When does stroke rehabilitation end?


  • Nicol Korner-Bitensky

    Corresponding author
    1. Centre de recherche interdisciplinaire en réadaptation du Montréal métropolitain (CRIR), Montreal, QC, Canada
    • Faculty of Medicine, School of Physical and Occupational Therapy, McGill University, Montreal, QC, Canada
    Search for more papers by this author

  • Conflict of interest: None declared.

Correspondence: Nicol Korner-Bitensky, School of Physical and Occupational Therapy, McGill University, 3630 Promenades Sir-William-Osler, Montreal, QC H3G 1Y5, Canada.

E-mail: nicol.korner-bitensky@mcgill.ca



This article examines key evidence on intervention effectiveness late poststroke; provides discussion on how this evidence impacts stroke rehabilitation at a clinical and national level; and explores strategies that should improve the way in which chronic stroke is addressed internationally.


Although the evidence is still small, early intensive intervention is increasingly recognized as a ‘good thing’. The only caveat, as suggested by the recent constraint-induced movement therapy (CIMT) trial, is that therapy that is too intensive, too early, might overwhelm the recovering brain [1]. Those receiving high-intensity CIMT had significantly less upper limb motor function improvement at 14 and 90 days compared with those receiving less intensive CIMT or traditional upper extremity (UE) therapy [1].

On the other end of the spectrum, the conversation and science about when does stroke rehabilitation end has rarely been addressed in clinical practice other than by patients and families who plead for continued rehabilitation while the experts deem that the plateau has occurred.

This commentary poses the question: What is stroke rehabilitation, and when does it end? The first part of the question is easier to answer than the second. The primary goal of early stroke rehabilitation can be succinctly summarized as:

  • preventing poststroke complications
  • minimizing impairments; and
  • maximizing function.

Success is often measured in the proportion of individuals returning home, or in functional gains occurring in the acute and sub-acute period. We have also come to recognize that successful stroke recovery includes resuming and sustaining community life, occupations, and meaningful activities. As we attempt to answer the second part of the question, ‘when does rehabilitation end?’, it is worthwhile to take a look at some of the myths that have driven the rehabilitation world.

For decades, the stroke rehabilitation community believed that the window of opportunity in which to provide rehabilitation was restricted to the first three-months poststroke [2]. What was gained in terms of motor control and function during that period was likely to be ‘as good as it gets’ in terms of recovery. The scientific evidence emerging over the past decade tells us that our perception of the three-month plateau is false. Indeed, some of the most exciting evidence in stroke plasticity and motor recovery has focused on what we now know to be a misnomer, the chronic stroke patient. Even for the most skeptical among us, it is hard to dispute the mounting evidence, both in quality and quantity, suggesting that rehabilitation interventions are effective in improving patient outcomes at both the impairment level and functional level late after stroke. In this invited commentary, I examine key evidence on intervention effectiveness late poststroke and provide a brief discussion on how this evidence impacts stroke rehabilitation at a clinical and national level. Finally, insights are provided on strategies that should improve the way in which chronic stroke is addressed.

How strong is the evidence?

As an international stroke rehabilitation research community, we have mounting evidence on the effectiveness of interventions offered in the chronic phase. Major trials such as the EXCITE [3], VA Robotics [4], and LEAPS [5] have shown that it is possible, not only to maintain function late poststroke, but also to actually reverse the downhill course. Here, I present some of the highlights of what we have come to know based on a decade of research in stroke rehabilitation focused on the chronic/late poststroke period.

Aerobic exercise – can it be done safely and does it work?

In the mid-1980s, as we began randomized clinical trials (RCTs) on treadmill training, we deliberated about the safety of intensive training; ever mindful of a concern that this was a largely fragile and compromised group of patients [6]. In the last decade, numerous studies have challenged this cautionary approach by testing the effectiveness of various forms of aerobic exercise (cycling, aquatic therapy, and treadmill training) offered in a ‘cocktail format’ (e.g., along with strength and flexibility training). For example, four high-quality RCTs [7-10] provide strong evidence (Level 1a) that aerobic exercise offered in a variety of formats significantly improves aerobic capacity in patients with chronic stroke when compared with a range of control therapies. In addition, there is moderate evidence (Level 1b) that aerobic exercise improves femoral neck bone mineral density on the paretic side [8], an especially encouraging finding given the high rate of falls and corresponding fracture risk in those with stroke [11, 12].

Constraint-induced movement therapy

The effectiveness of CIMT and modified CIMT (mCIMT) has been extensively studied in the chronic phase. Evidence from three high-quality RCTs [13-15] and five fair-quality RCTs [16-20] provide strong evidence (Level 1a) that mCIMT is more effective than conventional rehabilitation or no treatment for improving UE motor function in some patients with chronic stroke. There is also moderate evidence (Level 1b) from one high-quality RCT [21] that CIMT is more effective than control therapies (e.g., bilateral therapy; physical, cognitive, and relaxation exercises) for improving UE motor function. The question arises as to how we can begin to offer such intense therapy in many of our healthcare systems where the opportunity for poststroke rehabilitation in the chronic stage, if available, is unlikely to be offered at the required intensity. A recent study by Henderson and Manns [22] suggest one potential strategy to increase the use of mCIMT in daily practice – group-modified constraint-induced movement therapy. As well, Smania et al. [23] have tested the effectiveness of a reduced intensity mCIMT.

Improving executive function

If we move on to other important sequelae of stroke such as executive function, it is interesting to note that while there are only nine intervention studies (including single-subject designs, pre-post designs, and one RCT) focused on the chronic phase, seven have been published in just the last seven-years. They focus on various approaches to improving outcomes such as working memory: two used a remedial approach for improving working memory [24, 25]; four used strategy training in problem solving, planning, multitasking, and goal management [25-28], whereas the remaining three [29-31] relied on external compensatory approaches such as external cueing systems or checklists. Although the evidence from these studies is limited (Level 2a), it does suggest that working memory training compared with no intervention results in better working memory. There is also preliminary indication of generalization to everyday functioning. As well, there is limited evidence suggesting that strategy training in problem solving using various formats is more effective than no intervention at improving executive functioning and, possibly, everyday life skills [29-32]. Given the serious impact of executive disorders, these encouraging findings suggest the need for opportunities to receive rehabilitation aimed at enhancing executive functioning long after stroke.

Improving communication in poststroke aphasia

High-quality RCTs have shown the impact of various interventions aimed at improving communication in those with aphasia late poststroke. There is moderate evidence (Level 1b) from one high-quality RCT that communication groups are more effective than no therapy in improving language abilities in patients with chronic aphasia [33]. Similarly, a high-quality RCT [34] has investigated the effect of communication partners on communication skills among volunteers or caregivers of patients with aphasia in the chronic phase. Communication partner training was found to be more effective than no therapy in improving patient interaction and transaction skills. These are particularly interesting results given they use creative intervention solutions including the training of laypersons to increase the level of treatment intensity.

Where do we go with the evidence that stroke rehabilitation is effective late after stroke?

Twenty years has passed since Widén-Holmqvist et al. [35] posed the question of whether there is a basis for rehabilitation late after stroke. In response to the question ‘Do we have sufficient evidence to suggest that specific rehabilitation interventions are effective at both the impairment and function level when provided to those in the chronic phase of stroke?’, the answer is an overwhelming ‘yes’.

The most important question then is ‘where do we go with this knowledge?’ Foremost, we must reposition the management of stroke to recognize the need for sophisticated chronic disease management for this clientele. As a pragmatic strategy that can be offered at a national level, we likely require market scans within each nation to identify existing community programs that focus on exercise and health promotion programs that could potentially incorporate those with stroke in a safe and welcoming manner. Building national initiatives that include training of those working in health promotion and community fitness to allow them to meet the needs of those with stroke, where these programs are recognized and rewarded for their efforts, is likely a realizable and concrete initiative at a large-scale level.

Although policymakers, clinicians, and researchers can and should advocate for those with chronic stroke who require services, it is imperative that the millions of individuals with stroke and their families demand needed services. Thus, another strategy that is worth capitalizing on is the power of social media to help advocate for services late poststroke. Indeed, in contrast to the stroke consumer of a decade ago, the new-age consumer is not relying solely on their therapist to supply information about effective stroke interventions. Rather, knowledge about treatment effectiveness late poststroke, disseminated broadly through social media, should help drive the demand for services in this group. Major initiatives that use Internet information sharing are already underway in many countries, and these can be used to provide accurate evidence-based information to inform and empower people with stroke and their families to seek intervention late poststroke [36-44].

Another key strategy that we as an international community can embark on with a singular vision is that, in the next round of national guidelines, or preferably one international common stroke guideline, we include recommendations on the effectiveness of interventions according to stage of stroke recovery (early/late) and level of stroke severity. The clear evidence gathered from the hundreds of studies that have examined the benefit of various interventions in stroke rehabilitation allows us to move toward this more sophisticated approach to patient management, one that is based on the best evidence according to stage of poststroke recovery and patient-specific profile.


In conclusion, as an international stroke community of researchers and clinicians, we are focused on providing the best health care for those with stroke. In reality, no country is likely to have a system in place to provide intensive, evidence-based stroke management beyond the very narrow window of the first weeks poststroke. Together we can share successes and innovations that will allow us to narrow the gap between what we offer and what we should be offering to individuals with stroke given our strong evidence-based knowledge regarding treatment effectiveness.