How do we know whether patients have learned anything from what they practised? A ‘learning curve’ is a graphical representation of performance over time. Typically, performance on an outcome measure (Y-axis) is plotted at specific points in time (X-axis). Figure 8.1 presents a learning curve from a classic study by Stelmach in 1969, where performance was measured as the number of rungs climbed on the Bachman ladder (a novel ladder with unequally distributed rungs on either side of a vertical beam in the middle of the ladder). The purpose of the experiment was to compare the effectiveness of two different training regimes; ‘distributed practice’ (i.e. practice interspersed with rest, with more rest than practice) and ‘massed’ (i.e. more practice than rest).

Fig. 8.1 • The learning curve from Stelmach?s classic experiment on learning the Bachman ladder task (1969), with permission from the American Alliance for Health, Physical Education, Recreation and Dance.

Looking at figure 8.1, which schedule was more effective? Clearly, the distributed practice schedule resulted in a higher number of rungs climbed, compared to the massed practice schedule. Thus, motor performance appeared to be better following distributed practice.

In order to assess the effect of practice on learning, performance is usually tested again after a delay, with either a ‘retention test’ or a ‘transfer test’. A retention test occurs some time (minutes, days or weeks) after the practice has been completed, with the aim to determine how much the individual has retained. In the retention test, the task that has been practised is the same as the one that is assessed. A transfer test is different, in that the task that is assessed is similar but not identical to the task that has been practised. The purpose of a transfer test is to determine to what extent the person can transfer (carry over) what they have learned to a problem that is slightly different.

In Stelmach’s study, a retention test was used, following a short delay (4 minutes only). It is interesting to note that at the retention test, there is now hardly any difference between the two practice schedules (Fig. 8.2). What happened in that short period of time? Imagine yourself undertaking the Bachman ladder task under the massed practice schedule: stepping up and down as fast as you can, again and again. It is likely that fatigue would set in and impair your performance: your leg extensors would fatigue quickly with the fast concentric and eccentric contractions. However, after a short break the increased muscle tension reduces, circulation improves, your attention may also become more focused, and altogether you are able to deliver a much better performance.

Fig. 8.2 • The learning curve from the same experiment as Fig. 8.1 (Stelmach1969), with permission from the American Alliance for Health, Physical Education, Recreation and Dance.

Based on this classic, and similar experiments, it became clear that a distinction had to be made between ‘performance’ and ‘learning’. Motor performance is defined as:

‘the behavioural act of performing a skill at a specific time and in a specific situation.’

(Magill (2007, p. 440)

In contrast, motor learning can be defined as:

‘a set of processes associated with practice/ experience leading to relatively permanent changes in the capability for producing skilled action.’

(Schmidt, 1988, p. 346)

Three points are worth noting in the motor learning definition:

1. Practice or experience is required, i.e. an initial successful attempt would not be defined as a learned action stored in memory

2. Changes in performance are relatively permanent

3. Motor learning is concerned with skilled action, i.e. goal-oriented behaviour, but not reflex-like behaviour, which is largely automatic.

What is the moral of the study by Stelmach? The important point to take away is that one should not jump to conclusions about the effectiveness of a specific training schedule, if performance has only been tested immediately at the end of the training. The clinical equivalent would be to use an outcome measure at the start and immediately after practising a specific activity, and base conclusions about the effectiveness of the intervention on just this outcome. In this case, the clinician will only have tested the patient’s performance. Patients may appear to perform well following the end of an intervention, but effects may not be sustained without further practice. The need to include a follow-up after the end of an intervention is illustrated in Research box 8.1 below on cueing in patients with Parkinson’s disease (PD).

Research box 8.1 Short- and longer-term effects of cueing on the coordination of gait in Parkinson’s disease

The Rehabilitation in Parkinson’s Disease: Strategies for Cueing (ResCUE) trial, the first large scale randomised trial on cueing in Parkinson’s disease (PD), was designed to examine the effects of a home-based cueing programme on gait parameters, gait-related activity and health-related quality of life in people with PD (Nieuwboer et al. 2007). In this randomised cross-over trial, a total of 153 people with PD participated in a cueing intervention, which comprised a 3-week programme with nine treatment sessions of 30 minutes each. In the first week, patients tried auditory, visual as well as somatosensory cueing modalities, i.e. beeps administered via an earpiece, flashes from an LED mounted on a pair of glasses, and pulsed vibrations applied via a wristband-mounted device. Having tried each of these modalities, patients then continued training with the one they preferred. Most patients (67%) selected the auditory, while the remainder chose the somatosensory modality. While being instructed to time their heel strike according to the rhythm provided and to continue stepping through turning and other transitions, patients undertook a variety of tasks. These included starting and stopping walking, sideways and backward stepping, walking while doing another task (i.e. dual tasking), and walking outside over different surfaces. More information about the specific cueing intervention and the underlying evidence is detailed on the ReSCUE project CD-ROM (2002/3,http://www.rescueproject.org/). Depending on group allocation, this intervention was either preceded or followed by a 3-week period without any training, after which a 6-week follow-up period ensued.

The primary outcome measure was a composite score of five items on the Unified Parkinson’s Disease Rating Scale, reflecting gait and balance. Secondary outcomes included measures of gait and balance, functional activity, falls-related self-efficacy, participation and carer strain. All outcome measures were assessed – without patient using the cueing device – by a blinded assessor in weeks 3, 6 and 12.

Straight after the intervention, the results demonstrated significant improvements in the primary outcome measure of 4.2%, while median gait speed improved by 5 cm/s and median step length improved by 4 cm. Significant improvements were also found in balance tests, while self-reported confidence in gait-related activities had increased. A significant reduction of freezing episodes was found in so-called ‘freezers’ only. There were no significant changes in any of the other outcomes. However, at the 12-week follow-up assessment, most outcomes had deteriorated significantly compared to those after the intervention at week 6.

What do these findings mean for clinical practice? The results from this trial demonstrated that a 3-week programme of nine cueing sessions can result in significant – albeit small – improvements in specific aspects of gait and balance in people with mild to moderate PD. Interestingly, the effects were noted without patients wearing their cueing device, which indicates that some carry-over had taken place. However, although these effects were seen immediately after the intervention, they were not retained at follow-up. This suggests that the effects of cueing were more due to performance than learning, with the clinical implication that continuing use of cueing may be required to sustain its effects.

We mentioned earlier that some factors may mask learning. For example, in neurological rehabilitation, pain and fatigue are common (Chapter 14). Physical practice may exacerbate pain and fatigue, which in turn may have a detrimental effect on concentration. Patients may even get bored with an activity! All these factors may affect performance. So how may a therapist obtain a more accurate picture of what a patient has actually learned? The evidence from experiments such as that by Stelmach (1969) shows that it is essential to:

1. undertake a retention test sometime after the end of the intervention. This will help reduce any practice-related effects on pain, attention and fatigue and show how much has actually been retained in memory. Assessing performance immediately after an intervention will merely show performance effects.

2. utilise a transfer test, i.e. a slight variation from the task that has been practised. This will reveal to what extent the patient is able to adapt what has been learned. This means that the outcome measure must be related but not identical to the tasks that were practised in the intervention.

One point worth making here is that it is not advisable to change established, standardised outcome measures for the sake of a transfer test, as this would invalidate the measure. However, if one was using a simple measure such as time taken to stand up, the standing up movement could be done slightly differently on the transfer test, e.g. by using a different chair, placed on a different surface.

The transfer test (or task) is a crucial concept in clinical practice, as rehabilitation programmes – which can’t possibly incorporate all potential problems a patient may encounter after discharge – are designed on the assumption that what patients learn during treatment will carry over to their own activities of daily living in their own environment. This is more of a challenge than we may appreciate, and this topic will be further discussed below.

Perceptuo-motor learning: what do we learn?

What is actually stored in long-term memory during the skill acquisition process? Based on Schmidt’s motor control model (Schmidt and Wrisberg 2008, see Chapter 7), the following elements of information are stored:

In other words, motor learning is the process of working through the motor control process, time and time again, utilising any feedback available to improve performance. Through extensive repetition and trial and error, an increasingly reliable relationship is established between sensory input and motor output that is stored in long term memory – hence the term ‘perceptuo-motor’ learning.

In summary, this section has highlighted the difference between motor performance and learning and recommended that health care practitioners consider a delay between practice and outcome assessment, and utilise a transfer task to examine carry-over of what has been learned. From an information processing perspective, motor learning reflects the extent to which a GMP has been stored in long-term memory.

How do we learn? Stages of learning

Learning a new skill requires practice, and resultant changes in performance emerge over time. An understanding of this process, which can be described in stages, is important to therapists, as patients require different forms of guidance as they progress. The key stages, which were originally proposed by Fitts and Posner (1967), and later updated by Schmidt and Wrisberg (2008), are described below.

Verbal-cognitive stage

This is the first stage of skill acquisition, where the patient needs to get ‘the idea’ of the activity to be learned (e.g. walking with crutches for the first time). This stage is labelled verbal- cognitive as there is often a considerable amount of self-talk going on, with the patient guiding themselves through the solution (e.g. ‘left foot – right crutch…’). Information processing at this stage is predominantly conscious and deliberate; hence the term ‘cognitive’: this stage involves thinking through the activity and working out how to do it. Performance improves quickly but is often variable, as the patient tries different strategies to solve the problem.

What do you think is the main role of the therapist in this stage? Obviously, giving a clear explanation is crucial. Demonstration, either ‘live’ or on video, can be very useful to convey a clear image of the activity, without the need to use complex verbal descriptions. Manual guidance can also be effective in providing the patient with the necessary sensory information. Feedback on performance is also very important for this and the following stages and is discussed in detail later in this chapter.

Is there anything that therapists should try and avoid at this stage? It is easy to overload the patient with detailed information, or to talk, manually guide and demonstrate all at the same time, but this may well lead to information overload – causing the patient to get ‘lost’. Short-term memory has only limited capacity, restricted to around an average of 7 ± 2 chunks of information in healthy people, while it can only retain information for around 20 s. Short-term memory is vulnerable to interference, mainly caused by an overload of information.

Given the limitations of short-term memory, which are often even further compromised in people with a condition affecting the CNS (Chapter 9), it is essential that therapists select key information, e.g. where to look during a balance task (rather than prescribing the optimal position of individual joints). It is important for therapists to consider the patient’s capacity to process information. For example, physiotherapy gyms or occupational therapy kitchens are often lively environments, where music may be used to create an enjoyable atmosphere. While not denying the importance of this, therapists need to consider whether this environment actually enables the patient to focus their attention on the task in hand. As Chapter 9 will explain, information that is not attended to will be poorly remembered.

Motor stage or associative stage

Now that the patient has a clear idea of what to do and some idea of how best to do it, the process of putting this knowledge into practice and fine-tuning performance can begin. At this stage, improvements tend to be smaller, but performance becomes more consistent.

What is the main role of the therapist at this stage? Providing safe opportunities for practice (including trial and error where appropriate), allowing the patient to become proficient at the skill and solve more complex variations of the task. This could be done by adding ‘noise’ to the system, e.g. when practising walking, therapist could start talking at the same time (dual tasking), walking outside where the terrain is more variable (variable practice), interspersing walking with another task (random practice), all of which are examples of contextual interference – a concept we will discuss under random practice below.

With the patient gaining proficiency, the therapist should prepare for handing over control to the patient and gradually withdraw their input – eventually, the patient needs to be become as autonomous as possible.

Autonomous stage

This is where the patient masters the activity automatically and performance becomes more stable; little conscious attention is required. The patient may be able to dual task (e.g. walk with crutches and talk at the same time without this affecting their performance) and learns to adapt their performance to suit different situations. The main role of the therapist at this stage is to provide practice situations in which the patient can consolidate their performance and have the opportunity to problem-solve how the movement should be adapted to meet changing environmental conditions. At this stage, the therapist is usually no longer needed to provide detailed task instructions.

In summary, the process of skill acquisition progresses through stages. Different authors have proposed more or fewer stages (e.g. see Gentile 2000), and may have used slightly different terminology. However, it is important to bear in mind that the boundaries between the stages are rather blurred. Therefore, it is sufficient to have a broad understanding of these main stages of motor learning – and the key role of the therapist in each stage.

Facilitating skill acquisition

Having briefly introduced the role of the therapist in each of the stages of skill acquisition, it is now time to look in more detail at how this process can be facilitated. Several key factors impact on this process:

• Motivation

• Practice: amount, type and scheduling of practice,

• Feedback: type and content of feedback, attentional focus, feedback scheduling.

We will now discuss each of these points in more depth.

Motivation

Motivation is a key factor in any learning process; highly motivated patients are more likely to practise more and may require less therapist input than those who have low motivation. Although an in-depth discussion of the topic of motivation is beyond the scope of this text, it is important for therapists to have a good working understanding of motivation and how this may be enhanced.

Firstly, it may be helpful to differentiate between intrinsic motivation, which is directed at achieving internal satisfaction (e.g. enjoyment, having a sense of autonomy), and extrinsic motivation, which is aimed at external rewards, (e.g. recognition or positive feedback from the therapist). In order to nurture intrinsic motivation, it is important that the patient feels that what they are trying to achieve is meaningful and relevant to their own life and aspirations. E.g. patients may not automatically feel motivated to undertake straight leg raise exercises to improve quadriceps strength. However, the same exercise formulated as a short term goal to achieve the more functional goal of being able to go up and down stairs independently is more likely to be seen as relevant, and encourage a patient to practise.

A technique to enhance motivation and engagement is goal setting, which is widely recognised in rehabilitation settings as a useful tool, even though there is a lack of robust evidence on its effectiveness in clinical populations (Levack et al. 2006) as well as debate on how best to conduct the process (Playford et al. 2009). Despite these uncertainties, goal setting is recommended in a number of UK clinical guidelines (RCP 2008, SIGN 2010) and there is broad consensus that the characteristics of successful goals should follow the SMARTER acronym, i.e. they should be:

Specific: goals need to be specific in order to focus one’s effort. Goals such as ‘I want to be normal again’ lack precision about the aim, and about how progress can be ascertained. A more specific goal could be: ‘I want to be able to ascend and descend one flight of stairs without assistance’.

Measurable: as indicated above, it is necessary to be able to assess to what extent a goal has been achieved. For example, ‘I want to be able to ascend one flight of stairs in 2 minutes without assistance’. Outcome measures comprising timed tests (e.g. the Timed Up and Go test) may be useful to provide detailed information on progress in function and activity.

Agreed: it is important that a goal is agreed between the patient and the therapist, since ownership of the goal is key to intrinsic motivation. Other authors recommend that the ‘A’ stands for ‘achievable, while others argue this should stand for ‘ambitious’.

Relevant: as indicated above, goals need to be seen to be relevant by the patient. However, it is not always possible to formulate goals that meet this criterion. E.g. for a patient with traumatic brain injury, who has to have a plaster applied to his leg to prevent contracture of the ankle plantar flexors, the goal to increase range of movement in the ankle may not seem to be immediately relevant. However, having sufficient ankle movement to achieve 90 degrees of dorsiflexion is an essential requirement for weight bearing, which in turn is necessary to reach the longer term goal of being able to walk. Therapists often need to translate short-term, impairment-based goals into goals that are perceived to be relevant in the eyes of the patient.

Time-based: goals need a timeline, otherwise interventions could be endless. Setting deadlines is one of the most difficult aspects of goal setting in clinical practice however, due to the problems with predicting outcomes. For example, ‘In 1 week from now I want to be able to ascend one flight of stairs in 2 minutes without assistance’.

E: engaging: health care professionals may forget that therapeutic activities – however beneficial for the patient – can sometimes be tedious. Devising activities that are engaging and enjoyable is key to stimulating intrinsic motivation – the stepping stone to successful self-management after discharge. The rapidly growing field of ‘games for health’, using technology such as the Nintendo Wii and Microsoft Kinect, is an interesting development in this respect.

R: reviewed. In order to keep on track, or identify what holds patients back, goals need to be regularly reviewed.

In summary, motivation is key to engaging the patient in their rehabilitation process. Goal setting, based on the SMARTER acronym, is a technique for enhancing motivation. Finding out what motivates patients, integrating this with goal setting, designing targeted interventions and seeing patients achieve their own goals is probably one of the most rewarding aspects of clinical practice.

Practice: amount, types and schedules of practice

How much practice do patients need? Referring back to our case study at the start of this chapter, is there an optimum way to organise therapy sessions when a range of tasks need to be practised, e.g. should it be drill-like, or more variable? Are there any alternatives for practice when a patient is limited in their capacity to practice, e.g. because of pain, fear or fatigue?

Amount of practice

To master a skill, one needs to physically practise it, but how many times? It is clear that there is a lack of clarity on this issue. Ideally, several thousands of repetitions may be required to build and refine the required motor program and for this to be mastered at an automatic level, but this depends on the individual, their baseline performance, and the nature of the skill. For example, a group of people with stroke improved upper limb function significantly by practising an average of 322 repetitions per session (3 sessions a week for 6 weeks). (Birkenmeier et al. 2010). The consensus in the skill acquisition literature is, ideally, that ‘overlearning’ is required, i.e.continuing to practice beyond proficiency. Overlearning involves using the neural network required for the execution of a task over and over again, increasing efficiency and processing speed.

In reality, do patients get an opportunity to overlearn in the clinical setting? A study by de Wit et al. (2005) showed that, in the UK, stroke patients on stroke units (the accepted gold standard for hospital care after stroke) spent on average just 10% of their waking hours in therapy. Lang et al. (2009) showed that the average number of repetitions of task-specific training for people with stroke in a number of rehabilitation units in the USA was only 32 per therapy session. It is essential that therapists make the most of the very limited time available to increase opportunities for patients to practise the skills they need to be ready for discharge. One way to do so is to extend what we mean by ‘practice’.

Mental practice and observational learning

Traditionally, what we think of as practice is ‘doing’, i.e. physically performing an activity, over and over again. This is known as overt practice; the behaviour can be observed. However, ask professional dancers, athletes and musicians how they attain and maintain their level of proficiency and they are likely to reply that they use observation and going through an activity ‘in their mind’ as additional strategies. Below, we will briefly address each of these forms of covert practice and explain how they may be effective.

Observational learning can be defined as:

‘Learning a skill by observing a person performing the skill.’

(Magill 2007, p. 439)

Since the pioneering work by Bandura (1986), we have known that observation can be a powerful tool in learning. Social situations offer opportunities for role-modelling; a principle that is widely used in education in general and the field of skill acquisition in particular. However, it was not until relatively recently that a neural explanation for the phenomenon of observational learning emerged. Pioneers in observational learning, Di Pellegrino and his group, made their discovery in 1992 purely by serendipity. They observed monkeys interacting with objects, using single cell recordings of neurones in the monkeys’ cerebral cortex. To cut a long and fascinating story short, they discovered that the same network of neurones was activated when the monkey was looking at an experimenter interacting with an object, as when the monkey itself was interacting with the object (Rizzolatti and Fadiga 2005). This discovery led to the ‘direct matching hypothesis’, which postulates that by observing an actor interacting with an object, the visual representation of that activity is mapped onto the motor representation of the same activity (quite literally, a case of ‘monkey see, monkey do’). In other words, the visual map of the activity is reflected onto the motor map of the same activity. The neural network involved in this process of ‘reflection’ was aptly named the ‘mirror neurone system’. Much more work has been done since on the roles of the mirror neurone system including on humans, where a similar system has been confirmed. This discovery is so exciting, as it not only confirms earlier findings that observational learning can be effective, but it provides us with a much deeper understanding of how this process is thought to work.

Only gold members can continue reading. Log In or Register to continue