In search of the great discovery in Stroke Motor Recovery
The impact of stroke looms large
Stroke rehabilitation is a big deal. The financial and human cost of stroke looms large, and the pressure building on clinicians, researchers and funding agencies has surely reached boiling point. There is more information about stroke rehabilitation than ever before, and keeping up with the enormous amount of research is just not feasible or even possible. For those of us interested in recovery after stroke, the vortex of information can become overwhelming. At some point we really should stop and check. What direction are we heading in? What do we need to know before moving on? Do the key stakeholders – ie. stroke survivors, family, friends, clinicians and researchers have access to correct and relevant information? What do we need to consider before continuing with our quest to maximise the quality of rehabilitation after stroke?
Recovery, Rehabilitation and life transition
Recovery after stroke is a fascinating topic. Understandably, the first priorities are to survive a stroke, and then minimise the risk of another. Following this, attention focusses on any interventions that can enhance the recovery after stroke and maximising independence and quality of life. While signs and symptoms of stroke present their own weird and wonderful challenges, it is the parallel themes of spontaneous biological recovery, activity dependent rehabilitation and coping or transitioning to life after stroke that makes stroke rehabilitation a heavy proposition.
Acute Stroke Care leading the way
Acute stroke care is certainly moving in the right direction, (with the help of a hefty sum of financial support). Years of disappointment in the search for neuroprotective agents changed the focus toward a more proactive view of effective acute multidisciplinary care. Evidence shows that organised Acute Stroke Units improve both survival and recovery outcomes after stroke. No singular intervention or health discipline can claim any superiority in contributing to these favourable outcomes. This means that the benefits of acute stroke care is the cumulative effect of rapid assessment, streamlined investigations, prompt interventions, plus access to a specialised multidisciplinary team located in a designated space within a hospital. While some investigations such as access to CT/MRI and interventions such as thrombolysis remain quite specialised, many other key ingredients that are likely to contribute to favourable outcomes are wonderfully pragmatic. Examples of these ingredients include managing fever, blood sugar levels and ensuring early swallowing assessment (Middleton et al., 2011). Multiple Stroke Guidelines have been developed over the years that aim to provide explicit and accessible access to best evidence. Despite the simplicity of some of these interventions, clinicians and services still struggle to adhere to best practice – which is a great source of frustration! There is still on obvious divide between knowledge and implementation. The other challenge is to ensure best possible practice in acute care for every stroke patient, even when there is no access to a designated stroke services. This means more effective care for stroke in general medical wards, smaller rural health services and hospitals in lower income countries (Langhorne, de Villiers, & Pandian, 2012). Access to a multidisciplinary team with an interest in stroke care also allows for early rehabilitation. Early rehabilitation can provide several advantages. Prompt action to provide regular monitoring and interventions can prevent many complications such as infection, pain and pressure areas. Early intervention may also give an opportunity for early repair and re-organisation within the central nervous system. An example of an early interventions that provides a head start is body weight support treadmill training for non-ambulant patients, which can lead to more people becoming mobile after stroke, and improved mobility beyond the acute stage (Ada, Dean, Morris, Simpson, & Katrak, 2010; Dean et al., 2010).
Stroke Rehab – a mixed bag
Access to ongoing specialised multidisciplinary stroke rehabilitation improves outcomes, but accessing such services may well come down to luck (Enderby et al., 2016). Major questions remain how to best provide a quality service. Effective teamwork for any group of health professionals does not always come naturally and requires good leadership from many areas within the rehabilitation team. The old-school hierarchy medical model still lingers and may not always encourage the most effective interdisciplinary working. For those stroke survivors with rapid recovery, early discharge home can mean a quick shift to a community rehabilitation model which is usually either fragmented or non-existent. Rapid recovery and early discharge is obviously preferable, but dealing with any ongoing issues such as cognitive dysfunction, fatigue or depression can lead to significant ongoing issues, which can be difficult to manage in a more disjointed community setting.
Rehabilitation Potential – where is this heading?
Deciding on who would benefit from rehabilitation is becoming increasingly influenced by resource pressures. Even if a stroke service does employ quality health professionals who have patient priorities at heart, it is naive to think that the ‘selection’ for rehabilitation is not influenced by predicted length of stay and perceived ‘potential’ for improvement. Predicting rehabilitation ‘potential’ is often viewed as a dangerous game, as experienced clinicians know all too well that there are just so many variables to consider. Movement recovery is influenced by motor, sensory, visuospatial, psychological and other many other musculoskeletal complications and comorbidities. Predicting recovery is extremely difficult, no matter how experienced you think you are. Neuroscientists hope to assist in this process one day, with the addition of investigations and procedures that can more accurately measure the integrity of central nervous pathways needed for more optimal recovery (Stinear, Barber, Petoe, Anwar, & Byblow, 2012). This kind of research could be be viewed either with a glass half full or empty. Some hope that the ability to measure ‘potential’ would enable us to recognize potential in those people who would otherwise miss the opportunity receive therapy. Others may be concerned with the idea of a ‘self-fulfilling prophecy’ where there is a real belief that improvement cannot occur. This could lead to sub-optimal therapy, or even no access to therapy, with no opportunity for stroke survivors to engage in a process that could provide vital support and care. This is a prime example where evidence needs to be carefully placed into the context of a comprehensive, patient-centred rehabilitation model. At present, recovery research tells a consistent story – those that recovery well have the capacity respond to therapy – in other words those that recover better, do better! (Kitago & Krakauer, 2013) Belief mechanisms within the human psyche is very powerful, and can have an enormous negative impact on many health conditions such as chronic pain, chronic disability and depression. All stroke survivors have a right to know about specialized stroke rehabilitation. Perhaps predictive research should be complemented by research that examines access to therapy and true patient centered goal setting (McKenna, Martin, Jones, Gracey, & Lennon, 2015).
Compensation vs motor recovery – aim high, stay true.
Rehabilitation interventions for a number of stroke impairments inevitably lead into discussions about ‘recovery’ versus ‘compensation’(Levin, Kleim, & Wolf, 2009). In what circumstance does a speech therapist decide to move away from strategies to relearning expressive speech, to start a process that encourages alternative compensations such gesturing and communication technology? When does a physiotherapist or occupational therapist move away from encouraging and practicing arm and hand function recovery in the affected arm, and encourage an increased use of the unaffected arm for functional tasks? When does a physiotherapist decide that mobility can no longer progress to faster, more independent gait, and must remain a slow and careful gait with a quad stick? Of course goal setting influences all of these decisions, but the overarching stress of length of stay and safe discharge is always there. The unfortunate reality is that health professionals are often syphoned down a service delivery path where optimal recovery cannot realistically be reached, regardless of any ‘potential’. Movement quality can be more difficult to research, but while enhancing movement quality and performance may be preferable, it may not lead to measurable change. This may be due to a number of factors. Outcome measures used in research may not distinguish between new skills in movement performance and compensatory movements (Levin et al., 2009). In many circumstances, compensatory movements provide the meaningful functional change at a faster rate, which is often preferable. Changing movement quality and skill acquisition may take a little longer, and the challenge is determining reasons to train movement to another level. A proportion of people after stroke will have the capacity to change further and some will not. Hitching hips, trunk leans, major gait asymmetry, stiff hyperextending knees and internally rotated shoulders may well lead to pain and structural musculoskeletal changes sooner rather than later. Skilled and experienced therapists will help stroke survivors work toward preferred movement performance that fits with their goals and lifestyle – for the rest of their life!
Upper limb and hand recovery after stroke is great example where the requirement to tighten resources intersects with accumulating neuroscientific and clinical evidence. Over the past 15+ years there has been a growing field of neuroscientific research into the motor recovery of upper limb function. Functional MRI (Ward, 2015), Transcranial Magnetic Stimulation (Reis et al., 2008) and magnetoencephalography (Paggiaro et al., 2016) research is attractive reading for those of us interested in how the brain reorganizes itself after stroke. It gives us some useful information about how we can measure neuroplastic recovery processes. Brain stimulation also has given glimpses as potential adjunct to enhance active therapy. However, Its role in exciting or inhibiting areas of the brain with an aim of enhancing neuroplasticity has shown only small effects for hand function, although potential for reducing visuospatial neglect still looks promising (Müri et al., 2013). At the front line however, these interesting findings are not really having any direct influence on upper limb therapy, due to inaccessibility, cost and small effects (Ward & Kitago, 2016). We are still a long way of designing interventions based on this neuroscientific research, although the SEnse trial to improve for hand sensory function is a notable exception (Carey, Macdonell, & Matyas, 2011). Most forms of upper limb therapy are based on practice based experience, observation, motor control theories and are almost impossible to categorise and define. Constraint Induced movement Therapy (CIMT) stemmed from behavioural neuroscience experiments with macaque monkeys, based on the concept of learned non-use. Movement analysis reveals that upper limb recovery can be a mix of compensatory movements as well as relearning some movements again – both neuroplastic processes. It is disappointing that so many neurological therapists still show fanatical tendencies, and firmly plant themselves in quite vague categories or ‘approaches’. Defining a therapy intervention can be difficult enough, and methodological flaws in research reflect this. Current evidence suggests these differing ‘approaches’ are no more or less effective when it comes to stroke rehabilitation (Pollock et al., 2014). More clearly defined therapy protocols of CIMT and some task specific practice allow for some useful information to be extrapolated. Even these these protocols involved a blend of elements may all contribute to recovery – such as forced-use, repetitive practice, behaviour modification and intensity (Taub et al., 2013). Mixed approaches using a variety of verbal, tactile, passive and active movements, or those that intend to encourage motor priming (Stoykov & Madhavan, 2015) are very difficult to research, and will require further movement analysis. Movement analysis is any area of growing research, and is worth following as it may well help in identifying key therapeutic targets for movement retraining. Gait analysis has helped us identify key impairments after stroke (Arene & Hidler, 2015), and more research is needed with regard upper limb movement kinematics that will help in identifying key movements to aid upper limb function (more than just reach & grasp!) (Murphy & Häger, 2015). Despite this, upper limb therapy evidence remains quite limited. It has been suggested that upper limb and hand recovery after stroke may be more influenced by spontaneous biological recovery rather than any activity-dependent therapy intervention! Maybe in many circumstances therapists are just ‘overseeing and monitoring’ natural recovery. Of course, this is debateable as many interventions may be critical to stimulate recovery, enhance body schema and spatial awareness, coach postural control, encourage arm use and prevent complication such as contracture in key muscle groups and shoulder pain. Remember the brain is plastic, but the body is not! Again, people whose upper limbs recover better, do better! Animal studies indicate that spontaneous biological recovery is likely limited to a certain period after stroke, possibly within the first 3 months (Kitago & Krakauer, 2013). Within this, there may even be a critical period where neuroplastic potential is heightened. Therapy may be able to capitalise on this period, so further research such as the current collaborative work currently being undertaken by South Australian and London researchers will be worth following (McDonnell et al., 2015).
Is timing everything?
Often stroke interventions research include chronic stroke patients beyond 6 months, to demonstrate that any change is more likely due to the intervention, rather than spontaneous biological recovery. With the disappointing results of many stroke rehabilitation interventions that show small effects, it may well be worth considering interventions that start earlier that enhance the effects of spontaneous recovery and push into later subacute periods where rehabilitation progress is often disrupted by discharge (Stinear, 2016). The AVERT trial is an extremely impressive example of multicentre research in the acute phase. Results indicate that very intensive early mobilisation may not be suitable for many people after stroke (AVERT Trial Collaboration group et al., 2015). Perhaps the brain in the immediate phase following stroke has certain limitations and vulnerabilities that need considering. This may also explain why early, intense CIMT may also lead to less favourable outcomes (Dromerick et al., 2009). However, the EXCITE trial, another impressive multicentre trial involving sub-acute patients 3-9 months post stroke, showed some longer lasting benefits of CIMT for upper limb function that persist at 12-months (Wolf et al., 2006). Body weight support treadmill training for non-ambulant acute stroke patients has some clear benefits leading to more people recovering mobility at 12 months (Dean et al., 2010). Rehabilitation timing is likely to be very important, and combining interventions at certain times may be worth careful consideration with future research. There are many factors influenced by time and include, deconditioning, weakness, fitness, contracture and length of stay financial pressures.
Overdosed on evidence
While some say ‘timing is everything’, the biggest theme more recently has been ‘more is better’ (Lang et al., 2016). An intensive dose of practice is considered by many to be a necessary requirement for neuroplastic change. Constraint induced protocols for speech and upper limb function are very intensive programs that provide the strongest evidence to date. One of the main criticisms of the original CIMT upper limb protocol was the feasibility of such an intense 6-hour a day program. Intensity and ‘dose’ however are not the one in the same. The dose of task specific practice within CIMT is likely to be quite important, but the almost relentless daily reflection of upper limb use appears to be an effective form of behaviour modification, to increase upper limb use (Taub et al., 2013). Recent investigation investigating the dose of task specific practice suggests that the dose effect for this form of therapy will plateaux, which reminds that type and intensity of therapy may need to be examined further (Lang et al., 2016). It is not all about repetitions! We now must consider more innovative ways to ‘intensify’ task specific practice. Increasing sensorimotor feedback combined with attention, shaping and task specific tasks may further enhance plasticity. Sensory stimulation, facilitation techniques, robotics, supports, electrical stimulation, brain stimulation, virtual reality, imaging techniques, aerobic training, strength training and even pharmacological assistance with SSRIs (Mead et al., 2013) have the potential to prime the brain and gain more benefit from therapy. Most of these techniques complement each other and can be quite easily combined with part or whole task specific practice. What is certain – something must change (Bernhardt et al., 2016; Stinear, 2016)! In addition, many of these methods may provide some alternative for the majority of people with more severely affected upper limbs who cannot undertake many of the currently recommended task specific protocols.
Where to from here?
It is hard to predict where stroke rehabilitation will be in the next few decades. There are certainly some priorities to look at;
Education – practice is not following the evidence, particularly best practice guidelines. This is a fact. Something should change in the educational sector to ensure therapists are confident and equipped to perform adequate therapy. Perhaps less lectures and more practical sessions to run through a checklist of competencies? Many experienced therapists find themselves as managers or even academics, where paperwork takes over. Structured mentorship and clinical supervision should be mandatory. An audit feedback process is important, and guidelines may need to be more specific (and less vague!) (Janzen, McIntyre, Richardson, Britt, & Teasell, 2016).
Rehabilitation priorities – combining interventions, starting earlier and lasting longer (Stinear, 2016). What are the best outcomes to measure for people with more severe disabilities? How can we prevent these people from missing out on important care? We need to push and promote neurorehab, especially aspects such as cost-effectiveness (Turner-Stokes, Williams, Bill, Bassett, & Sephton, 2016).
Technology – we have an appetite for technology and its potential is enormous! Are robots cost-effective and useful? Does Virtual Reality improve motivation and create and environment for enhanced learning? How do we research or purchase new technology, even though it changes so fast? Staff training required? Maintenance costs? Safe to use at home? Does it need to be prescribed? So many questions! Tele rehabilitation needs more research, but could be a game changer (Laver et al., 2013). One thing is for sure – we will be using technology more, and it will change practice.
Living with stroke – once you survive a stroke, it is the transition to living with a disability for many years that may be the real priority (Crichton, Bray, McKevitt, Rudd, & Wolfe, 2016). Community rehabilitation must get creative to evolve and survive (Teasell et al., 2012). Long term health issues (Towfighi, Markovic, & Ovbiagele, 2012), complications and shifting priorities over time means that self-management (McKenna et al., 2015), goals setting, accessibility and support will need to be available from multiple services in the community (Rimmer, 2012). Community rehabilitation is the next frontier!
Stroke motor recovery research is moving in many directions. Let’s keep up the pressure and steer research and service delivery in the right direction to ensure a better life for those people who survive a stroke.
Associate Professor James McLoughlin
BAppSc (Physio), MSc (Clinical Neuroscience), PhD
James is an experienced neurological physiotherapist, and has been lecturer in stroke rehabilitation at Flinders University for over 10 years. He also Director at neurophysiotherapy.com.au
Ada, L., Dean, C. M., Morris, M. E., Simpson, J. M., & Katrak, P. (2010). Randomized trial of treadmill walking with body weight support to establish walking in subacute stroke: the MOBILISE trial. Stroke; a Journal of Cerebral Circulation, 41(6), 1237–1242.
Arene, N., & Hidler, J. (2015). Understanding motor impairment in the paretic lower limb after a stroke: a review of the literature. Topics in Stroke Rehabilitation. Retrieved from http://www.maneyonline.com/doi/abs/10.1310/tsr1605-346
AVERT Trial Collaboration group, Bernhardt, J., Langhorne, P., Lindley, R. I., Thrift, A. G., Ellery, F., … Donnan, G. (2015). Efficacy and safety of very early mobilisation within 24 h of stroke onset (AVERT): a randomised controlled trial. The Lancet, 386(9988), 46–55.
Bernhardt, J., Borschmann, K., Boyd, L., Thomas Carmichael, S., Corbett, D., Cramer, S. C., … Ward, N. (2016). Moving rehabilitation research forward: Developing consensus statements for rehabilitation and recovery research. International Journal of Stroke: Official Journal of the International Stroke Society, 11(4), 454–458.
Carey, L., Macdonell, R., & Matyas, T. A. (2011). SENSe: Study of the Effectiveness of Neurorehabilitation on Sensation A Randomized Controlled Trial. Neurorehabilitation and Neural Repair. Retrieved from http://nnr.sagepub.com.ezproxy.flinders.edu.au/content/25/4/304.short
Crichton, S. L., Bray, B. D., McKevitt, C., Rudd, A. G., & Wolfe, C. D. A. (2016). Patient outcomes up to 15 years after stroke: survival, disability, quality of life, cognition and mental health. Journal of Neurology, Neurosurgery, and Psychiatry. https://doi.org/10.1136/jnnp-2016-313361
Dean, C. M., Ada, L., Bampton, J., Morris, M. E., Katrak, P. H., & Potts, S. (2010). Treadmill walking with body weight support in subacute non-ambulatory stroke improves walking capacity more than overground walking: a randomised trial. Journal of Physiotherapy, 56(2), 97–103.
Dromerick, A. W., Lang, C. E., Birkenmeier, R. L., Wagner, J. M., Miller, J. P., Videen, T. O., … Edwards, D. F. (2009). Very Early Constraint-Induced Movement during Stroke Rehabilitation (VECTORS): A single-center RCT. Neurology, 73(3), 195–201.
Enderby, P., Pandyan, A., Bowen, A., Hearnden, D., Ashburn, A., Conroy, P., … Winter, J. (2016). Accessing rehabilitation after stroke–a guessing game? Disability and Rehabilitation, 1–5.
Janzen, S., McIntyre, A., Richardson, M., Britt, E., & Teasell, R. (2016). Building a Knowledge to Action Program in Stroke Rehabilitation. The Canadian Journal of Neurological Sciences. Le Journal Canadien Des Sciences Neurologiques, 43(5), 619–625.
Kitago, T., & Krakauer, J. W. (2013). Motor learning principles for neurorehabilitation. Handbook of Clinical Neurology, 110, 93–103.
Lang, C. E., Strube, M. J., Bland, M. D., Waddell, K. J., Cherry-Allen, K. M., Nudo, R. J., … Birkenmeier, R. L. (2016). Dose response of task-specific upper limb training in people at least 6 months poststroke: A phase II, single-blind, randomized, controlled trial. Annals of Neurology, 80(3), 342–354.
Langhorne, P., de Villiers, L., & Pandian, J. D. (2012). Applicability of stroke-unit care to low-income and middle-income countries. Lancet Neurology, 11(4), 341–348.
Laver, K. E., Schoene, D., Crotty, M., George, S., Lannin, N. A., & Sherrington, C. (2013). Telerehabilitation services for stroke. Cochrane Database of Systematic Reviews , (12), CD010255.
Levin, M. F., Kleim, J. A., & Wolf, S. L. (2009). What do motor “recovery” and “compensation” mean in patients following stroke? Neurorehabilitation and Neural Repair, 23, 313–319.
McDonnell, M. N., Koblar, S., Ward, N. S., Rothwell, J. C., Hordacre, B., & Ridding, M. C. (2015). An investigation of cortical neuroplasticity following stroke in adults: is there evidence for a critical window for rehabilitation? BMC Neurology, 15, 109.
McKenna, S., Martin, S., Jones, F., Gracey, J., & Lennon, S. (2015). The Bridges Stroke Self-Management program for Stroke Survivors in the Community: Stroke, Carer and HealthProfessional Participants’ Perspectives. Physical Medicine and Rehabilitation-International, 2(1), 1030–1036.
Mead, G. E., Hsieh, C.-F., Lee, R., Kutlubaev, M., Claxton, A., Hankey, G. J., & Hackett, M. (2013). Selective serotonin reuptake inhibitors for stroke recovery: a systematic review and meta-analysis. Stroke; a Journal of Cerebral Circulation, 44(3), 844–850.
Middleton, S., McElduff, P., Ward, J., Grimshaw, J. M., Dale, S., D’Este, C., … Levi, C. (2011). Implementation of evidence-based treatment protocols to manage fever, hyperglycaemia, and swallowing dysfunction in acute stroke (QASC): a cluster randomised controlled trial. The Lancet, 378(9804), 1699–1706.
Müri, R. M., Cazzoli, D., Nef, T., Mosimann, U. P., Hopfner, S., & Nyffeler, T. (2013). Non-invasive brain stimulation in neglect rehabilitation: an update. Frontiers in Human Neuroscience, 7, 248.
Murphy, M. A., & Häger, C. K. (2015). Kinematic analysis of the upper extremity after stroke – how far have we reached and what have we grasped? Physical Therapy Reviews: PTR, 20(3), 137–155.
Paggiaro, A., Birbaumer, N., Cavinato, M., Turco, C., Formaggio, E., Del Felice, A., … Piccione, F. (2016). Magnetoencephalography in Stroke Recovery and Rehabilitation. Frontiers in Neurology, 7, 35.
Pollock, A., Baer, G., Campbell, P., Choo, P. L., Forster, A., Morris, J., … Langhorne, P. (2014). Physical Rehabilitation Approaches for the Recovery of Function and Mobility After Stroke: Major Update. Stroke; a Journal of Cerebral Circulation, 45(10), e202–e202.
Reis, J., Swayne, O. B., Vandermeeren, Y., Camus, M., Dimyan, M. A., Harris‐Love, M., … Cohen, L. G. (2008). Contribution of transcranial magnetic stimulation to the understanding of cortical mechanisms involved in motor control. The Journal of Physiology, 586, 325–351.
Rimmer, J. H. (2012). Getting beyond the plateau: bridging the gap between rehabilitation and community-based exercise. PM & R: The Journal of Injury, Function, and Rehabilitation, 4(11), 857–861.
Stinear, C. M. (2016). Stroke rehabilitation research needs to be different to make a difference. F1000Research, 5. https://doi.org/10.12688/f1000research.8722.1
Stinear, C. M., Barber, P. A., Petoe, M., Anwar, S., & Byblow, W. D. (2012). The PREP algorithm predicts potential for upper limb recovery after stroke. Brain: A Journal of Neurology, 135(Pt 8), 2527–2535.
Stoykov, M. E., & Madhavan, S. (2015). Motor priming in neurorehabilitation. Journal of Neurologic Physical Therapy: JNPT, 39(1), 33–42.
Taub, E., Uswatte, G., Mark, V. W., Morris, D. M., Barman, J., Bowman, M. H., … Bishop-McKay, S. (2013). Method for enhancing real-world use of a more affected arm in chronic stroke: transfer package of constraint-induced movement therapy. Stroke; a Journal of Cerebral Circulation, 44(5), 1383–1388.
Teasell, R., Mehta, S., Pereira, S., McIntyre, A., Janzen, S., Allen, L., … Viana, R. (2012). Time to rethink long-term rehabilitation management of stroke patients. Topics in Stroke Rehabilitation, 19(6), 457–462.
Towfighi, A., Markovic, D., & Ovbiagele, B. (2012). Impact of a healthy lifestyle on all-cause and cardiovascular mortality after stroke in the USA. Journal of Neurology, Neurosurgery, and Psychiatry, 83(2), 146–151.
Turner-Stokes, L., Williams, H., Bill, A., Bassett, P., & Sephton, K. (2016). Cost-efficiency of specialist inpatient rehabilitation for working-aged adults with complex neurological disabilities: a multicentre cohort analysis of a national clinical data set. BMJ Open, 6(2), e010238.
Ward, N. S. (2015). Does neuroimaging help to deliver better recovery of movement after stroke? Current Opinion in Neurology, 28(4), 323–329.
Ward, N. S., & Kitago, T. (2016). Getting the right prescription for rehabilitation after stroke. Neurology. https://doi.org/10.1212/WNL.0000000000002653
Wolf, S. L., Winstein, C. J., Miller, J. P., Taub, E., Uswatte, G., Morris, D., … Others. (2006). Effect of constraint-induced movement therapy on upper extremity function 3 to 9 months after stroke: the EXCITE randomized clinical trial. JAMA: The Journal of the American Medical Association, 296(17), 2095–2104.