Article Text


Physiotherapy after stroke: define, divide and conquer
  1. J Marsden1,
  2. R Greenwood2
  1. 1Sobell Department of Movement Neurosciences, Institute of Neurology, Queen Square, London WC1, UK
  2. 2Regional Neurological Rehabilitation Unit, Homerton Hospital, London E9, and Acute Brain Injury Service, National Hospital for Neurology and Neurosurgery, Queen Square, London WC1, UK
  1. Correspondence to:
 Dr R J Greenwood
 RNRU, Homerton University Hospital, Homerton Row, London E9 6SR, UK;

Statistics from

Assessment of physiotherapy is difficult because of the complexity of the interventions

When rehabilitating people after stroke, physiotherapists often favour a particular “approach”. An “approach” is a theoretical construct based on a series of ideas and hypotheses about the retraining of movement that influences the content, structure, and aims of a therapy session.1 In this edition of the journal (see pp 503–8), van Vliet et al2 compare two approaches, “motor re-learning” and the “Bobath concept”, the latter presently being the preferred approach within the United Kingdom.1 The emphasis of “motor re-learning” is on context specific functional training using principles derived from motor learning theory, while that of the Bobath concept is on observational analysis and facilitation of normal movement using principles derived from neurodevelopmental and neurophysiological theory.

van Vliet and colleagues have previously shown that therapy using these two approaches differs, for example in the degree and type of feedback provided and extent to which everyday objects are used during the treatment session.3 Their single blind randomised controlled trial,2 which admitted 120 patients within two weeks of stroke, has now shown no differences in sensory-motor impairments or functional outcome between the two interventions up to six months after treatment onset. Whether these findings are the result of similarities between the two approaches far outweighing their differences or for other reasons is unclear and worthy of further exploration.


As van Vliet and colleagues observed,2 interpreting a study of this sort is made difficult by the variability between studies, not only in the content of the intervention itself but also in the context in which the intervention occurs. Physiotherapy interventions of this sort are complex and are “made up of various interconnecting parts.”4 If further studies of this sort are to be done, it will be important to define the practice protocols for each intervention rigorously, in particular how the type and difficulty of tasks used within a treatment session are chosen, and the ways in which treatment is progressed; observational analysis of the different interventions is an important step in this direction.1,3

Detailed definition of intervention protocols would facilitate study reproduction and comparison, identify the similarities and differences between interventions, and enable investigation of the effects of other variables such as the onset and intensity of treatment. Intensity of treatment is not necessarily the same as the duration of treatment, as an equivalence of treatment time between interventions may not guarantee an equivalence of treatment intensity. Future protocols should not only aim to define the components of a treatment session but also to define how the practice session is scheduled as this may be important in optimising the learning of a motor task. Factors such the type and frequency of feedback and whether different tasks are practised consecutively or their practice order randomised may affect the rate and extent to which a task is learnt, as well as the extent to which it may generalise to novel environmental situations both in health and after stroke.5,6

Given the wide differences in the severity and types of functional deficit after stroke, and the ever evolving ethos of the different approaches, formulating such protocols will be difficult. In addition, the other interventions and activities that occur outside the daily ∼30 minute therapy session may have a significant impact, to the extent that differences in the efficacy of one physiotherapy approach over another may be swamped by the similarities in activities that occur for the remaining ∼98% of the time. How studies can avoid this “noise” needs exploration.


The stroke population studied by van Vliet was heterogeneous, reflecting that seen in clinical practice. Such a holistic approach may overlook some of the intricacies of an approach. The different components of each approach are largely based on the theoretical construct with which it is underpinned; the effects of each component have been explored very little even in small proof of principle studies. The efficacy of different approaches and their components may be dependent on stroke pathology and the presenting impairments. Furthermore, a more detailed biomechanical assessment may reveal differences in the effects of two approaches. Although such differences may not be regarded as important unless they are revealed in snapshots of functional outcome, differences in movement patterns, and in particular movement efficiency, may be important for the ability to perform a task repetitively, as is required for everyday functioning, or for the gradual acquisition of a skill over a protracted period.

Thus two research strands are required to improve our current physiotherapy practice: first a top down, holistic approach, in which the practice content, schedule, and intensity are defined within the context of an overall standardised treatment package; second, a bottom up, reductionist approach where individual components of an intervention are assessed on specific patient populations in a hypothesis driven manner. As highlighted by van Vliet et al,2 one disadvantage of a reductionist approach is that it neglects potentially important interactions between treatment techniques, emphasising the need for a combined approach.

Assessment of physiotherapy is difficult because of the complexity of the interventions


View Abstract


  • Competing interests: none declared

Request permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Linked Articles