; 95% CI:0.47 0.92; P=0.0009; I2=60%) there were no difference in benefits for stroke patients with intracerebral hemorrhage (RR, 0.79; 95% CI, 0.61 1.00) when compared to patients with ischemic stroke (RR, 0.82; 95% CI, 0.70 0.97;P interaction =0.77). 6.4 What Elements of Care are Associated with Improved Outcomes? Why specialized stroke unit care improves patient outcomes remains unclear. It is likely that the processes of care and the structures that support these processes contribute to their success; however, the issue is complex. In the case of stroke rehabilitation, the unit of study is broad and involves the examination of complex care delivery systems. Furthermore, comparisons of studies, which appear to provide similar interventions, can be quite different. Several features associated with organised stroke unit care have been identified to contribute to the better outcomes: Co-ordinated multidisciplinary staff Regularly scheduled meetings Routine involvement of carers Staff specialization Standardized and early assessments Better diagnostic procedures Early mobilization Prevention of complications Better application of best-evidence" Attention to secondary prevention measures Evans et al. (2001) suggested specific components of acute stroke care that might be associated with decreased mortality and dependence, including: thrombolysis, physiological homeostasis, anticoagulation among patients with atrial fibrillation, early aspirin use and early mobilization. Processes of care were evaluated between a dedicated stroke unit, which included both acute and rehabilitative services and less organized stroke team, located on a general medical ward. Within the first seven days of admission, patients on the stroke unit were more closely monitored neurologically. A greater percentage of patients received oxygen therapy, nasogastric feeding and measures to prevent aspiration. Within the first four weeks of stroke, a greater percentage of stroke unit patients received a formal bedside swallowing assessment, a social work and occupational therapy assessment within 7 days, written evidence of rehabilitation goals and discharge/rehabilitation plans (Evans et al. 2001-see Table 6.4). Although both groups were comprehensively assessed and investigated, greater attention was paid to evaluations of consciousness, swallowing and communication among patients treated on the 6. The Elements of Stroke Rehabilitation pg. 9 of 54

stroke unit. Medical complications were more common among patients admitted to the general medical ward and appeared to be the factor most strongly associated with improved outcome among patients receiving care on the stroke unit. However, to what extent this factor and other unidentified factors contributed to the better outcome is unknown. Table 6.4 Differences in the Processes of Care Between a Stroke Unit and a Stroke Team (Evans et al. 2001). Process of Care Odds Ratio (95% CI) Associated with the Benefit of Stroke Unit Care Formal swallowing assessment with 72 hours 3.3 (2.0-5.4) Occupational therapy assessment with 72 hours 2.4 (1.4-4.0) Social work assessment within 72 hours 2.8 (1.1-6.9) Attention to secondary prevention 2.0 (1.2-3.3) Assessment of caregiver skill needs 11.3 (6.6-19.2) Written rehabilitation goals 2.5 (1.3-4.8) Indredavik et al. (1999) also found aggressive medical management including the use of intravenous saline solutions, oxygen therapy, heparin and Paracetamol to reduce fever was more frequent among patients managed on a stroke unit, compared to treatment received by patients on a general medical ward. Early mobilization was the most significant factor associated with discharge home at six weeks, although it remains unclear whether the benefit resulted from a decrease in medical complications such as deep vein thrombosis and pneumonia, or was due to positive psychological benefits. There were no differences in either the total mean hours of both occupational and physical therapy the groups received, which further highlights the intangible elements of a stroke unit that could account for the better outcomes. In a retrospective study, Ang et al. (2003) reported that patients treated within the integrated stroke unit had a shorter LOS and better functional outcome. The authors speculated that the main reasons for the improved outcome was due to seamless nature of care since patients did not have to be physically transferred to a different facility or wait have to wait for a bed to become available, before intensive rehabilitation therapies could begin. However, the report contained insufficient detail of the interventions provided within the two groups to assess the differences in care processes, which may have been responsible for the observed differences. Strasser et al. (2008) investigated the role of education within existing interdisciplinary rehabilitation units. A multiphase, staff training program compared training and information provision delivered over six-month period with information provision only. The group that received additional training discharged patients with a significantly greater gain in mean motor FIM score (+13.6). The authors speculated that the intervention taught the necessary skills and provided a useful conceptual model to positively impact on team dynamics. Commenting on the changes in stroke care at an Auckland Hospital between 1996 (prior to the establishment of a stroke unit) and 2001, following the establishment of a mobile stroke team, Barber et al. (2004) reported that while there were changes in the processes of stroke care since the implementation of the new stroke services, that there had been no corresponding decrease in mortality (14% in 2001 vs. 17% in 1996). However, greater proportions of patients were treated with aspirin within 24 hours of admission, and were discharged on anticoagulation therapy. Only 24% of patients were kept nil by mouth for 24 hours, compared to 46% in 1996. 6. The Elements of Stroke Rehabilitation pg. 10 of 54

Rudd et al. (2005), using data from the 2001-2002 National Stroke Audit (including England, Wales and Northern Ireland) evaluated the organization, processes of care and outcomes for stroke. The authors found that better processes of care, were more frequently associated with stroke unit care and decreased the risk of death considerably. The risk of death for patients who received care on a stroke unit was estimated to be 75% that of the risk for those receiving treatment on a non-stroke unit. While all of the above mentioned studies have focused on identifying the individual contributions of a variety of therapies or interventions, associated with a good outcome, Wade (2001) warns against the risk of committing a type III error (falsely rejecting the experimental hypothesis of the interactive effects of complex interventions are not considered), by pursuing such a course. He suggests that attempting to deconstruct the elements of specialized stroke rehabilitation therapies, in an effort to establish which isolated component(s) are effective may be flawed, by failing to recognise the interdisciplinary and complementary nature of the stroke rehabilitation. Ballinger et al. (1999) concluded that the types and duration of therapies provided by 13 physical and occupational therapists at four rehabilitation facilities treating stroke patients were heterogeneous and varied between institutions and individuals. It can be difficult to realize the same benefits associated with processes of care from clinical trials when they are translated into usual practice. As Kalra & Langhorne (2007) noted, most stroke units evolve in response to local patients needs, priorities and service arrangements, which may not be replicated in other settings. One of the elements of stoke unit care that has been associated with improved outcome is the prevention of complications. Complications are known to be common following acute stroke. Indredavik et al. (2008) followed 489 acute stroke patients who were admitted to a comprehensive stroke unit and subsequently enrolled in an early supported discharge service. Despite the benefit of the best model of care, medical complications were still common. During the first seven days following stroke, 64% of patients experienced at least one complication. The most common complications were pain, elevated temperature, stroke progression and urinary tract infections. Increasing stroke severity, advancing age and female gender were the strongest predictors of complications. Sorbello et al. (2009) also reported a high frequency of medical complications during the acute period following stroke, with or without early mobilization. 82% of patients experienced at least one complication, the most common being falls and urinary tract infections. These findings suggest that some complications experienced following stroke are difficult or impossible to prevent. Furthermore, it suggests that complications may not impact on stroke outcome as much as previously believed. In contrast to this finding, Govan et al. (2007), using a subset of data from the SUTC, found that patients receiving specialized stroke care had a lower incidence of chest infections, other infections and pressure sores. The prevention and treatment of complications was believed to be a contributing factor in improved outcomes. While it is tempting to try to reduce the gains that patients achieve while on stroke rehabilitation units to the sum of their parts, Whyte and Hart (2003) identified some factors which contribute to the difficulties encountered in attempting to unveil the effective elements of stroke rehabilitation: The broad range of treatments provided, as well as the poor definitions of treatments often described in published reports, means that reproducibility and dissemination of proven therapies may be difficult. 6. The Elements of Stroke Rehabilitation pg. 11 of 54

The intensities of treatments provided and the composition of therapy can vary across studies, even when evaluating similar therapies. The importance of patient participation, motivation and engagement is difficult to capture and can influence the result, again, when other factors may be constant between studies. Variations between individual therapists can occur as therapists respond to the responses and cues from patients they are treating. This effect can also result in subtle differences between like therapies and affect the study result. The therapist effect which refers to non-specific treatment effects brought about by the therapists personality, verbal communication skill or degree of warmth and empathy. Conclusions Regarding the Components of Stroke Units Which Contributes to Improved Outcomes The term 'stroke unit' is broad and describes hierarchal service organisation in stroke care along a continuum from more organised to less organised care. There is strong (Level 1a) evidence that individuals with stroke are more likely to survive, return home and regain independence if they received organised inpatient stroke unit care. Many elements contribute to the success of stroke rehabilitation unit. Although improved outcomes have been reported among trials evaluating stroke units, no causal mechanism(s) has been identified and verified. There is moderate (Level 1b) evidence that patients treated on a stroke unit have fewer medical complications. 6.5 Impact of Care Pathways and Guidelines Integrated Care Pathways (ICP) has been recently introduced in an attempt to improve the quality and consistency of stroke rehabilitation care. They have been seen as a means to translate the recommendations from national guidelines to a local setting. In some centres, ICPs have been developed to reduce lengths of hospital stay in an effort to reduce costs. ICPs can also be referred to as care mapping (Falconer et al. 1993). The definition of a care pathway may vary from one institution to another, although there are several common elements and include: being patient focused, the management is evidence-based, is multidisciplinary, documents in detail the clinical process and is constructed in a manner that facilitates an audit of outcomes (Edwards et al. 2004). However, the development and successful implementation of an ICP is time consuming and expensive and raises concerns over their associated opportunity costs. Sulch et al. (2000) described the development of an integrated care pathway as an organized, goaldefined and time management plan that has the potential of facilitating timely interdisciplinary coordination, improving discharge planning and reducing length of hospital stay. Other, less formal systems may include checklists of processes of care (Cadilhac et al. 2004). Kwan et al. (2007) suggested that the development of care pathways might be more appropriate for acute stroke management where they have the greatest potential to alter the highly complex processes of care, rather than in the 6. The Elements of Stroke Rehabilitation pg. 12 of 54

rehabilitative phase of stroke when well coordinated service is usually provided by an interdisciplinary team. Table 6.5 The Impact of Care Pathways Author, Year Country Pedro Score Hamrin et al. 1990 Sweden 4 (RCT) Falconer et al. 1993 USA 5 (RCT) Sulch et al. 2000 UK 6 (RCT) Duncan et al. 2002 USA No Score Sulch et al. 2002 UK 6 (RCT) Sulch et al. 2002 UK 6 (RCT) Cadilhac et al. 2004 Australia No Score Methods 280 acute stroke patients were assigned to be managed by a systematized care procedure with written care plans in accordance with the nursing process model or to be managed according to conventional care. 121 patients were randomized to either the Critical Path Method (CPM), which functions like an interdisciplinary team in the planning, monitoring and control of patient s program and progress, or to usual care (multidisciplinary rehabilitation team), which served as the control. 152 patients were randomized to be managed by an Integrated Care Pathway (ICP) based on evidence of best practice, professional standards and existing infrastructure for facilitating inter-disciplinary coordination, improving discharge planning and reducing length of hospital stay or were to be managed by conventional multi-disciplinary care (control). (see Figure) 288 stroke patients in 11 Dept. of Veteran Affairs Medical centres, hospitalized between 1998 and 1999 were followed prospectively for 6 mos. The degree of compliance with the AHCPR guidelines was correlated with FIM scores. Additional analyses from Sulch et al. 2000. Quality of life was assessed using the EuroQoL Visual Analogue Scale (EQ-VAS) at 6 mos. Additional analyses from Sulch et al. 2000, investigating the frequency of stroke specific assessments associated with either ICP or multidisciplinary care. Prospective, multicenter, single-blinded evaluating the following models of care: stroke units, mobile services, and conventional care. 21 selected processes of care (PoC) were identified and abstracted from medical records. The PoCs reflected aspects of care received within 24 hours of admission, documentation and general management, and included CT scan, swallowing assessment, allied health assessment, documentation of premorbid functioning and Outcomes Functional improvements within each group were equal and no significant differences between the two groups were found on any of the outcome measures. Of the 80 subjects who completed satisfaction questionnaires, general satisfaction scores at discharge were significantly lower in the CPM group compared to the control group. There were no differences between the groups in motor or cognitive functions at discharge and no differences in the average length of stay or hospital charges. There were no differences in mortality rates, frequency of institutionalization or LOS between the two groups. Conventional multidisciplinary care resulted in higher BI scores between 4 and 12 weeks and higher Quality of Life scores at 12 weeks and 6 months, compared to the ICP group patients. Average compliance scores in acute and post acute settings were 68.2% and 69.5%, respectively. Greater levels of adherence to the rehabilitation guidelines were associated with higher FIM scores. Patients receiving conventional multidisciplinary therapy had significantly higher QoL scores at 6 mos compared to patients in Integrated Care Pathway group (median 72 vs. 63, p<0.005). Increased frequency of stroke-related assessments with ICP, including testing for inattention (84% vs. 60%, p=0.015) and nutritional assessments (89% vs. 70%, p=0.024). Early discharge notifications to general practitioners were also higher among patients in the ICP group. Adherence to key PoC was higher in SUs than in other models. For all patients, adherence to PoC was associated with improved mortality at discharge and trends found with independence at home.468 of 1701 eligible patients participated. (28%) Overall PoC adherence rates for individual care models were SU 75%, mobile service 65%, and conventional care 52% (P<0.001). The adjusted odds of participants being alive at discharge if adhering to all or all but 1 PoC was significant (OR 3.63; 95% CI: 1.04 to 12.66; 6. The Elements of Stroke Rehabilitation pg. 13 of 54

Middleton et al. 2011 Australia 8 (RCT) Quality in Acute Stroke Care (QASC) study Panella et al. 2012 Italy 6 (RCT) discharge needs, measures to avoid aspiration, deep vein thrombosis and use of antiplatelet agents at discharge related to care models and participant outcomes. Data were collected at acute hospitalization (median 9 days) and at medians of 8 and 28 weeks after stroke. 1,696 stroke patients admitted to an acute stroke unit (ASU) within 48 hours of onset of symptoms were randomized to care on an ASU with standardized treatment protocols to manage fever, hyperglycaemia, and swallowing dysfunction with multidisciplinary team building workshops or to a control ASUs received only an abridged version of existing guidelines. The primary outcomes were 90- day death or dependency (modified Rankin scale [mrs] >/=2), Barthel index, and SF-36 physical and mental component summary scores. 14 hospitals were randomized to a clinical pathway (CP) group or to a non intervention/usual care (UC) arm. Healthcare workers in the CP arm received 3 days of training in quality improvement of CPs and in use of a standardized package including information on evidence-based key interventions and indicators. Healthcare workers in the usual-care arm followed their standard procedures. The primary outcome was 30-day mortality. 7-day mortality, LOS, readmission and institutionalization rates after discharge, dependency levels, and complication rates were also assessed. P=0.043). Important trends at 28 weeks were found for being at home (aor 3.09; 95% CI: 0.96 to 9.87; P=0.058) and independent (aor 2.61; 95% CI: 0.96 to 7.10; P=0.061), with complete PoC adherence. 19 ASUs were randomly assigned to intervention (n=10) or control (n=9). Intervention ASU patients were significantly less likely to be dead or dependent at 90 days (42% vs. 58%, p=0.002. Patients in the intervention ASU also had a better SF-36 mean physical component summary score (p=0.002). There were no significant differences between groups in mean BI scores or SF-36 mental component summary scores and no differences in 30-day mortality between groups. There was no difference in the 30-day mortality between groups (CP: 7.6% vs. UC: 10.5%, p=0.34). Compared with patients in the UC group, the patients in the CP arm had a significantly lower risk of mortality at 7 days (OR = 0.10; 95% CI 0.01 to 0.95) and significantly lower rates of adverse functional outcomes, expressed as the odds of not returning to pre-stroke functioning in their daily life (OR = 0.42; 95 CI 0.18 to 0.98). There were no differences between groups on the other outcomes. An updated Cochrane review (Kwan and Sandercock 2009 ) which included 3 randomized and 12 nonrandomized trials, reported no significant difference between care pathway and control groups in terms of death or alter the eventual discharge destination. In fact, patients managed with a care pathway were more likely to be dependent at discharge (P =0.04); less likely to suffer a urinary tract infection [(OR) 0.51, 95% (CI) 0.34 to 0.79], less likely to be readmitted (OR 0.11, 95% CI 0.03 to 0.39); and more likely to have neuroimaging (OR 2.42, 95% CI 1.12 to 5.25). Evidence from randomised trials suggested that patients' satisfaction and quality of life were significantly lower in the care pathway group (P = 0.02 and P < 0.005 respectively). This finding was confirmed by Hoeing et al. (2002) who found the structure of care (systematic organization, staffing expertise and technological sophistication was not associated with better functional outcomes whereas interestingly, compliance with AHCPR post stroke rehabilitation guidelines improved those same outcomes. The apparent paradox may signify the importance of using evidence or guidelines to assist rehabilitation clinicians in individualizing the rehabilitation of stroke patients as opposed to a one size fits all approach. In contrast, a more recent cluster-randomized controlled trial (Panella et al. 2012) reported that an evidence-based care pathway that was developed with the input from a multidisciplinary team resulted in a reduction in the odds of 7-day mortality, and increased the odds of return to pre-stroke function. Another cluster randomized controlled trial (Middleton et al. 2011) demonstrated that patients who 6. The Elements of Stroke Rehabilitation pg. 14 of 54

were allocated to a stroke unit with standardized evidence-based nursing protocols designed to improve the management of dysphagia, hyperglycemia and fever had better outcomes compared to patients who were randomized to a stroke unit without similar protocol. Sulch et al. (2000, 2002) randomized 152 stroke patients to a rehabilitation program of integrated care pathways (ICP), characterized as an organized, goal-defined and time managed plan with the potential to improve discharge planning and reduce length of hospital stay, or to a conventional multidisciplinary team (MDT) program of conventional rehabilitation. Patients receiving MDT care improved significantly faster between weeks 4 and 12 (median change in Barthel Index 6 vs. 2, p<0.01) and had higher Quality of Life scores, assessed by the EuroQol Visual Analogue Scale (EQ-VAS) at 6 months (72 vs. 63, p<0.005). Although intuitively care pathways should improve the quality of stroke care, the evidence does not support this conclusion. It may be that care pathways simply reinforce rather than change practice. It may also suggest that imposing a blueprint of care, rather than individualizing treatment, does not improve outcomes. Therefore, although organized interdisciplinary stroke rehabilitation units have been shown to improve outcomes, care pathways do not appear to be a contributing component to their success. There is evidence that the use of care pathways may actually be associated with poorer patient satisfaction and quality of life. In contrast, there appears to be strong evidence that adherence to clinical guidelines, which involves application of evidence-based practices at an individual patient level, does improve outcomes. Conclusions Regarding the Impact of Care Pathways Based on the results from three RCTs, there is conflicting (Level 4) evidence to suggest that care pathways improve stroke rehabilitation outcomes. There is moderate (Level 1b) evidence that care pathways do not reduce hospital costs or decrease length of hospital stays. There is strong (Level 1a) evidence that compliance with stroke rehabilitation guidelines and adherence to processes of care result in improved outcomes. There is insufficient evidence to support routine implementation of care pathways for acute stroke management or stroke rehabilitation as care pathways do not improve stroke rehabilitation outcomes or reduce costs. However, compliance with stroke rehabilitation guidelines improves outcomes. 6.6 Timing of Stroke Rehabilitation The results of several studies (Feigenson et al. 1977, Hayes and Carrol 1986, Wertz 1990) have suggested stroke rehabilitation should be initiated soon after stroke to achieve optimal results. In their review, Cifu and Stewart (1999) reported that there were four studies of moderate quality that demonstrated a positive correlation between early onset of rehabilitation interventions following stroke and improved functional outcomes (Table 6.6). These authors noted that Overall, the available literature demonstrates that early onset of rehabilitation interventions within 3 to 30 days post stroke is strongly associated with improved functional outcome. Ottenbacher and Jannell (1993) conducted a meta-analysis including 36 studies with 3,717 stroke survivors, and demonstrated a positive correlation between early intervention of rehabilitation and improved functional outcome. 6. The Elements of Stroke Rehabilitation pg. 15 of 54

Maulden et al. (2005) reported on the findings of the Post-Stroke Rehabilitation Outcomes Project (PSROP), an observational, prospective study, which enrolled 1,291 patients from six inpatients rehabilitation facilities in the US. Increases in the length of time from stroke onset to admission to rehabilitation were associated with lower discharge FIM scores and increased LOS for patients with both moderate and severe strokes. Days from stroke onset to admission was also a significant predictor of discharge total FIM score, discharge motor FIM score, discharge mobility FIM score and rehabilitation LOS in regression analysis. The strongest relationship between early admission to rehabilitation and improved functional outcome was among the most severely impaired patients (case-mix group 108-114). However, a literature review by Diserens et al. (2006) examining the potential benefits of early mobilization concluded that no randomized controlled trial had been conducted to enable a comparison of the effects of early (defined as within the first three days of stroke) vs. delayed (greater than 3 days). Individual Studies Several non-randomized trials investigating the effects of timing of rehabilitation interventions were identified (see Table 6.6). Table 6.6 Timing of Stroke Rehabilitation Author, Year Country PEDro Score Paolucci et al. 2000 Italy No Score Yagura et al. 2003 Japan No Score Musicco et al. 2003 Italy No Score Salter et al. 2006 Canada No Score Methods A case controlled study of 135 stroke patients who received: 1) rehabilitation within the first 20 days post-stroke (short onset) 2) rehabilitation 21 to 40 days post-stroke (medium onset) and rehabilitation 41 to 60 days (long onset) post-stroke. All patients received the same physical therapy program. 1056 patients were divided into 3 groups based on the time from stroke onset to admission to rehabilitation: Grp 1 <90 days; Grp 2 91-180 days; Grp 3 >180 days. 1,716 stroke patients, with moderate to severe disability, consecutively admitted to 20 Italian hospital rehabilitation centers in 1997 and 1998 were followed. Three negative patient outcomes were considered: death, early failure (premature, unwanted interruption of rehabilitation program; absence of any improvement at hospital discharge), and late failure in terms of severe disability (Barthel Index score <40) or poor quality of life (Medical Outcomes Study 36- Item Short-Form Health Survey; questionnaire score <80) 6 months after admission. 435 patients admitted to an inpatient stroke rehabilitation program within 150 days of a first unilateral stroke were studied. Patients were grouped according to time from onset of stroke to admission to the rehabilitation unit: 0 Outcomes Higher dropout rate was noted in the short onset group. Barthel Index scores in the short onset group showed significantly greater rate of improvement than the other two groups. Walking status improved in 71% of non-ambulatory patients in Grp 1, 55% in Grp 2 and 44% in Grp 3.ADLs improved in 67% of totally dependent patients in Grp 1 and 50% in Grps 2 and 3. 74% of patients in Grp 1 were discharged home compared to 60% in Grps 2 and 3. The presence of dementia and pressure ulcers on admission was associated with worse outcomes (OR for any failure or death=1.31; 95% confidence interval [CI], 1.03-1.67; OR=1.63; 95% CI, 1.12-2.37, respectively). Patients who initiated the rehabilitative procedures early (within 7d after stroke) had better long-term outcomes than did those who initiated the rehabilitation after more than 1 month (OR=2.12; 95% CI, 1.35-3.34) or from 15 to 30 days after the acute cerebrovascular event (OR=2.11; 95% CI, 1.37-3.26). FIM scores at admission and discharge as well as FIM change and FIM efficiency were significantly higher for early admission than for delayed admission patients. LOS was significantly longer among delayed admission patients. There were significant differences on admission 6. The Elements of Stroke Rehabilitation pg. 16 of 54

Gagnon et al. 2006 Canada No Score Huang et al. 2009 Taiwan No Score Bernhardt et al. 2008 Australia 8 (RCT) Sorbello et al. 2009 Australia 7 (RCT) Hu et al. 2010 Taiwan No Score Langhorne et al. 2010 Scotland 15, 16-30, 31-60, 61-90 and 91-150 days. and discharge FIM scores, FIM change and LOS between groups of patients admitted 0 15 and 16-30 days post stroke as well as between patients admitted from 16 30 days and 31-60 days. However, no significant differences in FIM admission, FIM discharge, FIM change scores or LOS were noted between patients admitted from 31 60 and 61-90 days or 61 90 and 91-150 days post stroke. 418 stroke patients, matched on age, gender and degree of stroke severity were retrospectively analyzed according to the time from stroke onset to admission to a stroke rehabilitation unit. Three cohorts were assembled; onset < 20 days (short), onset 20-40 days (moderate), and onset > 40 days (long). Admission, discharge FIM, FIM efficiency, FIM change and LOS were assessed. The charts of 76 patients admitted for inpatient rehabilitation following first-ever stroke were reviewed. Associations with, and predictors of, BI gains were examined at 1, 3, 6 and 12 months. 71 patients within 24 hrs of stroke onset were randomly assigned to receive standard care (SC) (n=33) or SC plus very early mobilization (VEM) (n=38). In the VEM group the goal of first was mobilization within 24 hours of stroke symptom onset. VEM continued daily for the first 14 days after stroke or until discharge (whichever was sooner) and was delivered by a nurse/physiotherapist. It was intended to provide twice the therapy compared with the SC group. The primary safety outcome was the number of deaths at 3 months. A good outcome, defined as a modified Rankin Score (mrs) of 0-2 at 3, 6 and 12 months was also assessed. Additional analysis from AVERT trial, examining the number and severity of complications between groups. Complications were classified according to type (complications of immobility, stroke-related, co-morbidity related, psychological and other) and severity. 154 patients admitted to a stroke intensive care unit were studied prospectively to examine the effects of timing of admission and intensity of therapy on outcome using regression analyses. 32 acute stroke patients (stroke onset < 36 hrs) were randomized to 1 of 4 nurse-led treatment protocols: (a) standard stroke unit care, (b) There were no significant differences between the groups (short, moderate long) on any of the variables assessed. LOS was similar (51 vs. 48 vs. 47 days), as was FIM efficiency (0.53 vs. 0.62 vs. 0.53). Time from stroke onset to rehabilitation was negatively associated with BI gains at 1 mos and 1 yr. There was no significant difference in the number of deaths between groups (SC, 3 of 33; VEM, 8 of 38; p=0.20). Almost all deaths occurred in patients with severe stroke. After adjusting for age, baseline NIHSS score and premorbid mrs score, the odds of experiencing a good outcome were significantly higher at 12 months for the VEM group (OR: 8.15, 95% CI 1.61-41.2, p<0.01). There was also a trend towards good outcome at 3 months, but not at 6 months. Within 3 months of admission there were no differences in the number or severity of complications between groups. Patients in the SC group experienced a total of 91 counts of complications while patients in the VEM group experienced 87. There were no differences in the severity of complications between groups. Stroke severity (assessed using the NIHSS scale), age, rehab intensity, admission BI score and onset to admission were all significant predictors of increases in BI scores, explaining 64% of the variability in model. Starting rehab one day sooner resulted in a 0.65 point increase in BI score. Timing of rehab was not a significant predictor an independent walking at the time of discharge, but was a predictor of BI score at discharge. With respect to the comparison of EM vs. control, the EM patients were significantly less likely to experience a medical complication resulting from immobility, after 6. The Elements of Stroke Rehabilitation pg. 17 of 54

7 (RCT) early mobilization (EM) where patients were encouraged to sit, stand or walk within 24 hrs of stroke and to repeat at least 4 x/day, (c) automated physiological monitoring (AM) or (d) combined EM and AM. Cumming et al. Additional analysis from AVERT trial. 2011 Australia 7 (RCT) Diserens et al. 2011 Switzerland 7 (RCT) Sundseth et al. 2012 Norway 7 (RCT) Bai et al. 2012 China 4 (RCT) 50 acute ischemic stroke patients with NIHSS score >6 were randomized to an early protocol in which patients were mobilized out of bed after 52 hours or to delayed protocol where patients were mobilized after seven days. All patients were treated with physiotherapy immediately after their admission The primary outcome was development of a serious complication during hospitalization. Secondary outcomes included minor complications, which did not increase length of stay significantly, % of patients with Rankin scores of 0-2 at 3 months and changes in cerebral blood flow. 65 patients admitted to a stroke unit within 24 hours after stroke were assigned to either VEM within 24 hours of admittance or mobilization between 24 and 48 hours (control group). Primary outcome was the proportion of poor outcome (modified Rankin scale score, 3-6). Secondary outcomes were death rate, change in neurological impairment (National Institutes of Health Stroke Scale score), and dependency (Barthel Index 0-17). 364 patients with hemorrhagic stroke were randomized to receive a standardized, threestage rehabilitation program or to a control group that was treated with standard hospital ward, internal medical intervention. Patients in the experimental group began therapy within 24 hours of admission. Therapy was provided for 45 min/day x 5 days/week. During the first month. Therapy continued in hospital and the community for 6 months. Outcomes assessed at baseline, 1, 3 and 6 months included the simplified Fugl-Myer Assessment Scale (FMA) and Modified Barthel Index (MBI). correcting for age, baseline NIH score and cointerventions, within the first 5 days of stroke. There was no longer a significant difference between groups at the 3-month assessment point. Patients in the very early and intensive mobilization group returned to walking significantly sooner than did standard stroke unit care controls (P=0.032; median 3.5 vs. 7.0 days). There were no differences in proportions of patients who were independent on the BI (score of 20) or who had achieved a good outcome on the Rivermead Motor Assessment Scale (score of 10-13) at either 3 or 12 months. VEM group assignment was a significant, independent predictor of independence on the BI at 3 months, but not at 6 months. VEM group assignment was a significant, independent predictor of good outcome on RMA at both 3 and 12 months. Eight patients were in the delayed group were transferred to other hospitals and were not included in the analysis. There were 2 (8%) severe complications in the 25 early mobilization patients and 8 (47%) in the 17 delayed mobilization patients (P < 0.006). There were no differences in the total number of complications or in any of the other outcomes. There were no differences in blood flow in the 26 patients (62%) who underwent serial transcranial Doppler ultrasonography. The median time to first mobilization was 13 hrs for patients in the VEM group and 33 hrs for patients in the control group. At 3 months, the odds of a poor outcome, death and dependency were non-significantly higher in the control group after adjusting for age and stroke severity: (OR, 2.70; 95% CI, 0.78-9.34; P=0.12; OR, 5.26; 95% CI, 0.84-32.88; P=0.08, and OR, 1.25; 95% CI, 0.36-4.34; P=0.73, respectively). Patients in both groups improved over the study period, although patients in the experimental group achieved significantly greater gains on both FMA and MBI scores (p<0.05). The greatest improvement was observed in the first month post stroke. 6. The Elements of Stroke Rehabilitation pg. 18 of 54

Although there is a strong correlation between early admission to stroke rehabilitation and improved functional outcomes, demonstrated in both individual studies and based on the results of meta-analysis, this relationship may not be one of cause and effect. Patients with severe strokes (and higher levels of impairment) are also more likely to experience medical complications, or be too impaired to participate in rehabilitation that may delay their admission to a stroke rehabilitation unit. In contrast, mild to moderate stroke patients with fewer medical complicating factors are more likely to be admitted sooner than later to a stroke rehab unit. The results from the individual studies are difficult to compare owing to the variations in the cohorts. The studies authored by Paolucci et al. (2000) and Gagnon et al. (2006) were similar with respect to their categorization of time from stroke onset, although the results were conflicting. Patients in the early onset cohort showed a greater rate of recovery in one study (Paolucci et al. 2000), but not in the other (Gagnon et al. 2006). Yagura et al. (2004) examined the differences in ambulation and ADL status between three groups of patients, divided according to the duration of time from onset of symptoms to stroke rehabilitation admission and reported that patients who were admitted within 90 days of their stroke achieved greater gains in ambulation, upper extremity and ADL function, compared to patients who had been admitted either 91-180 days or >180 days following stroke. However, while patients who were admitted earlier achieved better outcomes, all patients significantly benefited from rehabilitation regardless of their onset to admission time. Shah et al. (1990) found that interval between stroke onset and admission to rehabilitation was a predictor of achievement of rehabilitation potential among 258 patients recovering from first-ever stroke. A shorter onset time was associated with improved functional outcome. Similarly, Salter et al. (2006) found that early admission to rehabilitation was associated with improvements in ADL ability as measured by the FIM instrument, after controlling for the effects of patients age. The AVERT trial includes 3 publications Bernhardt et al (2008), Sorbello et al. (2009) and Cumming et al. (2011). In the first presentation of results, there were no reported differences between the early mobilization and conventional groups in death or the number and severity of complications 3 months. The authors attributed the lack of statistically significant between study groups to a small sample size. In the second publication, (Sorbello et al. 2009), there was no reported difference between groups in the frequency of medical complications between groups. Finally, in a last publication, Cumming et al. (2011) reported that patients randomized to the early group were able to walk 50 m unassisted, sooner were discharged from hospital slightly earlier (median 6 vs. 7 days) and a greater percentage were discharged home (32% vs. 24%). Additional results from AVERT phase II showed that the early mobilization protocol resulted in delivery of more and earlier therapy (van Wijk et al. 2012). Schedule (hours to first mobilization, dose per day, frequency and session duration) and nature (percentage out-of-bed activity) of therapy differed significantly between the VEM and standard care (SC) groups. Mobilization was earlier, happened on average 3 times per day in those receiving VEM, with the higher median proportion of out-of-bed activity in VEM session (85.5%) compared to median 42.5% in the standard care (van Wijk et al. 2012). In addition to VEM being effective in improving mobility and independence, economic evaluation suggests that VEM is potentially cost effective (Tay-Teo et al, 2008). Patients receiving VEM incurred significantly less costs at 3 months compared to standard care; and the difference in mean per patient total cost persisted at the 12-month assessment (Tay-Teo et al 2008). 6. The Elements of Stroke Rehabilitation pg. 19 of 54

The benefit of early mobilization was also investigated in an observer-blinded, pilot randomised controlled trial studying the key aspects of early stroke unit care. In the Very Early Rehabilitation or Intensive Telemetry After Stroke [VERITAS], early mobilization and intensive monitoring was incorporated within a 2x2 factorial study design (Langhorne et al. 2010). The early mobilization intervention arm utilized protocol based on the AVERT trial with respect to the timing, nature, and frequency of the intervention. (Langhorne et al 2010). Degree of mobilization activity, defined as the mean time spent upright per working day, was 61 (SD, 54) minutes in the early mobilization group compared with 42 (SD, 57) minutes with standard care. By Day 5, 74% of patients in the early mobilization group were independent in walking, compared to 44% of patients with standard care. There were also a trend to more patients on the early mobilization protocol being independent at 3 months (Langhorne et al. 2010). Although there was no difference in good outcome at 3 months (defined as modified Rankin Scale score of 0-2) between groups, the odds of medical complications were lower among subjects in the early mobilization group (OR 0.1, 95% CI: 0.001 to 0.9, p=0.04), after adjusting for age and severity of stroke. Pooling of individual patients' data from AVERT and VERITAS showed that time to first mobilization from symptom onset was significantly shorter among VEM patients (median, 21 hours; IQR: 15.8-27.8 hours) compared with standard care patients (median, 31 hours; IQR: 23.0-41.2 hours) [Craig et al 2010]. VEM patients had significantly greater odds of independence at 3 months compared with standard care patients (adjusted OR, 3.11; 95%CI 1.03-9.33) [Craig et al. 2010]. Diserens et al. (2012) also investigated whether early mobilization was safe and effective in preventing serious complications. While patients in the early group had a lower occurrence of severe complications the sample size was small and there were loses in the standard care group. Comparisons with other studies of early mobilization are difficult to make since the criteria used to define early were different. In contrast, Sundseth et al. (2012) provided very clear criteria for defining VEM. A large number of dropouts, particularly in the VEM group, resulted in a non significant trend towards a reduction in poor outcome for VEM patients. The impact of timing to rehabilitation on overall functional outcome was not limited to patients transferred from acute care to rehabilitation. Early initiation of rehabilitation predicted better functional outcome on patients with severe strokes in a stroke intensive care unit with a multidisciplinary stroke care team (Hu et al. 2010). After adjusting for stroke severity and age, patients who started earlier rehabilitation had higher Barthel Index scores at discharge. Commencing rehabilitation one day earlier in the stroke ICU resulted in an increase of the Barthel Index score by 0.65 points (Hu et al. 2010). There is increasing evidence that early mobilization, however it is defined, is not associated with negative effects and may be better for patients compared with delayed mobilization. Fears that hypoperfusion may be associated with early mobilizations, leading to worse outcomes appear to be unfounded. Conclusion Regarding the Timing of Stroke Rehabilitation There is limited (Level 2) evidence that early admission to stroke rehabilitation is associated with improved functional outcomes. There is strong (Level 1a) evidence that very early mobilization following stroke helps to reduce medical complications. 6. The Elements of Stroke Rehabilitation pg. 20 of 54

There is moderate (Level 1b) evidence that very early mobilization following stroke helps to decrease the time to achieve functional walking. Stroke patients who are appropriate candidates should be admitted to a rehabilitation unit or facility as soon as possible. 6.7 Intensity of Therapy 6.7.1 Intensity of Physiotherapy and Occupational Therapy When attempting to determine factors that contribute to the improved functional outcomes that are associated with specialized stroke rehabilitation, the intensity of rehabilitation therapies is often cited as an important element. Do patients who receive therapy for longer periods of time or at a higher level of intensity realize greater benefits compared to patients who receive conventional care? This hypothesis has been investigated extensively although these studies have found that intensity of therapy was only weakly correlated with improved functional outcome. However, Kalra and Langhorne (2007) have noted that there is evidence from neuroimaging studies showing that increased intensity of rehabilitation therapies results in greater activation of areas associated with the function towards which this therapy is directed. Definition of Intensity The definition of intensity or 'dosage' has been an unresolved issue in studies investigating the doseresponse relationship in rehabilitation therapies (Kwakkel 2006). Restrictions in measuring energy expenditure as a measure of activity intensity have resulted in estimates of therapy intensity in rehabilitation, measures such as the number of repetitions (frequency), the overall time spent in therapy or frequency of treatment sessions (Kwakkel 2006). While a universally accepted definition of the term intensity does not exist, it is usually defined as number of minutes per day of therapy or the number of hours of consecutive therapy. Studies evaluating the effects of increased intensity of therapy usually provide more therapy over a given course of total treatment time compared to the alternative, which receive a lesser amount. This weak association may be explained by differences in the time, duration and composition of therapies provided and/or the characteristics of the stroke patients under study. The American Congress of Rehabilitation Medicine Stroke Movement Interventions Subcommittee have recommended operational definitions for concepts that are common in stroke motor rehabilitation (Page et al. 2012). The recommendations focused on those terms regarded as often mislabelled, terms such as intensity and dosing (duration and frequency). The recommended definition for intensity is "the amount of physical or mental work put forth by the client during a particular movement or series of movements, exercise or activity during a defined period of time" (Page et al. 2012). Duration is defined as "length of time during which a single session is administered" and frequency is defined as "how often during a fixed period the regimen is administered" (Page et al. 2012). In addition, the delivery method and window of therapy have been identified as areas for further refinement (Page at al. 2012). The total amount of time that a patient spends engaged in rehabilitation activities vary considerably, between units, institutions and countries. Lincoln et al. (1996) observed that patients on a stroke 6. The Elements of Stroke Rehabilitation pg. 21 of 54

rehabilitation unit were engaged in interactive behaviours for only 25% of their time. De Weerdt et al. (2000) used behavioural mapping to quantify the amount of time patients spent in therapeutic activities on two rehabilitation units, one in Belgium and one in Switzerland. Patients were engaged in rehabilitation for a larger percentage of the day than those from Switzerland (45% vs. 27%). De Wit et al. (2005) also observed significant differences in the amount of time patients spent in rehabilitation activities among four European countries (Belgium, UK, Switzerland and Germany) Patients from Germany spent a larger percentage of the day in therapy time (23.4%), while those from the UK spent the least (10.1%). Therapy time ranged from 1 hour per day in the UK to about 3 hours per day in Switzerland. In all of the units, patients spent 72% of their time in non-therapeutic activities. Even more discouraging are the results from A Very Early Rehabilitation Trial (AVERT) (Bernhardrt et al. 2004, 2007) in which a cohort of 58 patients in 5 acute stroke units in Australia were observed. Patients engaged in moderate or high levels of activity for only 12.8% of their therapeutic day. 53% of the time, patients spent their time in bed and were alone 60% of the time. Although there was a direct relationship between stroke severity and activity, even patients with only mild stroke spent only 11% of their active day walking. Patients affected upper limbs were observed to be moving only 33% of the time, regardless of whether the patient was with a therapist or alone. A comparison between Australian patients and those in Norway (Bernhardt et al. 2008) revealed that patients admitted to acute stroke units in Trondheim spent an average of 21% less time in bed and 10% more time engaged in either sitting out of bed or in standing/walking activities compared with patients in Melbourne hospitals. There were differences between these two systems in terms of staffing ratios, policies and in the rehabilitation programs themselves. Duncan et al (2005) reviewed all RCTs and meta-analyses published to date examining the effect of intensity on improved functional outcome and concluded that there was weak evidence of a doseresponse relationship. The authors suggest that all subsets of patients may not benefit equally and could not recommend specific guidelines about the intensity or duration of rehabilitation therapies. 6.7.2 Previous Reviews and Meta-Analyses The results of four meta-analyses, suggest that increased intensity of therapy is beneficial (See Table 6.7). Langhorne et al. (1996) examining the effects of differing intensities of physical therapy showed significant improvements in activities of daily, living (ADL) function and reduction of impairments with higher intensities of treatment. Kwakkel et al. (1997) included 8 RCTs and one non-randomized experiment and found a small but statistically significant intensity-effect on ADL and functional outcome parameters. However, Cifu and Stewart (1999) identified only 3 moderate quality studies and one metaanalysis which examined the intensity of rehabilitation services and functional outcome) and reported that the intensity of rehabilitation services was only weakly associated with improved functional outcomes after stroke. Table 6.7 Reviews and Meta-Analyses Investigating the Intensity of Physiotherapy and Occupational Therapy in Stroke Rehabilitation Langhorne et al. (1996) Cifu and Stewart (1999) Kwakkel et al. (1997) Stern et al. 1970 Smith et al. 1981 Peacock et al. 1982 Sivenius et al. 1985 Kwakkel et al. 1997 Richards et al. 1993 Nugent et al. 1994 Kramer et al. 1997 Stern et al. 1970 Smith et al. 1981 Peacock et al. 1982 Sivenius et al. 1985 6. The Elements of Stroke Rehabilitation pg. 22 of 54

Sunderland et al. 1992 Wade et al. 1992 Richards et al. 1993 Keith et al. 1995 Werner 1996 Sunderland et al. 1992 Wade et al. 1992 Richards et al. 1993 Kwakkel et al. (2004) conducted an extension of previous meta-analysis and evaluated the benefit of augmented physical therapy which included 20 studies on several interventions: occupational (upper extremity), physiotherapy (lower extremity), leisure therapy, home care and sensorimotor training. After adjusting for differences in treatment intensity contrasts, augmented therapy was associated with statistically significant treatment effects for the outcomes of ADL and walking speed, although not for upper extremity therapy assessed using the Action Research Arm test. A 16-hour increase in therapy time during the first six-months following stroke was associated with a favourable outcome. Chen et al. (2002) examined the relationship between intensity of therapy and functional gains in a retrospective study of 20 sub acute rehabilitation facilities in the USA. Stroke patients made larger selfcare gains if they had lower self-care, higher mobility and cognition function at admission, longer, uninterrupted stays, received more intensive therapies and weren t admitted to a rehabilitation facility initially. Determinants of improvement in mobility included younger age, admission soon after impairment, higher self-care and cognition measures. Although admission function, length of stay and therapy intensity collectively contributed to greater functional gains, length of stay and therapy intensity did not always predict those gains. There was an interdependency between the domains of self-care, mobility and cognition, such that patients with deficits in self-care on admission made the greatest improvements if mobility or cognitively remained intact or relatively intact. Wodchis et al. (2005) studied a large cohort of stroke survivors (n=23,824) admitted to skilled nursing facilities in Ohio, Michigan and Ontario. For patients with an uncertain prognosis on admission the intensity of rehabilitation therapies was positively associated with an increased likelihood of going home. However, it should be noted that the weekly therapy time would not generally be considered to be intensive (The maximum category was 500+ min/week). Galvin et al. (2008) examined the effect of increased duration of exercise therapy on functional recovery post stroke. The results of the meta-analysis which included the results from 10 studies demonstrated that increased duration of exercise therapy time had a small but positive effect on activities of daily living as measured by the Barthel Index. The improvements were maintained over a 6-month period. Cooke et al. (2010) included the results from 9 RCTs representing 7 individual studies examining varying doses of the same exercise-based interventions. The authors meta-analyzed the studies on the basis of individual outcomes (ARAT scores, Motricity Index, handgrip strength, and comfortable walking speed) at the end of treatment and at follow-up. Most of the analyses contained the results from only 2-3 studies. Some small, but statistically significant treatment effects were reported. The authors concluded that there was some, but limited support in favour of greater therapy intensity. While the weight of evidence suggests that more intensive therapy is associated with greater rehabilitation gains, the European Stroke Organisation s 2008 Guidelines for the Management of Ischemic Stroke and Transient Ischemic Attack (Quinn et al. 2009) have indicated that at present available data do not allow for recommendations on minimal or maximum therapy times. 6. The Elements of Stroke Rehabilitation pg. 23 of 54

Nevertheless, the current standard of care at publicly-funded stroke rehabilitation facilities in the United States is to provide at least 3 hours of therapy. Failure to do so results in a loss of reimbursement (Conroy et al. 2009). Many trials have evaluated the efficacy of increased intensity of therapy and the relationship to improved functional outcomes. The results are presented in Table 6.8 Table 6.8 Studies of Stroke Rehabilitation and Therapy Intensities Author, Year Country PEDro Score Smith et al. 1981 Australia and USA 5 (RCT) Sivenius et al. 1985 Finland 5 (RCT) Sunderland et al. 1992 UK 6 (RCT) Richards et al. 1993 Canada 6 (Single-blind RCT) Kalra et al. 1994(a) UK 5 (RCT) Methods 133 recovering stroke patients randomized to receive intensive therapy, conventional therapy or no therapy. 95 consecutive stroke patients able to tolerate an intensive rehabilitation program were randomized at one-week following stroke to either an intensive physiotherapy program on a stroke unit or a control group receiving conventional physiotherapy on a general medical unit. 132 patients randomized to receive either conventional therapy (control) or enhanced arm therapy which included Bobath exercises, EMG biofeedback, micro-computer games and goal setting. 27 patients randomized to receive one of three therapies: 1) Early intensive therapy incorporating the use of a tilt table, resisted exercises and treadmill, beginning ~8 days post stroke, for 1.7 hrs/day x 5 weeks (experimental); 2) Early conventional therapy included traditional approach with therapy beginning ~9 days post stroke, for 1.8 hrs/day x 5 wks (control 1); or 3) Conventional therapy beginning 13 days post stroke, 0.72 hours/day x 5 wks (control 2). Analysis of 146 middle-band stroke patients taken from a sample of 245 stroke patients randomized at 2 weeks post stroke to a rehabilitation unit or a general medical unit after stratification by stroke severity. (Analysis of 1993 RCT). Outcomes Patients in the intensive and conventional therapy groups had greater improvement in their ADL scores compared to those who did not receive any therapy at 6 months. At 12 months, patients who had received intensive rehabilitation had significantly higher ADL scores when compared to those who received no therapy, whose scores had declined. Patients receiving intensive physiotherapy had significantly improved their level of ADL and mobility at 3, 6 and 12 months. The greatest gains were achieved in the first 3 months. Severe sub-group: No significant differences between the groups. Mild sub-group: Patients receiving enhanced therapy scored higher on the Extended Motricity Index, the 9-hole peg test and on the Motor Club Assessment than those receiving conventional therapy. At week 6, gait speed in the 2 control groups was similar and lower than the experimental group. By months 3 and 6, the gait speed between all groups was similar. The median Barthel Index (BI) scores of patients managed on the stroke unit were significantly higher when compared to patients on the medical unit (15 vs. 12). The rate of improvement in BI scores was faster for patients on the stroke unit and these patients had significantly shorter LOS (6 vs. 20 weeks). Significant gains were achieved at a faster rate without additional physiotherapy or occupational therapy in total. 6. The Elements of Stroke Rehabilitation pg. 24 of 54

Nugent et al. 1994 Australia No Score Sunderland et al. 1994 UK 6 (RCT) Werner and Kessler 1996 USA 6 (RCT) Logan et al. 1997 UK 8 (RCT) Feys et al. 1998 Belgium 6 (RCT) Kwakkel et al. 1999 Netherlands 8 (RCT) Lincoln et al. 1999 UK 7 (RCT) Ruff et al. 1999 USA 3 (quasi-rct) Parry et al. 1999 UK 7 (RCT) 54 stroke patients received a weight-bearing exercise (WBE) program that continued until patients scored 6 on the Motor Assessment Scale (MAS) for walking or until discharge. Follow-up of 97 patients participating in the 1992 study. 49 stroke patients were randomized to receive either intensive 12 week rehabilitation therapy for 1 hour a day for 4 days a week, or to no rehabilitation therapy (control). 111 stroke patients who were randomized to receive either enhanced services or received usual services. Patients in the enhance therapy group were seen significantly sooner following discharge (9 vs. 22.5 days), received more visits (6 vs. 2.5) and more therapy time (240 vs. 85 min) than patients receiving usual service. A multicentre trial of 100 patients (2 to 5 weeks after stroke onset) received either additional sensorimotor treatment 30 min x 5 days/wk x 6 wks (treatment) or a placebo treatment (control). 101 patients were randomized 14 days following stroke to receive one of 3 therapies: 1) arm training, 2) leg training or 3) basic rehabilitation only. Leg and arm treatments were applied for 30 min 5 days/week x 20 weeks. All patients received basic rehabilitation. 282 patients randomized to received one of 3 therapy groups: 1) routine physiotherapy (RPT), 2) routine therapy with additional PT from a qualified therapist (QPT) or 3) routine therapy with additional PT from a physiotherapist assistant (APT). The intervention groups (QPT, APT) received 10 hrs of additional physiotherapy over 5 weeks. 113 stroke patients received either therapy 6 days/wk or received therapy 7 days/wk. OT/PT/SLP interventions. Sub-group analysis of Lincoln et al. 1999 study. Significant overall Pearson correlation (r=.45) between the number of WBE repetitions and change in MAS scores. At one-year follow-up there were no significant differences between the 2 groups on any of the outcome measures (Extended Motricity Index, 9-hole peg test or Frenchay Activities Index). Significant improvements of FIM-motor scores were noted at 3 months for the treatment group; however no significant differences in FIM-MM gains were observed between the 2 groups from 3 to 9 months. Significant decline in SIP scores was observed in the treatment group at 3 months. At 3 months, the patients in the enhanced therapy group had higher EADL scores compared to patients receiving usual care. However, at 6 months, there were no differences between the two groups on any of the outcome measures. There were no differences in Fugl-Meyer (FM) scores between the groups at 6 wks. Patients in the experimental group scored higher on the FM at 6 and 12 months follow-up than the placebo treatment. No significant differences between the groups were observed in Action Research Arm (ARA) or Barthel Index (BI) scores. Patients in both groups improved significantly over time in FM, ARA and BI scores. At week 26, significant differences in median Action Research arm (ARA) scores between the three groups were observed. Median Barthel Index and ARA scores of patients in both arm and leg training groups were significantly higher when compared to the control group. No significant differences between the groups on any of the outcome measures post intervention were observed at 3 or 6 months (Rivermead Motor Assessment Arm scale, Barthel Index, Extended ADL index or Action Research Arm Test). No significant difference in gains achieved in any of the FIM components between the two groups. No significant difference between the two groups in length of hospital stay. For severe patients (score of 0 or 1 on RMA arm scale): No significant differences on any outcome measures between the groups following intervention at 6 mos. For less severe patients (score of 2 or more on RMA arm scale): No significant differences on any outcome measures between the groups. 6. The Elements of Stroke Rehabilitation pg. 25 of 54

Partridge et al. 2000 UK 8 (RCT) Kwakkel et al. 2002 Netherlands 8 (RCT) Pohl et al. 2002 Germany 6 (RCT) Slade et al. 2002 UK 7 (RCT) Sullivan 2002 USA 5 (RCT) Di Lauro et al. 2003 Italy 7 (RCT) Fang et al. 2003 China 8 (RCT) 114 patients randomized to receive either 30 minutes or 60 minutes of physiotherapy daily for an unspecified period of time. 6, 9 and 12 month follow up to Kwakkel et al. 1999. 60 ambulatory patients with hemiparesis, for at least 4 weeks, caused by right or left supratentorial ischemic stroke or intracerebral hemorrhage, exhibiting impaired gait were randomized into 1 of 3 groups: structured speeddependent treadmill training (STT) where increased in walking speed with each session; limited progressive treadmill training (LTT) in which training speed was increased by no more than 5% of the maximum initial walking speed each week; and, conventional gait therapy (CGT) involving physiotherapeutic gait therapy based on latest PNF and Bobath concepts. Patients were evaluated before training, 2 and 4 weeks post-training on over ground walking speed, cadence, stride length and FAC. 141 patients (62% with stroke) were randomized to receive conventional occupational therapy and physiotherapy or enhanced therapy (67% more) during in-patient rehabilitation. 24 patients with unilateral stroke within the MCA or basilar artery distribution resulting in unilateral hemiparesis were randomized to receive body weight treadmill training (BWSTT) at slow, fast or variable speeds after patients were stratified by locomotor severity. Treatment was provided for 20 minutes for 12 sessions over 4 weeks. Self-selected over ground walking velocity (SSV) was assessed at onset, middle and end of training and 1 and 3 months later. 60 acute stroke patients with some disability were randomized to receive intensive, early rehabilitation, beginning 24 hrs post stroke (2 hrs therapy daily) (n=29) or regular acute rehabilitation (45 min therapy daily) (n=31) for 14 days. Both groups of patients went on to receive continued rehabilitation at a specialized centre for 60 days. Patients with first-onset stroke consecutively admitted to the stroke centre. One group (n = 78) received additional early physiotherapy (AEP) for 45 minutes, five days a week for four weeks starting within the first week since stroke onset; No differences in profiles of recovery scale scores were noted between the groups at 6 weeks or at 6 months. No significant differences between the groups were found from 6 months onward for: Barthel Index, Functional Ambulation Categories, Action Research Arm test, comfortable and maximal walking speed, Nottingham health profile part 1, sickness impact profile- 68 and Frenchay activities index. After 4 weeks training period, the STT group scored significantly higher than the LTT and CGT for all outcome measures. For patients receiving enhanced therapy, there was a significant 14 day reduction in length of stay and no difference in function outcome, measured by the Barthel Index. SSV significantly increased in all groups compared to baseline at 1 month. Significantly greater improvement in SSV was with fast training speeds compared to slow and variable groups. There were no significant differences in the NIH or BI scores of patients at either 14 days or 180 days following treatment. Patients from the AEP group had a high drop-out rate (n = 28), but those remaining made relatively better functional recovery at 30 days than those from the RT group measured by the Barthel Index. There was significantly greater improvement from admission to day 6. The Elements of Stroke Rehabilitation pg. 26 of 54

Rodgers et al. 2003 United Kingdom 7 (RCT) Sonoda et al. 2004 Japan (No Score) GAPS 2004 UK 7 (RCT) Keren et al. 2004 Israel No Score Jette et al. 2005 USA No Score the routine therapy (RT) group (n = 78) received no professional rehabilitation therapy. 123 patients with stroke causing upper limb impairment within the previous 10 days were randomized to either an experimental group or into a control group. The experimental group received stroke unit care plus enhanced upper limb therapy from both a physiotherapist and an occupational therapist commencing within 10 days of stroke and available up to 30 minutes/day, five days/week for 6 weeks. The control group received stroke unit care. Historical comparison of 48 stroke patients treated admitted to a conventional stroke rehabilitation program in Dec 1999, compared to 58 patients treated by the Full-time Integrated Treatment (FIT) program. The key difference between the 2 programs was the intensity and frequency of treatment (80 minutes of OT/PT therapy 5x/week vs. same daily total of therapy time, but provided 7x/week, although patients were encouraged to remain active outside of structured sessions). A multisite (3 centre) single-blind randomized controlled trial (RCT) comparing the effects of augmented physiotherapy with normal physiotherapy on the recovery of mobility after stroke. 70 stroke patients admitted to hospital, able to tolerate and benefit from mobility rehabilitation were randomized to receive twice the amount of physiotherapy compared to the control group. Primary outcomes were mobility milestones (ability to stand, step and walk), Rivermead Mobility Index (RMI) and walking speed. Data were prospectively collected from 50 firstever stroke patients admitted for inpatient rehabilitation an average of 14 days post stroke. The relationship between functional outcome, measured by motor and cognitive FIM and the duration and intensity of therapy was examined. Retrospective analysis of 4,988 patients (993 with stroke), evaluating LOS and functional improvement. 30 demonstrated by Barthel Index scores, for patients in the AEP group compared to those in the RT group. There were no differences in other stroke outcomes such as motor function, neurological deficit severity, or psychiatric status. There was no significant difference between groups on any outcome measure (Action Research Arm Test, Motricity Index, Frenchay Arm Test, upper limb pain, Barthel ADL, Nottingham E-ADL) at 3 and 6 months after stroke. There was no significant difference in service costs between groups. Admission FIM scores between the 2 groups were similar (80.9, conventional vs. 81.2, FIT), however at discharge the FIT group had higher average FIM scores (97.1 vs. 105.0, p<0.01) and FIM efficiency, (change/los) (0.19 vs. 0.33, p<0.01). Hospital stays were also shorter for patients in the FIT group (72.9 vs. 81.1 days). The days of onset of stroke to admission into rehabilitation were 54 days for patients in the conventional group and 50 days for patients in the FIT group. The augmented therapy group received more direct time with therapists (62 vs. 35 min/day) and were more active (8.0% versus 4.8% time standing or walking) than normal therapy controls. There were trends favouring the augmented group in the outcomes of achievement to independent walking earlier, and higher RMI scores at three months. However, there was no significant difference in any other outcome. There was a correlation between duration of occupational therapy and gains in FIM motor and cognition scores, but no correlation with either physical therapy or SLP therapy. Predictors of improvement in motor FIM scores were younger age, admission FIM (both motor and cognitive) and receipt of SLP therapy. Predictors of improvement in cognitive FIM scores were admission FIM (cognitive), early admission to rehab and intensity of occupational therapy. Higher intensity of therapy was associated with shorter lengths of stay. Patients receiving < 1 hour of therapy/day had an average LOS of 21.4 (19.7-23.3) days, compared to those who received 1-1.5 hour/day (16.9;15.7-18.1) days or >1.5 hours/day (15.5;14.2-16.9) days. The odds of improving both independence in 6. The Elements of Stroke Rehabilitation pg. 27 of 54

Langhammer et al. 2007 Norway 6 (RCT) Langhammer et al. 2008 Norway 9 (RCT) Langhammer et al. 2009 Norway 8 (RCT) Huang et al. 2009 Taiwan No Score Hu et al. 2010 Taiwan No Score 75 stroke patients discharged from acute inpatient rehabilitation were randomized to receive an intensive outpatient exercise program emphasizing strength, balance and endurance or to a control group of regular exercise. Patients in the intensive group received physiotherapy during four periods with a minimum of 20 hours every third month. Treatment was carried out in home, at clinics or during inpatient rehabilitation. Subjects in the control group received follow-up treatment or rehabilitation only as required. Motor Assessment Scale (MAS, BI and grip strength measured with a vogorimeter (bars) were assessed at discharge from the acute ward and at 3,6,12 months. Additional outcomes reported from 2007 study. The primary outcome, assessed at admission, discharge and 3, 6, and 12 months post stroke onset was the Nottingham Health Profile (NHP). Additional outcomes reported from 2007 study. The main outcome was Instrumental Activities of Daily Living (IADL), motor function, gait performance, balance, grip strength, and muscle tone. Additional outcome measures included Motor Assessment Scale, 6-Minute Walk Test, Berg Balance Scale, Timed Up-and-Go Test, grip strength, Modified Ashworth Scale, and pulse monitoring. The patients were tested on admission, at discharge, and after 3, 6, and 12 months post stroke. The charts of 76 patients admitted for inpatient rehabilitation following first-ever stroke, were reviewed. Associations with, and predictors of BI gains were examined at 1, 3, 6 and 12 months. 154 patients admitted to a stroke intensive care unit were studied prospectively to examine the effects of timing of admission and intensity of therapy on outcome using regression analyses. mobility and functional independence were associated with increasing intensity of therapy. There were no differences between groups at any of the testing intervals. The mean differences from acute discharge to one year were: Intensive Group Regular Group MAS +2.5 +7.4 BI +5.3 +11.9 Grip strength 0.23 0.19 (paretic hand) There was a trend towards better Health-Related Quality of Life as measured by the NHP in the standard exercise group at 12 months (mean score: 16.2 vs. 17.9, p=0.05). On the 6 subscales, at 12 moths, there were no significant differences between the groups except in the area of physical mobility where patients in the standard group fared better (mean score: 17.9 vs. 36.6, p=0.04). Patients in both groups improved over time. The only significant difference between the groups was one component of the IADL scale, favouring the control group (ability to use the telephone independently). Time from stroke onset to rehabilitation was negatively associated with BI gains at 1 mos and 1 yr. Total OT and PT input (hrs) were positively associate with BI gains at 3 and 6 mos. Significant predictors of BI improvement at 1 yr were early rehab intervention, hemorrhagic stroke and lower BI score at admission. Stroke severity (assessed using the NIHSS scale), age, rehab intensity, admission BI score and onset to admission were all significant predictors of increases in BI scores, explaining 64% of the variability in model. An increase in 1 rehab session per day increased the odds of achieving a 2.8 increase in BI score, and a 27-fold increase in achieving independent ambulation at discharge. The effects appeared to be strongest for patients with severe vs. mild/moderate stroke. Askim et al. 2010 Norway 62 patients having suffered a stroke within the previous 14 days were randomized to receive Subjects in the treatment group received twice as much therapy compared with the control group (171 vs. 85 6. The Elements of Stroke Rehabilitation pg. 28 of 54

8 (RCT) either standard stroke unit care (including 2 daily sessions of therapy, 5 days/week) with early supported discharge or to an intensive therapy group which received the same care (SU + ESD) but with more intensive therapy: an additional 3 task-specific therapy sessions/week for 4 weeks following discharge from the SU and an additional session thereafter for the next 8 weeks. Each session lasted 30-50 minutes and additional training at home was encouraged. The primary outcome was Berg Balance Scale (BBS) score, assessed at 26 weeks. Additional outcomes included Motor Assessment Scale (MAS), BI and the Step test. Park et al. 2011 Korea 6 (RCT) Cauraugh et al. 2011 USA 6 (RCT) Van Wijk et al. 2012 Australia 8 (RCT) Discussion 25 subjects, an average of 28 months following stroke with hemiparesis who with a walking speed of <0.7 m/s were randomly assigned to either the experimental group or the control group, with 13 and 12 subjects, respectively. All subjects received a routine physical therapy in an inpatient setting. The subjects in the experimental group also received communitybased ambulation training (1 hr, 3X/week x 4 weeks). Outcomes were assessed before and after treatment. They included; 10-metre walk test, 6-minute walk test, community walk test, walking ability questionnaire and activitiesspecific balance. 18 patients with chronic stroke were randomized to receive one treatment protocol or 10. Patients in the single protocol group received 1 treatment session and those in the long-term practice group completed 10 treatment protocols over 16 months. All protocol sessions were 6 hours long (90 minutes 1 day/week/4 weeks) and were separated by 22 days. The treatment included bilateral arm movements and active neuromuscular stimulation. The primary outcome was the Box & Block test and was administered at baseline, at the end of one protocol, at 7 and at 16 months. Additional analysis from AVERT trial (Bernhardt et al. 2008) examining if patients who received early mobilization also received a greater amount of therapy compared to patients who did not. Adverse events at 3 months was the primary outcome. minutes/week). There were no statistically significant differences on any of the outcomes except a trend towards higher MAS scores (p=0.059) and gait speed (p=0.095) favouring the intensive group. At the end of the study period, subjects in the experimental group performed better the majority of outcomes compared with patients in the control group (p<0.05); 10-m walk test (mean decrease of 0.21 vs. 0.07 m/s), community walk test (mean increase of -13.61 vs. - 3.27 minutes), walking ability questionnaire (mean increase of 6.15 vs. 2.75 score) and activities-specific balance confidence scale (mean increase of 17.45 vs. 2.55 score). At the 16-month delayed retention test, when compared to the short-term group, the long-term group moved more blocks (43 vs. 32), and demonstrated faster premotor reaction times (158 vs. 208 ms). Patients in the VEM group received a greater dose of therapy (49 vs. 18 min/day, p<0.001), that was given more frequently (3.8 vs. 1 session/day, p<0.001) and they spent a greater percentage of the day out of bed (77% vs. 42%, p<0.001) compared with patients in the control group. There was no difference in the number of immobility-related adverse events between groups (18 vs. 22). 6. The Elements of Stroke Rehabilitation pg. 29 of 54

Fifteen good quality studies randomized patients to receive physical therapies at differing levels of intensity. The results from these studies are summarized in Table 6.9. Table 6.9 Summary of Functional Outcome Measures From RCTs with PEDro Scores 6 Evaluating the Intensity of Physiotherapy and Occupational Therapy Post-Stroke Author/Year PEDro Score Langhammer et al. 2007,08,09 9 Askim et al. 2010 8 Pohl et al. 2007 8 Partridge et al. 2000 8 Kwakkel et al. 1999 & 2002 8 Logan et al. 1997 8 Di Lauro et al. 2003 7 GAPS 2004 7 Lincoln et al. 1995, Parry et al. 1999 * 7 Slade et al. 2003 7 Sunderland et al. 1992 1994 (one-year follow-up) 6 * Richards et al. 1993 6 Werner and Kessler 1996 6 Feys et al. 1998 6 n at Randomization Intervention Outcome 75 Intensive outpatient physical therapy vs. standard care 62 Increased amount of therapy Motor Assessment Scale (-) Barthel Index (-) Nottingham Health Profile (-) Berg Balance scale (-) Berg Balance scale (-) Motor Assessment Scale (-) Barthel Index (-) 155 Gait trainer + PT vs. PT Walking independence (+) BI (+ at 4 weeks, - at 6 months) 114 Increased physiotherapy (60 vs. 30 min/day) 101 Arm training vs. leg training vs. control 111 Increased number of occupational therapy home visits 60 Intensive early rehabilitation 70 Regular PT vs. 2x regular PT 282 10 hours of additional physiotherapy over 5 weeks 141 67% increase in the amount of routine inpatient physio/occupational therapy 132 Enhanced upper extremity therapy 27 Early, intensive inpatient rehabilitation 49 Intensive 12 weeks of outpatient rehabilitation 100 Sensorimotor stimulation for 6 weeks Profiles of Recovery (- at 6 weeks and 6 mos) Barthel Index (- at weeks, 26 and 52) Walking ability (- at weeks 26 and 52) Dexterity (ARA) (+ at 26 weeks-arm training, - at week 52) Barthel Index (- at 3 and 6 mos) Nottingham Extended ADL (+ at 3 mos) (- at 6 mos) Barthel Index (-) NIH (-) Mobility measures (-) Rivermead Mobility Index (-) Walking speed (-) Barthel Index (- at 6 mos) Rivermead Motor Assessment Arm score (- at 6 mos) Length of stay (+) Barthel Index (-) Extended Motricity Index (+ at 6 mos) Motor Club Assessment (- at 1 yr) Gait velocity (+ at 6 weeks) (- at 3 and 6 mos) Change in FIM from 0-3 mos (+) Change in FIM from 3-9 mos (-) Brunnstrom-Fugl-Meyer test (- at 6 weeks) (+ at 3 and 6 mos) Barthel Index (-) 6. The Elements of Stroke Rehabilitation pg. 30 of 54

Pohl et al. 2002 6 Park et al. 2011 6 Cauraugh et al. 2011 6 60 Structured speeddependent treadmill training * Considered as a single study + Indicates a positive effect favouring treatment group - Indicates a non-significant difference between the intervention groups Action Research Arm test (-) walking speed, cadence, stride length and FAC at 4 weeks (+) 25 Community ambulation 10-m walk test (+) Community walk test (+) 18 1 vs. 16 treatment Box and block test (+) protocols The nature of specialized stroke rehabilitation services implies a greater intensity of therapy although this fact is not always documented. Several studies have attempted to determine the contribution of therapy intensity in stroke rehabilitation. However, illuminating the effect of greater intensity of therapy on functional outcome is difficult due to the variability of treatments provided, the timing and duration of their delivery and the outcomes that were assessed. Additionally, self-report of actual duration of therapy provided by physical therapists has been shown to be overestimated, compared with video recording (mean time 32 vs. 25 min) [Bagley et al. 2009]. Intensity of treatment is also dependent on the ability and the willingness on the part of the patient. The mechanism through which improved outcomes are achieved has not been well described. Fang et al. (2003) suggested that a program of greater intensity physiotherapy simply enabled patients to improve or achieve independence in ADL faster through compensation of the non-paretic limb, rather than necessarily through neurological improvements. Of the 15 good studies, many of them demonstrated a benefit on at least one testing but failed to demonstrate a difference when compared to conventional treatment at another point in time or among different stroke sub-types. Many of these studies demonstrated significant improvement on initial assessment; however, the benefit disappeared at a later date. The highest quality studies were associated with no benefit when compared to the control condition. The results of all of the three metaanalyses demonstrated a benefit of increased intensity of therapy; although the benefit was statistically significant, the clinical benefit was small. Conclusions Regarding the Intensity of Physiotherapy and Occupational Therapy There is strong (Level 1a) evidence that greater intensity of physiotherapy and occupational therapy results in improved functional outcomes. However, the overall beneficial effect is modest and the positive benefits associated with greater treatment intensities were not maintained over time. There is moderate (Level 1b) evidence, based on the results from a single study that the same therapies delivered more intensively over a shorter period of time results in faster recovery and earlier discharge from hospital. Greater intensities of physiotherapy and occupational therapy result in improved functional outcome. 6. The Elements of Stroke Rehabilitation pg. 31 of 54

6.7.3 Intensity of Aphasia Therapy Post Stroke The impact of the intensity of aphasia therapy post-stroke has also been studied. The most effective means of treating aphasia post stroke has yet to be determined, and studies investigating the efficacy of speech and language therapy for patients suffering aphasia post stroke have yielded conflicting results. One possible explanation for the observed heterogeneity of findings across studies is a difference in intensity of therapy. We have noted that the failure to identify a consistent benefit might have been due to the low intensity of speech-language therapy applied in the negative studies while higher intensities of therapy was present in positive studies (Poeck et al. 1989). Individual Studies of the Intensity of Language Therapy Post-Stroke The details of the studies investigating the intensity of aphasia therapy are presented in Table 6.10. Table 6.10 Intensity of Language Therapy of Aphasia Post Stroke Author, Year Country PEDro Score Meikle et al. 1979 UK 4 (RCT) David et al. 1982 UK 5 (RCT) Lincoln et al. 1984 6 (RCT) Shewan et al. 1984 Canada 5 (RCT) Methods 31 patients who had sustained a stroke 3 weeks prior and passed through the acute phase being left with disabling dysphasia were randomly assigned to 1 of 2 groups. One group received conventional speech therapy from a quality speech therapist while the other group received therapy from a nonprofessional volunteer. 155 aphasic stroke patients at 3 weeks post-stroke were randomised to receive either therapy from a speech language pathologist for 30 hours over 15 to 20 weeks or from an untrained volunteer providing support and encouragement for a similar time. 327 aphasic stroke patients who were able to cope with language testing assessment were randomised at 10 weeks post-stroke to receive 2, 1 hour therapy sessions per week at either a hospital or at home for 34 weeks or to receive no treatment. 100 aphasic stroke patients who failed to recover their language skills within the first 2 to 4 weeks post-stroke were randomized to one of 3 treatments: (1) language oriented therapy (LOT) provided by a speech language pathologist (SLP), (2) stimulation facilitation therapy (ST) provided by a SLP and (3) unstructured settings therapy (UNST) provided by nurses. Patients who did not want/were unable to participate formed a control group. Patients in each of the 3 treatment groups received 3, 1-hour sessions a week for 1 year. Outcome No significant differences were observed between the two groups on PICA scores. Patients in both groups showed improvement; however, no significant differences in FCP scores were noted between the groups. Patients in both groups demonstrated improvement; however, no significant differences in language recovery were noted between the groups on the Porch Index of Communicative Ability (PICA) Functional Communication Profile (FCP) and the Boston Diagnostic Aphasia (BDAE). No difference in Western Aphasia Battery (WAB) scores, its subsets Language Quotient (LQ) and Cortical Quotient (CQ) scores and Auditory Comprehension Test for Sentences scores between the groups. The LQ scores of patients in the treatment groups were significantly higher compared to the control groups. Individually, LOT and ST patients significantly improved compared to the control patients, but no significant differences were observed between the UNST and the control group. The CQ scores of the treatment groups were significantly higher compared to patients in the control group. Individually, ST patients had higher CQ scores than the controls but the LOT and UNST groups were not significantly different from the 6. The Elements of Stroke Rehabilitation pg. 32 of 54

Wertz et al. 1986 USA 6 (RCT) Hartman 1987 USA 6 (RCT) Brindley et al. 1989 UK 4 (RCT) Marshall et al. 1989 USA 5 (RCT) Poeck et al. 1989 Germany No Score (Cohort Study) 121 male veterans under the age of 75 years, between 2 to 4 weeks after onset of a single thromboembolic stroke with lesion confined to the left hemisphere and demonstrated language severity from 10 th -80 th percentile on PICA on entry into the study were randomized into one of three groups: (1) 8 to 10 hours a week of clinic treatment with speech therapy for 12 weeks followed by 12 weeks of no treatment; (2) 8 to 10 hours a week of home treatment by a trained volunteer for 12 weeks followed by no treatment; or (3) Treatment deferred for 12 weeks followed by 12 weeks of clinic treatment with a speech therapist. 60 right-handed patients with acute aphasia due to left hemisphere stroke were randomly assigned to receive 1 of 2 therapies for six months, beginning one month post-stroke: Conventional speech therapy provided by professional speech pathologists twice weekly or emotionally supportive counselling therapy, also provided by professional speech pathologists at the same intervals. Language function was measured by the Porch Index of Communicative Ability. 50 patients were also retested at 10 months post stroke. This study involved Broca s aphasic patients defined by the Boston Diagnostic Aphasic Examination without predominate apraxia who were 1 year post stroke. Two groups of 5 patients each received five hours of speech therapy for 5 days a week for 12 weeks. An intensive period of therapy was alternated with non-intensive period pre-course and a similar 12-week non-intensive period post-stroke. 121 males, 2 to 12 weeks onset of a single left hemisphere thrombosis infarct resulting in aphasia were randomized to receive: 1) home therapy treatment of aphasia given by wife, friend or relative, 2) treatment by Speech Language Pathologist or 3) treatment by Speech Language Pathologist deferred for 12 weeks. Therapy was provided for 8 to 10 hours a week for 12 weeks. 160 aphasic stroke patients with CT evidence of left hemisphere involvement only, beyond the acute stage of neurological illness were studied. Patients received intensive language treatment for 9 hr/week, for 6-8 weeks. These results were compared to those of a previous multicentre study of 92 German aphasic patients who did not receive language treatment. Patients were sub-grouped as early or late treated controls. After 1 st 12 weeks of treatment clinic treated patients performed significantly better than those deferred on the PICA. No significant difference were noted between home treated and clinic or between home treated and deferred treated patients. After 24 weeks of treatment there was no significant difference between any groups. No significant difference in the amount of improvement between the two groups. Significant improvement on Functional Communication Profile (FCP) details in movement, speech, reading, and overall score were noted during the intensive period. There was a significant ratio of improvement on FCP between intensive period and 2 nd non-intensive period in movement, speech and overall score. Language Assessment Remediation and Screening Procedure showed significant improvement in intensive period on sentence length increase, reduction in element omission, and increase in percentage of full utterances. At 12 weeks SLP group showed significantly more improvement than deferred group. Improvement noted in home treatment group did not differ from SLP group. At 24 weeks deferred treated group caught up to other 2 groups and no significant differences between groups was noted. In the early phase mean gains for each measure was significant for both treatment and control group on the token test and for repetition. About 2/3 of treatment patients showed a significant improvement in Aachen Aphasia Test. 6. The Elements of Stroke Rehabilitation pg. 33 of 54

Prins 1989 Amsterdam 5 (RCT) Bakheit et al. 2007 UK 7 (RCT) patients. 32 patients with aphasia for at least 3 months following a left hemispheric stroke were randomized to receive either systematic therapy (STAC) or conventional therapy (STIM). The STAC was comprised of a series of 28 different tasks on four levels: nonverbal, phonology, lexical-semantics and morphosyntax. The STAC group received treatment 2 times a week for 5 months. The STIM group received therapy during the same period of time with the same frequency as the STAC group. Patients were tested in two parts: subtests for auditory comprehension (items used as practice material in the STAC group) and 8 tests for auditory comprehension, reading comprehension and oral expression (items not used as practice material in either treatment group). Aphasic stroke patients were randomly allocated to one of 3 groups: 5 hours (intensive therapy group, n=51), 2 hours (standard therapy group n=46) while 19 patients were recruited for 2 hours per week of therapy and were treated by National Health Service (NHS) staff (NHS group). Subjects received speech and language therapy per week for 12 consecutive weeks starting as soon as practicable after the stroke. The Western Aphasia Battery (WAB) assessments were at baseline and 4, 8, 12 and 24 weeks after the start of therapy. No significant differences were noted between the groups on any of the test batteries. There were no results reported for the non-treatment control group. Patients in the intensive, standard and NHS received, on average 4.3, 1.6 and 0.57 hr/week of therapy. The improvement in aphasia was least in patients who were in the NHS group. The mean WAB scores at week 12 for the intensive, standard and NHS groups were 70.3, 66.2 and 58.1, respectively. There was no treatment effect of intensive therapy (p > 0.05), but there was a statistically significant difference between the standard study and the NHS groups (p = 0.002 at week 12 and 0.01 at week 24). The results from the studies in Table 6.10 are summarized in Table 6.11. Table 6.11 Efficacy and Intensity of Language Therapy on Aphasia Post Stroke Author, Year PEDro Score n at Randomization Intensity of Therapy Impact of Aphasia Therapy vs. Control Bakheit et al. 2007 7 116 36 hrs given over 38 - sessions in 12 weeks Lincoln et al. 1984 6 327 2, 1-hour sessions per - week for 34 weeks Wertz et al. 1986 6 121 8 to 10 hours a week for + 12 weeks Hartman 1987 6 60 2 times a week for 6 - months David et al. 1982 5 155 30 hours over 15 to 20 - weeks Shewan et al. 1984 5 100 3, 1-hour session a week + for 1 year Marshall et al. 1989 5 121 8 to 10 hours a week for + 12 weeks Prins et al. 1989 5 32 2 sessions a week for 5 - months Meikle et al. 1979 4 31 Minimum of 3 and - 6. The Elements of Stroke Rehabilitation pg. 34 of 54

maximum of 5 sessions a week for 45 minutes. Brindley et al. 1989 4 10 5 hours over 5 days a week for 12 weeks Poeck et al. 1989 Not Scored 160 9 hours a week for 6 to 8 weeks + +/- The most recent and largest RCT examined by Bakheit et al. (2007) failed to uncover a benefit of intensive aphasia therapy as assessed using the Western Aphasia Battery. The average length of stroke onset was one-month. The authors reported that the majority of patients receiving intensive treatment weren t able to tolerate it. Patients were either too ill or refused therapy and actually had lower WAB scores compared with patients who received less intensive, standard therapy (68.6 vs. 71.4). While this study was considered to be negative, patients who received an average of 1.6 hours of therapy (standard group) per week had significantly higher scores than those who received only.57 hours of therapy (NHS group). Patients in the highest intensity therapy group received an average of 4 hours of therapy per week. Therefore, depending on how intensive is defined, this trial could be considered positive. Bhogal et al. (2003) observed that a significant treatment effect was achieved among studies which provided a mean of 8.8 hours of therapy per week for 11.2 weeks compared to trials that only provided approximately 2 hours per week for 22.9 weeks. On average, positive studies provided a total of 98.4 hours of therapy while negative studies provided a total of 43.6 hours of therapy. Consequently, total length of therapy was significantly inversely correlated with mean change in Porch Index of Communicative Abilities (PICA) scores. The hours of therapy provided in a week was significantly correlated to greater improvement on the PICA and on the Token Test. And finally, total hours of therapy were significantly correlated with greater improvement on the PICA and the Token Test. The authors concluded that intense therapy over a short amount of time could improve outcomes of speech and language therapy for stroke patients with aphasia (Bhogal et al. 2003). In a Cochrane Review by Kelly H et al (2010), intensive speech language therapy (SLT) was compared to conventional SLT by four subcomparisons. When compared to conventional SLT, the intensive SLT approach was found to result in benefits in participants written language skills with some indications of improvements in their receptive language and severity of impairment measures. However, the number of drop-outs from the intensive SLT groups was significantly higher than the conventional SLT group suggesting that such an intensive approach to therapy may not be suitable for all patients. Cherney et al (2011) performed a systematic review of treatment studies which directly compares conditions of higher and lower intensity treatment for aphasia. Results were described based on International Classification of Functioning, Disability and Health-ICF (WHO 2001). Results at the ICF's language impairment and communication activity/ participation levels tend to be equivocal for both acute and chronic aphasia; with no clear differences between intensive and non-intensive treatment across studies. Conclusions Regarding the Intensity of Language Therapy There is conflicting (Level 4) evidence that more intensive language therapy results in greater improvements compared with less-intensive therapy. It is uncertain whether more-intensive language therapy is better than less-intensive therapy. 6. The Elements of Stroke Rehabilitation pg. 35 of 54

6.8 Durability of Rehabilitation Gains Functional recovery (the ability to perform activities despite impairment) and improvement in communication may continue for months after neurological recovery is complete (Stineman and Granger 1998). Between 6 months and 3 years post stroke the average level of functional ability is maintained (Dombovy et al. 1987, Borucki et al. 1992). Beyond 3-5 years, slight decreases were noted, most likely related to the effects of increasing age and comorbidity (Stineman and Granger 1998). Therefore, in the absence of a new event, it has long been thought that stroke patients tend to maintain gains made in rehabilitation over the long-term. 6.8.1 Previous Reviews Evans et al. (1995) reviewed 11 studies published between 1980 and 1993 that evaluated rehabilitation treatments, which included an untreated control group (Table 6.12). The outcomes of mortality, discharge location and functional ability were assessed. Three of the papers evaluated the rehabilitation of individuals with disabilities other than stoke. Their analysis revealed that treatment on a rehabilitation unit resulted in greater odds of survival, higher rates of discharge to home, higher rates of remaining at home at 8-12 month follow-up, and higher levels of functional ability at discharge. However, the difference in survival and functional independence had disappeared at the 12-month follow-up period, suggesting that many patients who are discharged from rehabilitation may deteriorate medically, physically, and functionally. Table 6.12 Evans et al. (1995) Multidisciplinary Rehab vs. Medical Care and Durability of Rehab Gains Applegate et al. 1990* Garraway et al. 1980a, 1980b Indredavik et al. 1991 Kalra et al. 1993 Lefton et al. 1983* Rubenstein et al. 1984* Sivenius et al. 1985 Smith et al. 1982 Stevens et al. 1984 Strand et al. 1985 Wood-Dauphinee et al. 1984 * includes non stroke geriatric patients Bagg (1998) stated that this finding accentuated the need to assess the effectiveness of outpatient and home- based therapies after discharge from inpatient rehabilitation programs, as well as the role of maintenance therapy for individuals with stroke requiring long-term institutionalization. This is discussed in greater detail in the last section on Community Reintegration. Gresham et al. (1995) noted that studies examining long-term outcomes have reached mixed conclusions. Some studies found functional gains were maintained (Indredavik et al. 1991, Smith et al. 6. The Elements of Stroke Rehabilitation pg. 36 of 54

1991, Strand et al. 1985) while others did not (Garraway et al. 1980a, 1980b, 1981, Sivenius et al. 1985, Stevens et al. 1984 Sunderland et al. 1992, 1994, Wade et al. 1992). Fourteen studies evaluated the durability of the gains achieved through specialized rehabilitation. The results are presented in Table 6.13. Table 6.13 The Durability of Rehabilitation Gains Author Year Country PEDro Score Garraway et al. 1980 (a) UK 5 (RCT) Garraway et al. 1980 (b) UK 5 (RCT) Smith et al. 1981 Australia and UK 5 (RCT) Stevens et al. 1984 UK 6 (RCT) Sivenius et al. 1985 Finland 5 (RCT) Strand et al. 1985 Sweden No Score Davidoff et al. 1991 USA No Score Methods 311 consecutive patients with moderate to severe strokes, admitted within 7 days of onset of symptoms were randomized to receive treatment on either a stroke unit or one of 12 medical units on call for emergency admissions. Follow-up study of 192 stroke patients from a study. 133 recovering stroke patients were randomized to one of 3 groups: 1) intensive outpatient rehabilitation therapy 6hrs/day x 3 days/week, 2) conventional outpatient therapy (3 hrs/day, 3 days/week) or 3) home visits by a nurse with no therapy. 228 acute stroke patients were randomized to receive care on a stroke rehab ward or a general medical ward. 95 consecutive stroke patients able to tolerate an intensive rehabilitation program were randomized at one-week following stroke to either an intensive physiotherapy program on a stroke unit or a control group receiving conventional physiotherapy on a general medical unit. 293 acute (within 7 days) stroke patients were allocated to receive care on either a non-intensive stroke ward (n=110) or a general medicine ward (n=183). The ADL scores of 139 middle-band stroke patients who had received inpatient rehabilitation were reviewed at admission and discharge and at one-year follow-up from inpatient rehabilitation. Outcome A greater proportion of stroke unit patients was classified as independent when compared to medical unit patients, 50% vs. 32% at 60 days. When comparing only survivors, the proportion of independent patients rose to 62%. A greater proportion of stroke unit patients was referred to physiotherapy (PT), with shorter delays between admission and the start of PT. At one year, there were no longer significant differences in the proportion of patients that was classified as independent. 55% of stroke unit patients and 52% of medical ward patients were assessed as independent. At 3 months, the greatest improvement in ADL scores was observed in patients who received intensive therapy, followed by patients who received conventional therapy and finally by patients who did not receive any therapy. A similar pattern was repeated at 12 months. The greatest proportion of patients whose scores deteriorated at both 3 and 12 months was reported in patients who did not receive any therapy. Patients on the stroke ward received more occupational and speech therapy. A significantly greater percentage of patients on the stroke ward was assessed as independent in dressing at one-year follow-up. There were nonsignificant differences (but a trend in favour of the stroke unit) in mortality, rate of discharge home and ADL function at one-year. Patients receiving intensive physiotherapy significantly improved their level of ADL and mobility at 3, 6 and 12 months. The greatest gains were achieved in the first 3 months. After one year, fewer stroke ward patients required longterm hospitalization. Greater proportion of stroke ward patients was independent in walking, personal hygiene and dressing. There was a significant improvement in the modified Kenny self-care scale between stroke onset, discharge from rehab and follow-up. There was significant improvement in ADL scores from rehab discharge to follow-up in a subset of patients who also received outpatient therapy. 6. The Elements of Stroke Rehabilitation pg. 37 of 54

Indredavik et al. 1991 Norway 7 (Single-blind RCT) Borucki et al. 1992 USA No Score Ferrucci et al. 1993 Italy No Score Juby et al. 1996 UK 6 (RCT) Indredavik et al. 1997 Norway 7 (RCT) Leonard et al. 1998 USA No Score Broeks et al. 1999 The Netherlands (No Score) Indredavik et al. 1999 Norway 220 acute (within 7 days) stroke patients randomized to either a combined acute/rehabilitation stroke unit or a general medical unit. 71 stroke rehabilitation inpatients were divided into 2 age categories (<69 yrs and > 70 yrs) and followed for up to 24 months following discharge. Self-reported disability and neural impairment were measured in 50 stroke patients at discharge from a rehabilitation program and at 3 and 6 months later. 315 stroke patients were randomized to receive care on either an interdisciplinary stroke unit or to care on a general medical and geriatric unit an average of 2 weeks post stroke onset. 5-year follow-up study of 220 stroke patients examining long-term survival and functional state of stroke, initially randomized to either a combined acute/rehabilitation stroke unit or a general medical unit. The functional outcomes of 68 stroke patients were assessed 1 to 5 years after discharge from rehabilitation. FIM scores were obtained through telephone interviews. 54 patients with first-ever stroke who underwent inpatient stroke rehabilitation were assessed at 3-8 and 16 weeks and at 4 years. Average age at time of stroke was 53.2 years 220 unselected hospitalized stroke patients randomized to receive care on either a stroke unit or a general medical ward. 10-year follow-up study Patients who were treated on the combined stroke unit were more likely to have been discharged home, were less likely to have been institutionalized and were more likely to have higher Barthel Index scores at both 6 weeks and 1 year. The 6-week mortality rate was lower for patients treated on the combined stroke unit, but not at one year. No differences in the Barthel Index (BI) scores for the younger vs. the older patients at 6, 12 or 24 months. The frequency of new or progressive disease was the same for both age groups while death or placement in a nursing home was more frequent among the older patients. The probability of maintaining discharge BI scores at 24 months was 55% for the younger group and 25% for the older group. Although this difference was non-significant, the possibility of a type II error. Functional disability was significantly reduced at 3 and 6 months. There was also a reduction in neural impairment, as measured by the Fugl-Meyer scale. At both 6 months and 1 year, stroke unit patients had higher Nottingham Extended ADL scores. At one year, stroke unit patients had better scores on the General Health Questionnaire. At 6 months, stroke unit patients had higher Barthel Index scores and Rivermead Mobility Index scores compared to patients treated on the other units. Cognitive readjustment was better for patients on the stroke unit at 6 months. 5 years following stroke, a greater proportion of patients originally treated on the stroke unit were alive, residing at home with higher BI scores when compared to patients treated on the general medical unit. Patients FIM scores improved significantly from admission to discharge from rehabilitation. There was a trend towards improvement of scores up to one year. There were non-significant declines in FIM scores in years 4 and 5. Most upper arm recovery, assessed by the Fugl-Meyer scale occurred during the first 16 weeks post stroke. Improvement continued for after 16 weeks in 10 patients. For 13 patients recovery only started after 16 weeks. After 4 years, fair to good recovery (FM score >20) was observed in 31 patients. At 4 years 52 patients had Barthel Index scores > 60 and 33 patients had BI scores of 100. There was a group of patients that achieved a BI score of 100 and had a non-functional affected arm. Only 14/54 patients had intact sensory function and only 9/54 patients had normal muscle tone. Serious shoulder pain persisted in 11/54 patients at 4 years. At 10-years post stroke, a greater proportion of patients initially treated on the stroke unit were alive (25 vs. 13%), residing in their homes (20 vs. 8%) and had BI 6. The Elements of Stroke Rehabilitation pg. 38 of 54

7 (RCT) of Indredavik et al. (1991). scores 60 (20 vs. 8%) compared to patients treated on a general medical ward. Lincoln et al. 2000 UK 6 (RCT) 315 stroke patients were randomized to receive care on either a stroke unit or to care on a general medical and geriatric unit. 5-year follow-up study of Juby et al. 1996 (n=159). Relative risks of death, death or disability and death or institutional care were all in direction of favourable outcomes for patients initially treated on the stroke unit. Kwakkel et al. 2002 Netherlands 8 (RCT) Drummond et al. 2005 UK 6 (RCT) Bernhardt et al. 2008 Australia 8 (RCT) Discussion 110 stroke patients with a primary middle cerebral artery lesion were randomly allocated to one of three groups for a 20 week rehabilitation program with an emphasis on (1) upper limb function, (2) lower limb function or (3) immobilization with an inflatable pressure splint (control group). Follow up assessments within and between groups were compared at 6, 9, and 12 months after stroke. No significant differences between the groups were found from 6 months onward for: Barthel Index, Functional Ambulation Categories, Action Research Arm test, comfortable and maximal walking speed, Nottingham health profile part 1, sickness impact profile- 68 and Frenchay activities index. 10 year follow up of Juby et al. 1996 Of 176 patients originally allocated to receive treatment on a stroke unit, 122 (69%) were dead, 31 were disabled (Barthel Index 0-17) and 9 were in institutionalized care. (8 were untraced and 4 refused to give consent for follow-up). Of the 139 allocated to a conventional ward, 111 (80%) were dead, 9 were disabled, 2 were in institutional care. (7 were untraced and 4 refused to give consent for follow-up). The relative risks for: death (0.87, 95% CI; 0.78 to 0.97), death or disability (0.91, 95% CI; 0.94 to 1.05) and death or institutional care 0.91 (0.83 to 1.00). 71 patients within 24 hrs of stroke onset were randomly assigned to receive standard care (SC) (n=33) or SC plus very early mobilization (VEM) (n=38) until discharge or 14 days. The primary safety outcome was the number of deaths at 3 months. A good outcome, defined as a modified Rankin Score (mrs) of 0-2 at 3, 6 and 12 months was also assessed. 18% of patients screened were suitable for recruitment. There was no significant difference in the number of deaths between groups (SC, 3 of 33; VEM, 8 of 38; p=0.20). Almost all deaths occurred in patients with severe stroke. After adjusting for age, baseline NIHSS score and premorbid mrs score, the odds of experiencing a good outcome were significantly higher at 12 months for the VEM group (OR: 8.15, 95% CI 1.61-41.2, p<0.01). There was also a trend towards good outcome at 3 months, but not at 6 months. Five good (PEDro > 6) quality studies evaluated the durability of rehabilitation gains. The results are summarized in Table 6.14 Table 6.14 Summary of Outcome Measures From RCTs With PEDro 6 Evaluating the Durability of Rehabilitation Gains Author/Year PEDro score Bernhardt et al. 2008 8 N at Randomization Intervention Outcome 71 Early mobilization mrs score 0-2 (+/- at 3 mos, - at 6 mos, + at 12 mos) Kwakkel et al. 2002 8 101 Upper extremity vs. lower extremity therapy vs. control condition Barthel Index (- at weeks, 26 and 52) Walking ability (- at weeks 26 and 52) Dexterity (ARA) (+ at 26 weeks-arm training, - at week 52) 6. The Elements of Stroke Rehabilitation pg. 39 of 54

Indredavik et al 1991, 1997 and 1999 7 * Stevens et al. 1984 6 Juby et al. 1996, Lincoln et al. 2000 and Drummond et al. 2005* 6 220 Inpatient stroke rehabilitation Barthel Index (+ at 6, 52 weeks, 5 and 10 years) 228 Inpatient stroke rehabilitation Independence in dressing (+ at 12 mos) 315 Stroke unit care Nottingham EADL (+ at 6 mos and 1 year) General Health Questionnaire (+ at 1 year) BI scores (+ at 6 mos) Unfavourable outcomes (death, death/disability, death/institutionalization) (+ at 5 years) Death (+ at 10 years) * Considered as a single study. Same patients followed forward in time. + Indicates a positive effect favouring treatment group - Indicates a non-significant difference between the intervention group All of these studies reported improvement in the functional outcome of stroke rehabilitation patients compared to the control group (general medical ward) anywhere between 12 months and 10 years following stroke. The relative benefit attributed to stroke rehabilitation appears to be very robust. However, the absolute gains achieved through stroke rehabilitation appear to be less robust. Stevens et al. (1984) found selective continued improvement from four to 12 months. In contrast, patients in the control group actually declined in function. Indredavik et al. (1997, 1999) reported a decline in scores associated with functional outcome between five and 10 years post stroke, although the Barthel Index scores of patients treated on the stroke unit were higher compared to control group patients. Davidoff et al. (1991) reported a significant improvement in ADL scores between rehabilitation discharge and one year. Leonard et al (1998) found that FIM scores improved for the first year and then plateaued, with a non-significant decline over the next four to five years. Bernhardt et al. (2008) demonstrated that early mobilization during the first 2 weeks following stroke was associated with a good outcome at 12 months following stroke. The program was also found to be cost-effective. Mean total costs over the 12 month study period were AUD $13,559 for the AVERT group and AUD $21,860 for the standard care group (Tay-Teo et al. 2008). A large follow-up study, AVERT III, is planned to examine the effects of additional early mobilization (3x/day for 14 days) following acute stroke. Conclusions Regarding the Durability of Rehabilitation Gains There is strong (Level 1a) evidence that relatively greater functional improvements are made by patients rehabilitated on specialized stroke units when compared to general medical units and the effects are maintained over both the short-term and long-term. There is strong (Level 1a) evidence that functional outcomes achieved through stroke rehabilitation are maintained and actually improve for up to one year. There is moderate (level 1b) evidence that by five years post-stroke functional outcomes plateau and may decline. By ten years, overall functional outcome scores significantly decline although it is unclear to what extent the natural aging process and comorbidity may contribute to these declines. 6. The Elements of Stroke Rehabilitation pg. 40 of 54

Greater functional improvements made on interdisciplinary stroke rehabilitation units are maintained over the long-term. 6. The Elements of Stroke Rehabilitation pg. 41 of 54

6.9 Summary 1. There is limited (Level 2) evidence that the improvement in disability seen in stroke rehabilitation cannot be explained on the basis of natural recovery of the neurological impairment alone. 2. The term 'stroke unit' is broad and describes hierarchal service organisation in stroke care along a continuum from more organised to less organised care. 3. There is strong (Level 1a) evidence that individuals who received organised inpatient stroke unit care are more likely to survive, return home and regain independence. 4. Although improved outcomes have been reported among trials evaluating stroke units, no causal mechanism(s) has been identified and verified. There is moderate (Level 1b) evidence that patients treated on a stroke unit have fewer medical complications. 5. There is conflicting (Level 4) evidence that patients suffering from hemorrhagic stroke have worse long-term outcomes compared with patients with ischemic stroke. 6. There is conflicting (Level 4) evidence to suggest that care pathways improve stroke rehabilitation outcomes. There is moderate (Level 1b) evidence that care pathways do not reduce hospital costs or decrease length of hospital stays. There is strong (Level 1a) evidence that compliance with stroke rehabilitation guidelines and adherence to processes of care result in improved outcomes 7. There is limited (Level 2) evidence that early admission to stroke rehabilitation units directly results in improved functional outcomes. Stroke patients should be admitted to stroke rehabilitation units as soon as they are medically stable. 8. There is strong (Level 1a) evidence that very early mobilization following stroke helps to reduce medical complications. 9. There is moderate (Level 1b) evidence that very early mobilization following stroke helps to improve functional recovery and decrease the time to achieve functional walking. 10. There is strong (Level 1a) evidence that greater intensity of physical and occupational therapy results in improved functional outcomes. However, the overall beneficial effect is modest and the positive benefits associated with greater treatment intensities were not maintained over time. 6. The Elements of Stroke Rehabilitation pg. 42 of 54

11. There is moderate (Level 1b) evidence, based on the results from a single study that the same therapies delivered more intensively over a shorter period of time results in faster recovery and earlier discharge from hospital. 12. There is conflicting (Level 4) evidence whether more-intensive language therapy is efficacious in treating aphasia post-stroke. 13. There is strong (Level 1a) evidence that relatively greater functional improvements are made by patients rehabilitated on specialized stroke units when compared to general medical units and the effects are maintained over both the short-term and long-term. There is strong (Level 1a) evidence that functional outcomes achieved through stroke rehabilitation are maintained and actually improve for up to one year. There is moderate (Level 1b) evidence that these same functional gains decline after five years. 6. The Elements of Stroke Rehabilitation pg. 43 of 54

References Andersen KK, Olsen TS, Dehlendorff C, Kammersgaard LP. Hemorrhagic and ischemic strokes compared: stroke severity, mortality, and risk factors. Stroke 2009; 40:2068-2072. Askim T, Morkved S, Engen A, Roos K, Aas T, Indredavik B. Effects of a Community-Based Intensive Motor Training Program Combined With Early Supported Discharge After Treatment in a Comprehensive Stroke Unit. A Randomized, Controlled Trial. Stroke 2010; 41:1697-1703. Bagg SD. Outcome predictors and the effectiveness of stroke rehabilitation. Physical Medicine and Rehabilitation: State of the Art Reviews 1998; 12:581-592. Bai Y, Hu Y, Wu Y et al. A prospective, randomized, single-blinded trial on the effect of early rehabilitation on daily activities and motor function of patients with hemorrhagic stroke. J Clin Neurosci 2012. Bagley P, Hudson M, Green J, Forster A, Young J. Do physiotherapy staff record treatment time accurately? An observational study. Clin Rehabil 2009 Sep; 23(9):841-5. Bakheit AM, Shaw S, Barrett L, et al. A prospective, randomized, parallel group, controlled study of the effect of intensity of speech and language therapy on early recovery from poststroke aphasia. Clin Rehabil 2007; 21:885-894. Ballinger C, Ashburn A, Low J, Roderick P. Unpacking the black box of therapy -- a pilot study to describe occupational therapy and physiotherapy interventions for people with stroke. Clin Rehabil 1999; 13:301-309. Bernhardt J, Dewey H, Thrift A et al. Inactive and alone: physical activity within the first 14 days of acute stroke unit care. Stroke. 2004; 35:1005-1009. Bernhardt J, Dewey H, Collier J et al. A very early rehabilitation trial (AVERT). International Journal of Stroke. 2006; 1:169-171. Bernhardt J, Chitravas N, Meslo IL, Thrift AG, Indredavik B. Not all stroke units are the same: a comparison of physical activity patterns in Melbourne, Australia, and Trondheim, Norway. Stroke 2008; 39:2059-2065. Bernhardt J, Dewey H, Thrift A, Collier J, Donnan G. A very early rehabilitation trial for stroke (AVERT): phase II safety and feasibility. Stroke 2008; 39:390-396. Bernhardt J, Thuy MN, Collier JM, Legg LA. Very early versus delayed mobilisation after stroke. Cochrane Database Syst Rev 2009; CD006187. Bhogal SK, Teasell R, Speechley M. Intensity of aphasia therapy, impact on recovery. Stroke 2003; 34:987-993. 6. The Elements of Stroke Rehabilitation pg. 44 of 54

Borucki S, Volpe BT, Reding M. The effect of age on maintenance of funcional gains following stroke rehabilitation. J Neuro Rehab 1992; 6:1-5. Broeks JG, Lankhorst GJ, Rumping K, Prevo AJ. The long-term outcome of arm function after stroke: results of a follow-up study. Disabil Rehabil 1999; 21(8):357-64. Brindley P, Copeland M, Demain C, Martyn P. a Comparison of the speech of ten chronic Broca's aphasics following intensive and non-intensive periods of therapy. Aphasiology 1989; 3:695-707. Cauraugh JH, Naik SK, Lodha N, Coombes SA, Summers JJ. Long-term rehabilitation for chronic stroke arm movements: a randomized controlled trial. Clin Rehabil 2011; 25:1086-1096. Chan DKY, Cordato D, O'Rourke F, Chan DL, Pollack M, Middleton S, Levi C. Comprehensive stroke units: A review of comparative evidence and experience. Int J Stroke 2013; 8:260-264 Chen CC, Heinemann AW, Granger CV, Linn RT. Functional gains and therapy intensity during subacute rehabilitation: a study of 20 facilities. Arch Phys Med Rehabil 2002; 83(11):1514-23. Cherney RL, Patterson JP, Raymer AM. Intensity of aphasia theraoy: Evidenc e and efficacy. Curr Neurol Neurosci Rep 2011; 11(6):560-569 Chiu D, Peterson L, Elkind MS, Rosand J, Gerber LM, Silverstein MD. Comparison of outcomes after intracerebral hemorrhage and ischemic stroke. J Stroke Cerebrovasc Dis 2010; 19:225-229. Cifu DX, Stewart DG. Factors affecting functional outcome after stroke: a critical review of rehabilitation interventions. Arch Phys Med Rehabil 1999; 80:S35-S39. Conroy BE, DeJong G, Horn SD. Hospital-based stroke rehabilitation in the United States. Top Stroke Rehabil 2009; 16:34-43. Cooke EV, Mares K, Clark A, Tallis RC, Pomeroy VM. The effects of increased dose of exercise-based therapies to enhance motor recovery after stroke: a systematic review and meta-analysis. BMC Med 2010; 8:60. Craig LE, Bernhardt J, Langhorne P, Wu O. Early mobilization after stroke: an example of an individual patient data meta-analysis of a complex intervention. Stroke 2010; 41(11):2632-2636. Cumming TB, Thrift AG, Collier JM, Churilov L, Dewey HM, Donnan GA et al. Very early mobilization after stroke fast-tracks return to walking: further results from the phase II AVERT randomized controlled trial. Stroke 2011; 42(1):153-158. David R, Enderby P, Bainton D. Treatment of acquired aphasia: speech therapists and volunteers compared. J Neurol Neurosurg Psychiatry 1982; 45:957-961. Davidoff GN, Keren O, Ring H, Solzi P. Acute stroke patients: long-term effects of rehabilitation and maintenance of gains. Arch Phys Med Rehabil 1991; 72:869-873. 6. The Elements of Stroke Rehabilitation pg. 45 of 54

Diserens K, Moreira T, Hirt L et al. Early mobilization out of bed after ischaemic stroke reduces severe complications but not cerebral blood flow: a randomized controlled pilot trial. Clin Rehabil 2012; 26:451-459. De Wit L, Putman K, Dejaeger E, et al. Use of time by stroke patients: a comparison of four European rehabilitation centers. Stroke 2005; 36:1977-1983. De Weerdt W, Selz B, Nuyens G, et al. Time use of stroke patients in an intensive rehabilitation unit: a comparison between a Belgian and a Swiss setting. Disabil Rehabil 2000; 22:181-186. Di Lauro A, Pellegrino L, Savastano G, Ferraro C, Fusco M, Balzarano F, Franco MM, Biancardi LG, Grasso A. A randomized trial on the efficacy of intensive rehabilitation in the acute phase of ischemic stroke. J Neurol 2003t; 250(10):1206-8. Diserens K, Michel P, Bogousslavsky J. Early mobilisation after stroke: review of the literature. Cerebrovasc Dis 2006; 22(2-3):183-190. Dobkin BH. Focused stroke rehabilitation programs do not improve outcome. Arch Neurol 1989; 46:701-703. Dombovy ML, Basford JR, Whisnant JP, Bergstralh EJ. Disability and use of rehabilitation services following stroke in Rochester, Minnesota, 1975-1979. Stroke 1987; 18:830-836. Duncan PW, Zorowitz R, Bates B, Choi JY, Glasberg JJ, Graham GD et al. Management of Adult Stroke Rehabilitation Care: a clinical practice guideline. Stroke 2005; 36(9):e100-e143. Edwards SG, Thompson AJ, Playford ED. Integrated care pathways: disease-specific or process-specific? Clin Med 2004; 4:132-135. Evans A, Perez I, Harraf F, Melbourn A, Steadman J, Donaldson N et al. Can differences in management processes explain different outcomes between stroke unit and stroke-team care? Lancet 2001; 358(9293):1586-1592. Evans A, Harraf F, Donaldson N, Kalra L. Randomized controlled study of stroke unit care versus stroke team care in different stroke subtypes. Stroke 2002; 33(2):449-55. Evans RL, Connis RT, Hendricks RD, Haselkorn JK. Multidisciplinary rehabilitation versus medical care: A meta-analysis. Soc Sci Med 1995; 40:1699-1706. Falconer JA, Roth EJ, Sutin JA, Strasser DC, Chang RW. The critical path method in stroke rehabilitation: lessons from an experiment in cost containment and outcome improvement. Qual Rev Bull 1993; 19:8-16. Fang Y, Chen X, Li H, Lin J, Huang R, Zeng J. A study on additional early physiotherapy after stroke and factors affecting functional recovery. Clin Rehabil 2003;17(6):608-17. 6. The Elements of Stroke Rehabilitation pg. 46 of 54

Feigenson JS, McCarthy ML, Greenberg SD, Feigenson WD. Factors influencing outcome and length of stay in a stroke rehabilitation unit. Part 2. Comparison of 318 screened and 248 unscreened patients. Stroke 1977; 8:657-662 Ferrucci L, Bandinelli S, Guralnik JM, Lamponi M, Bertini C, Falchini M, Baroni A. Recovery of functional status after stroke. A postrehabilitation follow-up study. Stroke 1993; 24:200-205. Feys HM, De Weerdt WJ, Selz BE, Cox SG, Spichiger R, Vereeck LE, Putman KD, Van Hoydonck GA. Effect of a therapeutic intervention for the hemiplegic upper limb in the acute phase after stroke: a single-blind, randomized, controlled multicenter trial. Stroke 1998; 29:785-792. Foley N, Salter K, Teasell R. Specialized stroke services: A metanalaysis comparing three models of care. Cerebrovasc Dis 2007; 23:194-202Forster A, Young J. The clinical and cost effectiveness of physiotherapy in the management of elderly people following a stroke. Bradford Elderly Care and Rehabilitation Research Department, UK, 2002. Gagnon D, Nadeau S, Tam V. Ideal timing to transfer from an acute care hospital to an interdisciplinary inpatient rehabilitation program following a stroke: an exploratory study. BMC Health Serv Res. 2006; 6:151. Galvin R, Murphy B, Cusack T, Stokes E. The impact of increased duration of exercise therapy on functional recovery following stroke--what is the evidence? Top Stroke Rehabil 2008; 15(4):365-377. Garraway WM, Akhtar AJ, Hockey L, Prescott RJ. Management of acute stroke in the elderly: follow-up of a controlled trial. Br Med J 1980; 281:827-829. Garraway WM, Akhtar AJ, Prescott RJ, Hockey L. Management of acute stroke in the elderly: preliminary results of a controlled trial. Br Med J 1980; 280:1040-1043. Garraway WM, Akhtar AJ,Smith DL, Smith ME. Community Health 1981; 35:39-44. The triage of stroke rehabilitation. J Epidemiol Govan L, Langhorne P, Weir CJ. Does the prevention of complications explain the survival benefit of organized inpatient (stroke unit) care? further analysis of a systematic review. Stroke 2007; 38:2536-2540. Gresham GE, Duncan PW, Statson WB. Poststroke Rehabilitation Guideline Panel. Clinical Practice Guideline No 16 (AHCPR Publication No 95-0662). US Department of Health and Human Services, Public Health Service, Agency for Health Care Policy and Research, Rockville MD, 1995. Hamrin EK, Lindmark B. The effect of systematic care planning after acute stroke in general hospital medical wards. J Adv Nurs 1990; 15:1146-1153. Hartman J, Landau WM. Comparison of formal language therapy with supportive counseling for aphasia due to acute vascular accident. Arch Neurol 1987; 44:646-649. 6. The Elements of Stroke Rehabilitation pg. 47 of 54

Hayes SH, Carroll SR. Early intervention care in the acute stroke patient. Arch Phys Med Rehabil 1986; 67:319-321. Hu MH, Hsu SS, Yip PK, Jeng JS, Wang YH. Early and intensive rehabilitation predicts good functional outcomes in patients admitted to the stroke intensive care unit. Disabil Rehabil 2010; 32:1251-1259. Huang HC, Chung KC, Lai DC, Sung SF. The impact of timing and dose of rehabilitation delivery on functional recovery of stroke patients. J Chin Med Assoc 2009; 72:257-264. Indredavik B, Bakke F, Slordahl SA, Rokseth R, Haheim LL. Stroke unit treatment. 10-year follow-up. Stroke 1999; 30:1524-1527. Indredavik B, Bakke F, Solberg R, Rokseth R, Haaheim LL, Holme I. Benefit of a stroke unit: a randomized controlled trial. Stroke 1991; 22:1026-1031. Indredavik B, Slordahl SA, Bakke F, Rokseth R, Haheim LL. Stroke unit treatment. Long-term effects. Stroke 1997; 28:1861-1866. Indredavik B, Rohweder G, Naalsund E, Lydersen S. Medical complications in a comprehensive stroke unit and an early supported discharge service. Stroke 2008; 39:414-420. Jette DU, Warren RL, Wirtalla C. The relation between therapy intensity and outcomes of rehabilitation in skilled nursing facilities. Arch Phys Med Rehabil 2005; 86:373-379. Jorgensen HS, Nakayama H, Raaschou HO, Olsen TS. Intracerebral hemorrhage versus infarction: stroke severity, risk factors, and prognosis. Ann Neurol 1995; 38:45-50. Juby LC, Lincoln NB, Berman P. The effect of a stroke rehabilitation unit on functional and psychological outcome: a randomised controlled trial. Cerebrovasc Dis 1996; 6:106-110. Kalra L, Dale P, Crome P. Improving stroke rehabilitation. A controlled study [see comments]. Stroke 1993; 24:1462-1467. Kalra L. The influence of stroke unit rehabilitation on functional recovery from stroke. Stroke 1994; 25:821-825. Katrak PH, Black D, Peeva V. Do stroke patients with intracerebral hemorrhage have a better functional outcome than patients with cerebral infarction? PM R 2009; 1(5):427-433. Kelly H, Brady MC, Enderby P. Speech and language therapy for aphasia following stroke. Cochrane Database of Systematic Reviews 2010, Issue 5. Art. No.: CD000425. DOI: 10.1002 /14651858. CD000425.pub2. Kelly PJ, Furie KL, Shafqat S, Rallis N, Chang Y, Stein J. Functional recovery following rehabilitation after hemorrhagic and ischemic stroke. Arch Phys Med Rehabil 2003; 84:968-972. 6. The Elements of Stroke Rehabilitation pg. 48 of 54

Keren O, Motin M, Heinemann AW, O'Reilly CM, Bode RK, Semik P, Ring H. Relationship between rehabilitation therapies and outcome of stroke patients in Israel: a preliminary study. Isr Med Assoc J 2004; 6:736-741. Kwakkel G, Wagenaar RC, Koelman TW, Lankhorst GJ, Koetsier JC. Effects of intensity of rehabilitation after stroke. A research synthesis. Stroke 1997; 28:1550-1556. Kwakkel G, Wagenaar RC, Twisk JW, Lankhorst GJ, Koetsier JC. Intensity of leg and arm training after primary middle-cerebral-artery stroke: a randomised trial. Lancet 1999; 354:191-196. Kwakkel G, Kollen BJ, Wagenaar RC. Long term effects of intensity of upper and lower limb training after stroke: a randomised trial. J Neurol Neurosurg Psychiatry 2002; 72:473-479. Kwakkel G, van Peppen R, Wagenaar RC, et al. Effects of augmented exercise therapy time after stroke: a meta-analysis. Stroke 2004; 35:2529-2539. Kwakkel G. Impact of intensity of practice after stroke: Issues for consideration.disabil Rehabil 2006; 28(13 14):823 830. Kwan J, Sandercock P. In-hospital care pathways for stroke. Cochrane Database Syst Rev 2002;(2):CD002924. Review. Kwan J, Sandercock P. In-hospital care pathways for stroke. The Cochrane Database of Systematic Reviews 2004, Issue 4. Art.No.: CD002924.pub2. DOI: 0.1002/14651858.CD002924.pub2.update 2009 Kwan J. Care pathways for acute stroke care and stroke rehabilitation: from theory to evidence. J Clin Neurosci. 2007; 14:189-200. Langhammer B, Lindmark B, Stanghelle JK. Stroke patients and long-term training: is it worthwhile? A randomized comparison of two different training strategies after rehabilitation. Clin Rehabil 2007; 21:495-510. Langhammer B, Stanghelle JK, Lindmark B. Exercise and health-related quality of life during the first year following acute stroke. A randomized controlled trial. Brain Inj 2008; 22:135-145. Langhammer B, Stanghelle JK, Lindmark B. An evaluation of two different exercise regimes during the first year following stroke: a randomised controlled trial. Physiother Theory Pract 2009; 25:55-68. Langhorne P, Wagenaar R, Partridge C. Physiotherapy after stroke: more is better? Physiother Res Int 1996; 1:75-88. Langhorne P, Stott D, Knight A, Bernhardt J, Barer D, Watkins C. Very early rehabilitation or intensive telemetry after stroke: a pilot randomised trial. Cerebrovasc Dis 2010; 29:352-360. Langhorne P, Fearon P, Ronning OM, Kaste M, Palomaki H et al. Stroke Unit Care Benefits Patients with Intracerebral Hamorrhage: Systematic Review and Meta-analysis. Stroke 2013; 44:3044-3049 6. The Elements of Stroke Rehabilitation pg. 49 of 54

Laver K, Ratcliffe J, George S, Lester L, Walker R, Burgess L et al. Early rehabilitation management after stroke: what do stroke patients prefer? J Rehabil Med 2011; 43(4):354-358. Leonard CT, Miller KE, Griffiths HI, McClatchie G. A sequential study assessing functional outcomes of first-time stroke survivors 1 to 5 years after rehabilitation. J Stroke Cerebrovascular Disease 1998; 7:145-153. Lincoln NB, Willis D, Philips SA, Juby LC, Berman P. Comparison of rehabilitation practice on hospital wards for stroke patients. Stroke 1996; 27:18-23. Lincoln NB, Husbands S, Trescoli C, Drummond AE, Gladman JR, Berman P. Five year follow up of a randomised controlled trial of a stroke rehabilitation unit. BMJ 2000; 320(7234):549. Lincoln NB, McGuirk E, Mulley GP, Lendrem W, Jones AC, Mitchell JR. Effectiveness of speech therapy for aphasic stroke patients. A randomised controlled trial. Lancet 1984; 1:1197-1200. Lincoln NB, Parry RH, Vass CD. Randomized, controlled trial to evaluate increased intensity of physiotherapy treatment of arm function after stroke. Stroke 1999; 30:573-579. Lind K. A synthesis of studies on stroke rehabilitation. J Chronic Dis 1982; 35:133-149. Logan PA, Ahern J, Gladman JR, Lincoln NB. A randomized controlled trial of enhanced Social Service occupational therapy for stroke patients. Clin Rehabil 1997;11:107-113. Marshall RC, Wertz RT, Weiss DG, Aten JL, Brookshire RH, Garcia-Bunuel L, Holland AL, Kurtzke JF, LaPointe LL, Milianti FJ. Home treatment for aphasic patients by trained nonprofessionals. J Speech Hear Disord 1989;54:462-470. Maulden SA, Gassaway J, Horn SD et al. Timing of initiation of rehabilitation after stroke. Arch Phys Med Rehabil. 2005;86:S34-S40. Middleton S, McElduff P, Ward J et al. Implementation of evidence-based treatment protocols to manage fever, hyperglycaemia, and swallowing dysfunction in acute stroke (QASC): a cluster randomised controlled trial. Lancet 2011;378:1699-1706. Milianti FJ, Brannegan R. Comparison of clinic, home, and deferred language treatment for aphasia. A Veterans Administration Cooperative Study. Arch Neurol 1986; 43:653-658. Meikle M, Wechsler E, Tupper A, Benenson M, Butler J, Mulhall D, Stern G. Comparative trial of volunteer and professional treatments of dysphasia after stroke. Br Med J 1979; 2:87-89. Musicco M, Emberti L, Nappi G, Caltagirone C; Italian Multicenter Study on Outcomes of Rehabilitation of Neurological Patients. Early and long-term outcome of rehabilitation in stroke patients: the role of patient characteristics, time of initiation, and duration of interventions. Arch Phys Med Rehabil 2003; 84(4):551-8. 6. The Elements of Stroke Rehabilitation pg. 50 of 54

Nugent JA, Schurr KA, Adams RD. A dose-response relationship between amount of weight-bearing exercise and walking outcome following cerebrovascular accident. Arch Phys Med Rehabil 1994; 75:399-402. Ottenbacher KJ, Jannell S. The results of clinical trials in stroke rehabilitation research [see comments]. Arch Neurol 1993; 50:37-44. Page SJ. Intensity versus task-sepcificity after stroke: how important is intensity? Am J Phys Med Rehabil 2003; 82(9):730-2. Page SJ, Schmid A, Harris JE. Optimizing terminology for stroke motor rehabilitation: Recommendations from the American Congress of Rehabilitation Medicine Stroke Movement Interventions Subcommittee. Arch Phys Med Rehabil 2012; 93:1395-9. Panella M, Marchisio S, Brambilla R, Vanhaecht K, Di SF. A cluster randomized trial to assess the effect of clinical pathways for patients with stroke: results of the Clinical Pathways for Effective and Appropriate Care Study. BMC Med 2012 Jul 10; 10(1):71. Paolucci S, Antonucci G, Grasso MG, Morelli D, Troisi E, Coiro P, Bragoni M. Early versus delayed inpatient stroke rehabilitation: a matched comparison conducted in Italy. Arch Phys Med Rehabil 2000; 81:695-700. Paolucci S, Antonucci G, Grasso MG, Bragoni M, Coiro P, De AD, Fusco FR, Morelli D, Venturiero V, Troisi E, Pratesi L. Functional outcome of ischemic and hemorrhagic stroke patients after inpatient rehabilitation: a matched comparison. Stroke 2003; 34:2861-2865. Park HJ, Oh DW, Kim SY, Choi JD. Effectiveness of community-based ambulation training for walking function of post-stroke hemiparesis: a randomized controlled pilot trial. Clin Rehabil 2011; 25(5):451-459. Parry RH, Lincoln NB, Vass CD. Effect of severity of arm impairment on response to additional physiotherapy early after stroke. Clin Rehabil 1999; 13:187-198. Partridge C, Mackenzie M, Edwards S, Reid A, Jayawardena S, Guck N, Potter J. Is dosage of physiotherapy a critical factor in deciding patterns of recovery from stroke: a pragmatic randomized controlled trial. Physiother Res Int 2000; 5:230-240. Patel M, Potter J, Perez I, Kalra L. The process of rehabilitation and discharge planning in stroke: a controlled comparison between stroke units. Stroke 1998; 29:2484-2487. Poeck K, Huber W, Willmes K. Outcome of intensive language treatment in aphasia. J Speech Hear Disord 1989; 54:471-479. Pohl M, Mehrholz J, Ritschel C, Ruckriem S. Speed-dependent treadmill training in ambulatory hemiparetic stroke patients: a randomized controlled trial. Stroke 2002; 33(2):553-8. 6. The Elements of Stroke Rehabilitation pg. 51 of 54

Pohl M, Werner C, Holzgraefe M et al. Repetitive locomotor training and physiotherapy improve walking and basic activities of daily living after stroke: a single-blind, randomized multicentre trial (DEutsche GAngtrainerStudie, DEGAS). Clin Rehabil. 2007; 21:17-27. Prins RS, Schoonen R, Vermeulen J. Efficacy of two different types of speech therapy for aphasic stroke patients. Applied Psycholingistics 1989; 10:85-123. Quinn TJ, Paolucci S, Sivenius J, Walker MF, Toni D, Lees KR. Evidence-based stroke r-ehabilitation: an expanded guidance document from the European stroke organisation (ESO) guidelines for management of ischaemic stroke and transient ischaemic attack 2008. J Rehabil Med 2009; 41:99-111. Richards CL, Malouin F, Wood-Dauphinee S, Williams JI, Bouchard JP, Brunet D. Task-specific physical therapy for optimization of gait recovery in acute stroke patients. Arch Phys Med Rehabil 1993; 74:612-620. Rodgers H, Mackintosh J, Price C, Wood R, McNamee P, Fearon T, Marritt A, Curless R. Does an early increased-intensity interdisciplinary upper limb therapy programme following acute stroke improve outcome? Clin Rehabil 2003; 17(6):579-89. Rudd AG, Hoffman A, Irwin P, Lowe D, Pearson MG. Stroke unit care and outcome: results from the 2001 National Sentinel Audit of Stroke (England, Wales, and Northern Ireland). Stroke 2005; 36:103-106. Ruff RM, Yarnell S, Marinos JM. Are stroke patients discharged sooner if in-patient rehabilitation services are provided seven v six days per week? Am J Phys Med Rehabil 1999; 78:143-146. Salter K, Jutai J, Hartley M, Foley N, Bhogal S, Bayona N, Teasell R. Impact of early vs delayed admission to rehabilitation on functional outcomes in persons with stroke. J Rehabil Med. 2006 Mar; 38(2):113-7. Shah S, Vanclay F, Cooper B. Efficiency, effectiveness, and duration of stroke rehabilitation. Stroke 1990; 21:241-246. Shewan CM, Kertesz A. Effects of speech and language treatment on recovery from aphasia. Brain Lang 1984; 23:272-299. Sivenius J, Pyorala K, Heinonen OP, Salonen JT, Riekkinen P. The significance of intensity of rehabilitation of stroke--a controlled trial. Stroke 1985; 16:928-931. Slade A, Tennant A, Chamberlain MA. A randomized controlled trial to determine the effect of intensity of therapy upon length of stay in a neurological rehabilitation setting. J Rehabil Med 2002; 34:260-266. Smith DS, Goldenberg E, Ashburn A, Kinsella G, Sheikh K, Brennan PJ, Meade TW, Zutshi DW, Perry JD, Reeback JS. Remedial therapy after stroke: a randomised controlled trial. Br Med J (Clin Res Ed) 1981; 282:517-520. 6. The Elements of Stroke Rehabilitation pg. 52 of 54

Sonoda S, Saitoh E, Nagai S, Kawakita M, Kanada Y. Full-time integrated treatment program, a new system for stroke rehabilitation in Japan: comparison with conventional rehabilitation. Am J Phys Med Rehabil 2004; 83(2):88-93. Sorbello D, Dewey HM, Churilov L, Thrift AG, Collier JM, Donnan G, Bernhardt J. Very early mobilisation and complications in the first 3 months after stroke: further results from phase II of A Very Early Rehabilitation Trial (AVERT). Cerebrovasc Dis 2009; 28:378-383. Stevens RS, Ambler NR, Warren MD. A randomized controlled trial of a stroke rehabilitation ward. Age Ageing 1984; 13:65-75. Stineman MG, Granger CV. Outcome, efficiency, and time-trend pattern analyses for stroke rehabilitation. Am J Phys Med Rehabil 1998; 77:193-201. Strand T, Asplund K, Eriksson S, Hagg E, Lithner F, Wester PO. A non-intensive stroke unit reduces functional disability and the need for long-term hospitalization. Stroke 1985; 16:29-34. Strasser DC, Falconer JA, Stevens AB, et al. Team training and stroke rehabilitation outcomes: a cluster randomized trial. Arch Phys Med Rehabil 2008; 89:10-15. Stroke Unit Trialists Collaboration. Organised inpatient (stroke unit) care for stroke. The Cochrane Database of Systematic Reviews 2001, Issue 3. Art. No.: CD000197. DOI: 10.1002/14651858.CD000197. Stroke Unit Trialists Collaboration. Organised inpatient (stroke unit) care for stroke. The Cochrane Database of Systematic Reviews 2013, Issue 9. Art No:CD000197.DOI:10.1002/14651868. CD000197.pub3 Sulch D, Evans A, Melbourn A, Kalra L. Does an integrated care pathway improve processes of care in stroke rehabilitation? A randomized controlled trial. Age Ageing 2002; 31(3):175-179. Sulch D, Kalra L. Integrated care pathways in stroke management. Age Ageing 2000; 29:349-352. Sulch D, Melbourn A, Perez I, Kalra L. Integrated care pathways and quality of life on a stroke rehabilitation unit. Stroke 2002 June; 33(6):1600-4. Sullivan KJ, Knowlton BJ, Dobkin BH. Step training with body weight support: effect of treadmill speed and practice paradigms on poststroke locomotor recovery. Arch Phys Med Rehabil. 2002; 83(5):683-691. Sunderland A, Fletcher D, Bradley L, Tinson D, Hewer RL, Wade DT. Enhanced physical therapy for arm function after stroke: a one year follow up study. J Neurol Neurosurg Psychiatry 1994; 57:856-858. Sunderland A, Tinson DJ, Bradley EL, Fletcher D, Langton HR, Wade DT. Enhanced physical therapy improves recovery of arm function after stroke. A randomised controlled trial. J Neurol Neurosurg Psychiatry 1992; 55:530-535. 6. The Elements of Stroke Rehabilitation pg. 53 of 54

Sundseth A, Thommessen B, Ronning OM. Outcome After Mobilization Within 24 Hours of Acute Stroke: A Randomized Controlled Trial. Stroke 2012 Jun 14. Tay-Teo K, Moodie M, Bernhardt J, Thrift AG, Collier J, Donnan G, Dewey H. Economic evaluation alongside a phase II, multi-centre, randomised controlled trial of very early rehabilitation after stroke (AVERT). Cerebrovasc Dis 2008; 26:475-481. The Glasgow Augmented Physiotherapy Study (GAPS) group. Can augmented physiotherapy input enhance recovery of mobility after stroke? A randomized controlled trial. Clin Rehabil 2004; 18:529-537. Turton A, Pomeroy V. When should upper limb function be trained after stroke? Evidence for and against early intervention. Neurorehabilitation 2002; 17(3):215-224. van Wijk R, Cumming T, Churilov L, Donnan G, Bernhardt J. An Early Mobilization Protocol Successfully Delivers More and Earlier Therapy to Acute Stroke Patients: Further Results From Phase II of AVERT. Neurorehabil Neural Repair 2012; 26:20-6. Wade DT, Collen FM, Robb GF, Warlow CP. Physiotherapy intervention late after stroke and mobility [see comments]. BMJ 1992; 304:609-613. Wade DT. Research into the black box of rehabilitation: the risks of a Type III error. Clin Rehabil 2001; 15:1-4. Weiss A, Suzuki T, Bean J, Fielding RA. High intensity strength training improves strength and functional performance after stroke. Am J Phys Med Rehabil 2000; 79.(4):369-376. Werner RA, Kessler S. Effectiveness of an intensive outpatient rehabilitation program for postacute stroke patients. Am J Phys Med Rehabil 1996; 75:114-120. Wertz RT, Weiss DG, Aten JL, Brookshire RH, Garcia-Bunuel L, Holland AL, Kurtzke JF, LaPointe LL, Wertz RT. Communication deficits in stroke survivors. An overview of classification and treatment. Stroke 1990; 21:II16-II18. Wodchis WP, Teare GF, Naglie G, et al. Skilled nursing facility rehabilitation and discharge to home after stroke. Arch Phys Med Rehabil 2005; 86:442-448. Whyte J, Hart T. It's more than a black box; it's a Russian doll: defining rehabilitation treatments. Am J Phys Med Rehabil 2003; 82:639-652. Wood-Dauphinee S, Shapiro S, Bass E, Fletcher C, Georges P, Hensby V, Mendelsohn B. A randomized trial of team care following stroke. Stroke 1984; 15:864-872. Yagura H, Miyai I, Seike Y, Suzuki T, Yanagihara T. Benefit of inpatient multidisciplinary rehabilitation up to 1 year after stroke. Arch Phys Med Rehabil 2003; 84(11):1687-91. 6. The Elements of Stroke Rehabilitation pg. 54 of 54