For any carer or parent of a child or adult who has Cornelia de Lange Syndrome there are two broad issues that relate to interventions for selfinjurious behaviour. The first is prevention. Whilst we noted in Chapter 4 that self-injurious behaviour is not necessarily associated with Cornelia de Lange Syndrome any more than we would expect by chance, it is still true that a significant proportion of children and adults with Cornelia de Lange Syndrome show self-injurious behaviour. Therefore, it is important to be aware that selfinjury can develop and that proactive prevention of the development of more self-injurious behaviour is important to bear in mind. We hope that a clear understanding of the causes of self-injurious behaviour, outlined in Chapter 5, and the interventions that are described in this chapter should help parents and carers prevent the development of more severe self-injurious behaviour (see Box).

Four important things to know about preventing self-injury from developing
  1. Pain and discomfort can lead to self-injury and the cause needs to be identified quickly and resolved with medical treatment.
  2. Find stimulating activities for children to have available so that they do not show mild self-injury to gain stimulation.
  3. If you see mild self-injury, such as banging teeth with hands, be aware of how you and others are responding to it. Are you rewarding the behaviour?
  4. A functional communication system is important to prevent socially reinforced self-injury developing. Any child or adult must have an effective and reliable way of letting others know what he or she wants to happen next.

The second issue is interventions for selfinjurious behaviour when it is occurring. This is the focus of this chapter in which we describe how to use the assessment information that has been collected to create an intervention that can succeed. Before describing the different interventions that might be implemented on the basis of assessment information, there are some general issues which are important to note. 

The first issue is the importance of caring for injuries that result from self-injurious behaviour. It is, of course, important to ensure that any injuries are dealt with appropriately to prevent infection and also to ensure that unseen injury has not taken place. It is particularly important to try to get wounds to heal rapidly to prevent infection but also because the longer a wound is open the more people find it itches and consequently they may respond to this by picking or scratching the wounds and a viscious scratch-itch cycle develops. Covering the wounds and seeking medical advice on promoting healing at this stage is important. The second issue is that it is very important when putting together an intervention that others are involved in this process and all information on the intervention is shared. This means ensuring that all people who come into contact with the person showing self-injury are aware of the intervention strategy that is being used and are constantly updated on any changes that may take place.  

It is important to seek professional help and advice at various stages. Some interventions are more easy to implement than others but it is always important to try and work in collaboration with a professional who has experience of implementing interventions. For psychological intervention this kind of professional help should be available from educational psychologists, clinical psychologists and nurses and teachers who may have been trained in behavioural methods (also known as behaviour modification or applied behaviour analysis). There is no doubt that psychological interventions are difficult to implement and a team approach is more likely to lead to success. 

It is also important to note that intervention for self-injurious behaviour is more of a process than something that happens once and then does not need to be repeated. We have outlined the process at various points in this book and would emphasise the importance of knowing background information about Cornelia de Lange Syndrome as well as information about self-injurious behaviour in all people who have intellectual disabilities. It is also critical that anyone thinking about an intervention for self-injurious behaviour is thoroughly familiar with the potential causes and has conducted a systematic assessment to try to determine which cause or causes are most applicable to the person they intend to work with. 

Once this has been done a plan can be formed and the intervention can be implemented. At this point you are only at the first stage of intervention and it is critical that two things now take place. First, whatever intervention you conduct, you must evaluate the effect of the intervention. This means keeping a regular and systematic record of the self-injury in order to determine whether or not you are being effective. The second is that you persist with an intervention for a reasonable period of time in order to give the intervention time to work. You should not expect rapid and immediate results as it may take time for people to learn different ways of behaving. These two issues are constantly referred to as we consider the different types of interventions that might be implemented. On the basis of the constant evaluation that is conducted it is extremely important that you are prepared to modify the intervention and try different things. This will be based primarily on your understanding of why the self-injurious behaviour is occurring but will also be based on the results of the evaluation. This overall process of intervention can be repeated many times and it is important that any effective intervention is constantly reviewed and that those involved in the intervention are prepared to change. 

In the sections on intervention methods you will see that we mainly target self-injurious behaviour and the way in which this behaviour can be decreased. However, it is important to note that increasing adaptive behaviours (the term ‘adaptive behaviour’ usually refers to independent living skills and other behaviours such as communication) alongside decreasing selfinjurious behaviour is an important aspect of intervention and does seem to lead to longer term change. It is often tempting to focus only on the self-injurious behaviour and not on the adaptive behaviours. We would emphasise that increasing adaptive behaviours is every bit as important as decreasing self-injurious behaviour and we would strongly advise that as much attention is paid to this aspect of the intervention. In the population of people with intellectual disabilities and, we believe, within Cornelia de Lange Syndrome it is the case that more severe self-injurious behaviour is associated with a greater degree of intellectual disability. Individuals who show self-injurious behaviour are more likely to have poor expressive communication and limited adaptive behaviours. This is important for us to know as it will determine the types of intervention that might be implemented. Because of this association between the greater severity of intellectual disability and self-injurious behaviour, in the section on psychological interventions we tend to focus more on interventions that are appropriate people with severe intellectual disability. However, it is important toremember that if someone has speech then other types of intervention are available and we include a brief description of these methods.

Keeping records to evaluate interventions

There is little doubt that the most important  aspect of intervening is keeping a record of  whether or not the intervention is working. In  this section we will describe how these records  may be kept and how we can evaluate whether  or not our intervention is working.

What to record

It may seem obvious to say that it is the self-injurious  behaviour that needs to be recorded but it  is important that we define exactly what we are  going to record in order that everyone involved  records the same thing. There are a number of  ways in which we might record self-injurious  behaviour. First, we need to define what we  mean for any given individual by self-injurious behaviour. It is better to use specific terms and to  write out a definition of what we mean so that  we can agree when the behaviour is occurring  and when it is not. We call these operational definitions.  So, rather than use the term self-injurious  behaviour when we are evaluating an intervention  we would keep records of, for example,  skin-picking or head-banging so that we are clear  about the target behaviour. Each of these behaviours  might be individually defined so that everyone  records the same thing. Box below gives some  examples of operational definitions for self-injurious  behaviours.

Operational definitions of self-injurious behaviour

The table below provides some examples of operational definitions of different types of self-injurious behaviour that we have used in our research. As you can see they are very specific but this helps make sure we all mean the same thing. It is important to define exactly what we are going to record to ensure that everyone involved agrees on which behaviour has been seen.

Self-injuriousbehaviour Operational definition
Body to object bangingMovement of the hand or body down onto an object (excluding body throwing and slapping surface)
Body hittingMovement of the hand or object down onto the trunk of the body
Body pickingUse of the finger nail(s) to scratch or pick at the body.
Body pokingPressing the tip of a single finger or thumb into the body.
Eye pokingPressing the tip of a single finger or thumb into the eye.
Face hittingRapid movement of the hand making contact with the face
Face pickingUse of the finger nail(s) to scratch or to pick at the face including cheek and jaw line.
Hand bitingEnclosing and clamping teeth down onto fingers or hand.
Head bangMovement of head towards and making contact with a surface (e.g. tables, walls floors)

A second way of recording is that rather than  focusing on the behaviour itself we can record or  measure the effect of the behaviour. So, we  could record the number of scratch marks on  someone’s arm, the size of an abrasion or  whether a new scratch mark has appeared in a  given period of time, such as a day. A third way  of recording is to make a rating of all the selfinjurious  behaviour across a given time period.  So, we may rate on a scale of 1 (not at all) to 5  (nearly all the time) how frequent the self-injurious  behaviour has been in, say, an hour or half  a day or a day. (see Box below for examples of how  these records can be kept).
Keeping records of self-injurious behaviour

The table below gives an example of how to record the frequency of self-injurious behaviour. The person completing the record indicates the number of times they have seen the behaviour in each hour. A completed record gives us an idea of how frequently self-injury is occurring.


The average is calculated by dividing the total number of self-injurious responses occurring in a day by the number of hours that the person was observed for (e.g. day 2 is 15/3 = 5; day 3 is 12/3 = 4). Recording behaviour in this way before and after interventions indicate how successful the intervention has been in reducing the frequency of self-injury.

These three methods  of keeping a record of self-injurious behaviour  give us different types of information in order to evaluate an intervention. The methods have different  strengths and limitations. Clearly counting  the self-injurious behaviour is difficult because  we need to carry out almost constant observations  (we discuss ways of doing this later in this  chapter). However, this is the most accurate way  of evaluating an intervention. The other two  methods are easier to do but they tend to be less  reliable when it comes to evaluation. It is important  that you use the most reliable method that  you can and that you check that everyone is  recording the same thing.


How to record

Just as there are different things that we might record so there are a number of ways in which we can carry out the record keeping for self-injurious behaviour. We can simply count how many times the behaviour occurs within a given time period, for example, an hour, a day, or a week. This is called a frequency count and the record tends to be accurate. However, the method involves constant observation and this may not always be practical and we discuss methods of what is known as time sampling below. Whist frequency counts are the most common way of recording they may not give the most accurate picture of behaviours that have a very low frequency but tend to occur for long periods of time. An example this is eye pressing which may only occur once or twice in an hour but when it does occur it could occur for 20 minute periods. Generally speaking it is better to record how long these sorts of behaviours occur for, as we really want to decrease the amount of time spent in the behaviour. Frequency counts are more useful for behaviours that have a short and stable duration. 

It is of course often impractical to observe someone all the time and consequently we can use methods of sampling in order to keep an accurate record. There are two ways in which we can sample. One is that instead of watching someone for a whole day we could watch them for, say, three half-hour periods throughout a day in order to record the level of self-injurious behaviour. A second way in which we can sample is that when we are observing we can use some time sampling methods in order to get accurate records. There are two methods of time sampling that we would recommend: momentary time sampling and partial interval time sampling and these are described in box below.

Momentary time sampling and partial interval time sampling

Momentary time sampling
Momentary time sampling involves recording occurrences of behaviour and possible antecedents at intervals throughout a day or across several days. The normal procedure is for the observer to have a walkman or watch that beeps every, say, 5th minute to remind them to observe. The observer then looks at the target individual for a few seconds and records a) whether or not self-injury (or other problem behaviour) is occurring and b) what else is happening i.e. attention, demand, playing alone etc. The observer then looks away and does not observe the individual until the next bleep. The bleeps can be spaced as far apart as is practical, although the ideal interval size will depend on the frequency of the behaviour being observed. This is a useful technique as it ensures that observations are random and not just taking place because the target behaviour is occurring. It is also easy to carry out and does not require lengthy observations of several hours at a time.
Partial interval sampling
Partial interval sampling works in a similar way to momentary sampling but in this case, observations are continuous across the whole period. The session is split into sections (ranging from 10 seconds to several minutes, depending on the behaviour) and as before, a walkman or watch is used to signal moving from one section to the next. Target behaviours and environmental events are ticked off if they occur at any point during a particular section. This process is more time consuming than momentary time sampling but can be more accurate if the time periods for momentary time sampling are too long. 

In both of these methods we would use a personal cassette player with a pre-recorded prompt on the tape to make a record of the self-injurious behaviour on paper. This will give us a reasonably accurate record of the self-injury without having to observe the person all the time.

Using the data that you collect

The best way to evaluate the effects of an intervention is to keep a record of the self-injurious behaviour before the intervention starts (this is called the baseline) and then to keep the same record whilst the intervention is running. The most useful thing to do is to plot the information on a graph and then you have a good visual picture of whether the behaviour is increasing, staying at the same level or decreasing. It is also useful to use the graph to keep a record of any unforeseen incidents that you think might have affected the self-injurious behaviour so that you can see whether or not a pattern emerges. The Box below describes how to plot a graph based on the information you have collected.

Plotting graphs

The Box about Operational definitions of self-injurious behaviour gave an example of how to record the frequency of self-injurious behaviour. Sometimes it is much easier to see patterns in the data when you plot it in a graph. In particular this is useful for seeing how successful an intervention has been at reducing the severity of self-injury in a particular individual. The graph below is an example of how you might plot this data. It shows the frequency of self-injury before and after the implementation of an intervention.


The graph enables us to see that self-injury is markedly decreased after the intervention apart from on day 10 when it increases again. As is indicated on the graph the individual had a bad ear infection on this day which might explain this transient increase in behaviour.


Interventions for self-injury related to medical causes

In Chapter 2 we described the medical conditions that are more commonly associated with Cornelia de Lange Syndrome and that may give rise to pain and discomfort. It should be noted that whilst these medical conditions are associated with Cornelia de Lange Syndrome it is, of course, likely that children and adults with the syndrome will, at some time have medical conditions that affect anybody. In Chapter 5 we discussed how we think self-injurious behaviour is related to pain and discomfort and in Chapter 6 we discussed how a potential relationship between a medical condition and self-injurious behaviour can be assessed. If it is thought that pain and discomfort might be related to self-injurious behaviour from the results of an assessment then the most effective way of trying to decrease the self-injury is to deal with the medical cause. This means involving medical staff who are prepared to systematically evaluate potential sources of pain and discomfort and try methods to relieve the resultant pain and discomfort if the root cause cannot be cured. Whilst this strategy seems obvious we believe it is very important that it is vigorously pursued and that all possible medical causes are ruled out before implementing psychological interventions. The Cornelia de Lange Syndrome website is a good resource for learning more about the treatment protocols for medical conditions that are associated with Cornelia de Lange Syndrome.

We also noted in Chapter 2 the possibility that individuals with Cornelia de Lange Syndrome might have a peripheral sensory neuropathy. This may give rise to unusual sensations and feelings in the feet, legs, hands, fingers and arms and this may lead to biting or scratching of the area where this sensation is experienced. From the research literature that we have examined there is very little indication that there is an effective intervention for these sensations. However, there is a case report in a medical journal that describes hand and arm massage as effective in reducing self-injurious behaviour in a woman who has Cornelia de Lange Syndrome1 (see Box below).

Massage for very severe selfinjurious behaviour in a girl with Cornelia de Lange Syndrome
This case report was presented by Dossetor, Couryer & Nicol in 19911. A young girl (LH) with Cornelia de Lange syndrome was described who, at age 14, had been displaying severe self-injury for ten years. Medication had been tried but had been unsuccessful in treating the behaviour. Only splints and a helmet could prevent LH from engaging in self-injurious behaviour. Functional analysis had failed to reveal any consequences of the behaviour.

As a part of a new treatment programme LH received a 30 minute massage twice a day. From the first day, she enjoyed and relaxed with the massage. It had a beneficial effect on her behaviour and mood for the rest of the day. After the third day LH indicated that she wanted to massage others as well. Thus a form of reciprocal play developed. After six months, her self-injurious behaviour improved so much that she was taken off all medication for the first time in ten years. She wore no splints or helmet and her injuries healed. The authors indicated that the improvements had been maintained 18 months later. Any relapses that had occurred had been mild.

It is possible that this massage worked by in some way alleviating or cancelling out these sensations that were experienced by the individual. This seems a possible intervention and at present the only way to know whether it may work is to systematically evaluate the effect of this intervention in the way we have described above.

Interventions for self-injury related to sensory reinforcement

In Chapter 5 we noted that there is good evidence that self-injurious behaviour can occur simply because the sensory or perceptual effects of self-injury are in some way rewarding. The type of reward can be of two types. First, there can be alleviation of pain and discomfort (this process would be referred to as sensory negative reinforcement) and interventions related to medical causes can be explained by this process. Second there can be reward that is positive from the stimulation that comes from the self-injury. This seems most likely for self-injurious behaviour that involves mild eye pressing and perhaps finger noted that there is no evidence that the type of self-injurious behaviour is necessarily related to any particular cause, and any form of self-injury can be reinforced by sensory consequences.

For self-injury that appears to be related to the positive sensory consequences there are three basic strategies. These involve increasing alternative forms of sensory input, decreasing the sensory input that arises from the behaviour and increasing the available rewards for not showing the behaviour. Before considering any of these interventions it is important to think through a number of issues. First, it is has been argued that some mildly self-injurious behaviours are developmentally appropriate2. This means that all children at some time will suck and bite their hands and if you watch any child then at some point you may see them scratch or, for example, bang their head. For some of these behaviours, particularly mouthing hands, this is a way in which the very young child will explore their environment. However, the problem is that because development may be much slower and may plateau at an earlier stage for children with Cornelia de Lange Syndrome, then these behaviours may stay in the child’s repertoire for a much longer period.

The second issue is deciding whether or not to intervene. For mild self-injurious behaviours that are not causing considerable tissue damage it is important to think through what the person will do if they are not showing these behaviours. In other words is there a good alternative to the behaviour that you may want to decrease. For behaviours that are injurious there is clearly a good case to be made for trying to decrease the behaviour. However, for behaviours that are extremely mild and no alternative can be developed then there is perhaps a less strong case. The other factors to take into account when considering whether to intervene is the extent to which the behaviour is additionally handicapping for an individual. Some behaviours can make children look very unusual and for this reason it may be important to try to decrease the behaviour

Increasing a specific type of reinforcement

When self-injurious behaviour is shown because the sensory reinforcement that is experienced is positively rewarding, then increasing the type of sensory reinforcement can lead to decreases in self-injurious behaviour. In practice this means trying to understand the kind of sensation or perception that the child gets from the self-injurious behaviour and then trying to present this kind of stimulation in a different way. (See Box below)

Providing stimulation in a more adaptive way
Judith Favell and colleagues described how Dane, a 14 year old boy with profound intellectual disability with light\dark perception only, showed eye poking for approximately 40% of the time3. When Dane was provided with toys with visually striking properties e.g. brightly coloured toys, mirrors and translucent coloured shapes the eye poking decreased to about 10% of the time. One interpretation of this decrease is that as visual stimulation comes from the toys there is no longer a need for eye poking.

This is, of course, not always practical or easy to do, for example for head-banging. However, for hand biting or hand mouthing then providing different activities that stimulate the child’s hands and mouth can be effective. The real task is finding an activity that the child likes as much, and preferably more, than the effect of the self-injurious behaviour whilst ensuring that this does not look as unusual as the behaviour and that the child does not tire of the activity quickly. This is where a process of trial and error and an active imagination is important.

Increasing the general level of sensory reinforcement

The second strategy is to try to increase the amount and variety of sensory reinforcement that is available for any child or adult as much as is possible and practical. There are a number of ways in which this can be achieved. One way is by ensuring that there is a high turnover of sensory activities. This means that the child’s activities are changed at very regular intervals such that the child does not become bored with the activity that is available to them and reverts to self-injurious behaviour to seek stimulation. The second strategy is to ensure that there are a variety of sensory activities available to the child. This means that rather than providing activities or toys that are all stimulating to the child in the same way, that different toys and activities giving very different effects are made available. The third strategy is to try to increase the overall level of activity that the child is experiencing. This means ensuring that the level of stimulation is generally kept high so that the child does not need to revert to self-injurious behaviour to provide stimulation. Inevitably this means a good deal of trial-and-error to find activities that child finds reinforcing. Low input, high return toys are particularly good for this purpose and local toy libraries for children with intellectual disabilities can be a very good resource to find activities and try them out (see Box 7.8).

Toys that can provide stimulation to compete with behaviour
Glyn Murphy and her colleagues described how David, a 14 year old boy with profound disabilities, partial sight and hearing problems, showed high levels of a self-stimulatory behaviour (rocking) even when he had toys to play with4. However, when one of the toys was adapted to vibrate when it was touched, the self-stimulatory behaviour went down from 85% of the time to 15% of the time and active toy touching increased from 6% of the time to 94% of the time. Although the rocking is not self-injurious, this example does show how providing stimulating toys can effectively compete with the stimulation that comes from a repetitive behaviour. A similar finding was reported by Jon Bailey and Lee Meyerson in 1970 for the effect of vibration on headbanging.

Sensory extinction

Sensory extinction was first described in the 1970’s and referred to a process whereby the sensory consequences of a behaviour are eradicated such that the reward no longer occurs and consequently the behaviour will decrease. This has been demonstrated for both self-stimulatory behaviours, such as plate spinning and handwaving, and some self-injurious behaviours (see Box below).

Changing the consequences for self-stimulatory behaviour
Arnold Rincover first reported the method of sensory extinction in 1978. He described Reggie, 14, who had profound intellectual disability and visual impairments who would spin objects, particularly plates error on hard surfaces in a repetitive way5. When the sound from the plate spinning was muffled, by padding the table top, the amount of time Reggie engaged in plate spinning dropped from about 60% of the time to nearly zero. Rincover argued that this was because the behaviour was rewarded by the noise of the plate spinning and when this was muffled the behaviour was extinguished.

However, it should be noted that the evidence for self-injurious behaviour is rather limited. Basically, the procedure consists of removing any sensory consequences to the selfinjurious act. In practice this may mean padding or covering the area that is targeted for self-injury such that the individual does not experience any feedback whatsoever. There are two major problems with this technique. The first is that the person may well try to remove the pad or protection in order to seek the stimulation and the second is that the padding often needs to be thick and obtrusive and this may make the child look unusual. We would certainly recommend that if this type of intervention is tried then it is important to combine the intervention with the two strategies that have been described above (increasing the specific type of sensory reinforcement and increasing the general level of sensory reinforcement) and/or the reinforcement competition strategy that we describe next.

The other way in which the sensory consequences of self-injury can be minimised is by a carer preventing the responses when they occur (so called responses prevention) There are a number of problems with this method. First, for high rate behaviours this can be very demanding of carers. Second, there is a risk that by providing a social response to the behaviour it may become socially reinforced, although this can be minimised by the social response being very cool (no speech, no eye contact). Third, although the child may learn not to show a behaviour he or she will not necessarily learn what to do instead. For these reasons this type of intevention needs to be implemented with caution and certainly alongside another strategy.

Reinforcement competition

This strategy can be used in addition to the interventions described above and is considered in much more detail below. Basically, the intervention consists of presenting a reward for the individual when they are either not showing the selfinjurious behaviour or they have not shown the behaviour for a given period of time. The different ways in which this can be achieved are described in Section A comment on punishment in which the importance of identifying the right reward and frequent changes of rewards is also described. For a summary of the main points about selfinjury that is rewarded by sensory stimulation see Box below.

Four important things to know about intervening for self-injury maintained by sensory stimulation
  1. Providing an alternative form of stimulation can decrease self-injury, the closer the type of stimulation is to the effect of self-injury the better.
  2. Alternative forms of stimulation, such as toys, should be low input with high return, changed regularly and with new forms introduced often.
  3. Present rewards when the person touches toys or other stimulating objects. These can be from you but might be in the object itself (objects that vibrate or light up when touched).
  4. Present varied and stimulating activities throughout a day.

Interventions for self-injury related to social reinforcement

Interventions for self-injury related to social reinforcement

 In Chapter 5 we focused on the fact that selfinjury can occur because of the social reinforcement that is presented by others. There is a good deal of evidence in the research literature that this is a common cause of self-injurious behaviour in people with intellectual disabilities and from our research we have no evidence to suggest that the cause is not appropriate for people with Cornelia de Lange Syndrome (see Boxes 6.2 and 6.12). In the broader research literature on social reinforcement in people with intellectual disabilities the evidence is that for approximately 70 percent of people the cause of self-injurious behaviour is likely to be social reinforcement (see Box below).

How common is social reinforcement for self-injurious behaviour
Brian Iwata and colleagues carried out a review of all the studies of self-injury that had used analogue methods as an assessment for whether the self-injury was socially reinforced (6). 152 children and adults had taken part in the assessments and the results showed that 23% showed self-injury that was rewarded by attention, 35% by escape from tasks and just under 10% for other social reasons. 20% were thought to show self-injury because of the sensory stimulation but the evidence for this figure is less strong.

Additionally, there is good evidence that when the cause is social reinforcement psychological interventions can be effective, although they can be rather difficult to sustain over long periods of time. In this section we will describe interventions for self-injury that is related to social reinforcement and, just as for the section on assessment, we will differentiate between self-injury that occurs because of a positive social reinforcement process (i.e. self-injury that tends to be rewarded by social attention or by tangible items such as drinks or activities) and self-injury that occurs because of negative reinforcement (i.e. self-injury that tends to be rewarded by escape from task demands or social contact).

Self-injury maintained by positive social reinforcement

Self-injury maintained by positive social reinforcement

There are two general strategies that can be implemented when the results of an assessment show that self-injurious behaviour is maintained by the social attention that follows the self-injury or the presentation of more tangible things such as food, drinks or activities. The first is to try to generally increase access to these reinforcers. That is, to increase the overall level of attention from others and increase the availability of the more tangible items that have been reinforcing the self-injurious behaviour. There are a number of specific ways in which this can be done and these are described below. The second general strategy attends to the issue that attention is extremely rewarding to almost any child or adult and that it cannot always be presented. It is important to ask the question why attention is so rewarding and reinforcing and consequently understand why the person wants attention at such regular intervals.

Apart from the intrinsically rewarding nature of social attention, it is likely that other people often act as the link between the individual who shows self-injurious behaviour and things that they wish to access. That is, people with severe intellectual disabilities will inevitably be highly reliant on other people to satisfy their needs. The general strategy therefore is one of trying to increase levels of adaptive behaviour and remove the barriers that might exist in the environment that limit an individual’s access to the things that they want or need and thus require other people to present. Whilst this may not seem to be a priority when self-injurious behaviour is occurring, it is an important background strategy that will help any psychological intervention to be effective. We fully acknowledge that increasing adaptive behaviour is difficult when working with people with severe or profound intellectual disability. However, there is good evidence that people can acquire new skills when the precision teaching methods of applied behaviour analysis are employed with consistency over time. We should not make the assumption that any child, no matter how disabled, cannot learn new skills. This is particularly important with regard to communication and we discuss this in more detail below.

Changing the consequences
If the results of an assessment show that the selfinjurious behaviour is caused and maintained by the attention that follows the self-injury then it is important to change this consequence (see Box below).
The importance of what happens after self-injury
It is very important to find out what happens after self-injury and to think about whether it might be a reward. The most common rewards are social attention and escape from tasks and these can take many forms. Telling someone off, even by shouting, can be rewarding, as can giving eye contact or trying to distract someone from the behaviour by presenting alternative activities. It is critical that you check whether things that happen after self-injury might be rewarding. If there is a chance that it is, then the response must be changed.

There is very strong evidence in the research literature that doing so can decrease self-injurious behaviour 7. However, this strategy should never be used alone and there are a number of issues that need to be carefully thought through before trying to change the consequences. When self-injurious behaviour is caused and maintained by the attention that occurs after the self-injury the temptation is to stop presenting this attention because this should lead to decreases in the self-injurious behaviour. Whilst it is certainly true that the self-injurious behaviour will eventually decrease there are two basic problems with this strategy.

The first is known as an extinction burst8. That is when a behaviour has been reinforced in the past and the reinforcement is withdrawn the behaviour does not stop immediately. Rather it increases in both frequency and intensity before finally stopping. The Box below shows an example of an extinction burst.

Extinction burst: Things get worse before they get better.


The graph above shows what happened over a 45 minute period when we did not present the normally occurring social reward for a problem behaviour that was being shown by a 14 year old girl (in this example this was aggression) 9. Within about 10 minutes the behaviour increased in frequency by about 50% and the aggression was also more intense (harder hits and hair pulls). It took about 45 minutes before the extinction burst was over.

The problem of an extinction burst is evident in our daily lives. If you think about the last time you lost something you will almost certainly have repeatedly returned to the place where the thing was even though you now know it is not there. It is very similar for a child or adult who has been used to receiving attention after showing self-injurious behaviour. Even though in the short term the reward does not come they will continue to show the behaviour, often more vigorously, for some time before stopping. The problem here of course is that a good deal of injury can be experienced by the child during this extinction burst. This procedure of extinction therefore needs careful consideration before being implemented. It is possible to use protective devices when trying this strategy in order to protect the person from any injury. We would strongly recommend that professional help is sought before considering that this kind of intervention.

The second problem with simply stopping a reward, such as attention, that usually follows the self-injurious behaviour is that the child or adult no longer has any way of gaining the reward. So, whilst self-injury has stopped, the need that the child or adult has for attention has not been removed and then two things may happen. The first is that a different behaviour, most likely a problem behaviour if the person has a limited repertoire of adaptive behaviours, may replace the self-injurious behaviour and the second is that the self-injury may simply come back at a later date. Therefore, when thinking about the use of extinction it is critical that the procedure is combined with a way in which alternative behaviours that can replace the self-injury are taught.

Increasing alternative behaviours
In the section on changing the consequences we clearly identified a problem with the strategy of simply stopping the reinforcement that usually follows self-injurious behaviour. If we do this we leave the person without a way of gaining the thing that they want or need. In the research literature there is very strong evidence that teaching a form of communication that can gain attention under these circumstances can be very effective and it is critical that if the function of self-injurious behaviour is to gain attention then the child has a reliable and effective way of communicating that they want attention. We cannot emphasise enough the importance of trying to increase the child’s ability to effectively influence the behaviour of others under these circumstances.

Below we consider the various forms of communication that can be taught to a child in order to help to decrease a behaviour that occurs because it is positively reinforced. Whilst this approach focuses on the child learning to communicate their need for attention to others, it is important to remember that self-injurious behaviour can be positively rewarded by things other than attention. So, it is important that the child is able to communicate a need for more tangible items, such as drinks, food or activities, or that they are able to access them in some other way. There is some research evidence that children with severe intellectual disabilities can learn to use micro-switches that can control aspects of their immediate environment such that they are able to satisfy their needs without an adult being present (see Box 7.15). Whilst this research is at an early stage there is no reason why children and adults with Cornelia de Lange Syndrome should not also be able to learn in precisely the same way. The more control the child or adult has over their environment and the better their ability to control and acquire the things that they want and need the less they will need to rely on self-injurious behaviour.

For a summary of the main points about selfinjury that is rewarded by social attention see Box below.

Five important things to know about intervening for self-injury maintained by attention.
  1. Modify your response to the self-injury. If you must respond, to protect and restrict, be ‘cool’, no speech, no eye contact, no fun!
  2. Present positive and fun attention when self-injury is not occurring. Set a timer to remind you.
  3. Find or teach a communicative response that can get attention from you and others. Make sure everyone responds to it.
  4. Increase independence to give control to the person and reduce their reliance on your attention.
  5. Make sure everyone is doing the same thing.

Self-injury maintained by negative reinforcement

Self-injury maintained by negative reinforcement

We described in Chapter 5 how self-injurious behaviour may occur when an aversive event or task is presented to the individual and that the self-injury is then reinforced or rewarded by the removal of the aversive event or task. Under these circumstances there are a number of strategies that can be used to try to manage the behaviour and decrease the self-injury in the long term.

Reducing the aversive nature of tasks and demands First it is important to think through why any task may be experienced as aversive or unpleasant. There can be a number of reasons for this and trying to see the task from the child or adult’s point of view is an important way of trying to understand why escape from the task is a big reward. Often the tasks that might be presented to children and adults with a severe intellectual disability are not necessarily intrinsically rewarding or the rewards are so long term that the activity is not experienced as worthwhile. It is always worth considering any task or event and asking yourself what is in it for the child?

Sometimes an event is experienced as aversive simply because it evokes fear or anxiety. We may not know how this came about in the first place but often it is because a particular event was associated with a very unpleasant experience at some time in the past and consequently the event now evokes fear. When the child is exposed to this aversive event they may experience the fear or anxiety that occurred in the past and show self-injurious behaviour until they are removed from the situation or the situation is removed from them.

Under these circumstances we can draw on the methods that are normally used to treat phobias in order to try to reduce the degree to which the child experiences the situation as fear or anxiety provoking. One strategy that may be used is called graded exposure. This means closely examining a feared event or situation and breaking it down into small components. Then the person can be gradually exposed to small parts of the event or situation and rewarded for not showing the behaviour when they are in this situation. An example of this kind of intervention is shown in Box below.

A graded approach to a feared stimuli
A colleague of ours described an intervention for a young boy who showed self-injury mainly when he was being taken toward the toilet. In the past the toilet had been associated with painful attempts to pass faeces, due to constipation, and although this problem was now resolved his avoidance of the toilet persisted and he escaped by self-injuring. The intervention consisted of first rewarding the boy for walking a short distance toward the toilet (but going no further, turning around and going away) and then for getting a bit closer, then touching the door, then opening the door and so on. It took time but eventually the boy was again able to use the toilet and did not selfinjure when being taken.

A second strategy that can be used for fear and anxiety provoking events is called “flooding”. This refers to exposing the person to the situation and not allowing them to remove themselves from the situation until the fear and anxiety have decreased. This is a very effective intervention for phobias. However, it is likely that if this intervention is tried then the self-injurious behaviour might increase dramatically, because of the extinction burst that was described above and in Box 7.13, in the short term and consequently this is a difficult strategy to employ.

The research literature shows that for negatively reinforced behaviour the type of event that will lead to self-injurious behaviour in children and adults with intellectual disabilities is an unpleasant or unwanted task or demand11. Under these circumstances we may well expect to see more self-injurious behaviour in teaching settings and in one-to-one and school environments. Again, it is important to think through why the child might find the tasks aversive or unpleasant. If we can do this then we may be able to make the task a more positive experience and consequently the child will not show self-injurious behaviour to escape. There are a number of ways in which we can look at tasks and think about why they are aversive.

First, tasks may be experienced as too difficult and need breaking down into small steps in order that there are more opportunities to reinforce the child. Part of this process is called a task analysis and is a common method that is used to help in teaching adaptive behaviours for children and adults with severe intellectual disabilities. An example of a task analysis is given Box below.

Breaking a task down into small steps
Here is an example of how hand washing can be broken down into small steps so that, initially, each one can be rewarded. Later rewards could be given every two steps, then three and so on. This is called forward chaining. A different method is backward chaining. Here everything would be done for the person except the last step. The person is prompted to do the last step and then rewarded. Next time the person is prompted to do the last two steps and then rewarded, then the last three and so on. The advantage of this method is that the reward naturally comes at the end of the task.
  1. Turn on tap. 
  2. Wet hands. 
  3. Turn off tap. 
  4. Pick up soap. 
  5. Rub soap on hands. 
  6. Put soap back. 
  7. Turn on tap. 
  8. Rinse hands. 
  9. Turn off tap. 
  10. Pick up towel. 
  11. Dry hands. 
  12. Return towel (or, if you are male, drop on the floor for someone else to pick up).

Once the task analysis has been completed then it is possible to reward the child for completing each component of a task rather than only at the end of the long task. Over time it is possible to put components of the task together and consequently only present the reward, for example, when two or three components have been completed. This process of reward fading helps to move towards a more natural process of teaching where the reward comes when a task is complete.

It is also important to consider other aspects of the task that may make the event aversive and consequently lead to a burst of self-injurious behaviour. There is some evidence that a very high rate of demands as opposed to few demands that are spaced by a good period of time is experienced as more aversive and thus more likely to lead to self-injurious behaviour11. Similarly, the kinds of prompts that are given to a child or adult to complete a task may also be experienced as more or less aversive. Physical prompts that are high rate and firm may well be experienced as much more unpleasant than more gentle prompts. A trial-and-error process of the components of any teaching session should be able to reveal those parts of task demands that are experienced as the most aversive and can be modified in some way.

There are some general aspects of teaching sessions or task demands that can make the experience more pleasant and thus less likely to evoke self-injurious behaviour when it is reinforced by escape from a teaching situation. First, the amount of reward available for participating in small aspects of the teaching session needs to be high and reliably presented. The second is that when resistance to teaching sessions is prominent it is possible to gradually increase the level of demand or the time for which the session is conducted gradually when combined with high levels of reward and reinforcement. (See Box below)

A little bit longer each time
When children or adults will not take part in any teaching activity, even for a short time, then one strategy to use is shaping in which the person is rewarded for spending increasing longer amounts of time taking part in a task. So, first of all you might only reward someone for sitting down at the table for a second or two, then when they have sat for, say, 20 seconds, then a minute, then two and so on. You can also use the same method for increasing the amount of demands. You could present a reward for just touching the task for a second, then for picking up part of the task, then completion of a small bit of the task and so on. You may need to use some prompts initially but these can be faded away to verbal instructions.

Gradual increase and patience is the key. Finally, there is a growing body of research that shows that locating more difficult tasks and events within a stream of easier tasks leads to lower rates of problem behaviour and this includes self-injurious behaviour12.

So, there are a number of ways in which it is possible to decrease the aversiveness of a task or event such that when self-injurious behaviour is reinforced by escape, the behaviour is made less likely to occur. All of these strategies need considering in the broader context of the importance of increasing adaptive behaviour in all people with learning disabilities and, for children, the school curriculum and setting. Additionally, it is important to consider whether some tasks or aversive events are really necessary and whether there are alternative solutions to the person acquiring a particular skill when they find the teaching of that skill or adaptive behaviour highly unpleasant.

In addition to these specific strategies that might be adopted to make tasks less aversive there are two general things that might be considered and for which there is research evidence that they are helpful. The first is the importance of task variety. Everyone has the experience that repetitive tasks with very little variability are more boring and consequently more aversive than a series of different tasks. Second, there is growing evidence that allowing people choice over the type of tasks that they will undertake does lead to less problem behaviour than usually occurs in response to demands. Whilst these general strategies seem obvious, it is well worth keeping an eye on whether there are varied tasks being presented and whether the individual has choice over what kind of task they are carrying out.

Changing the consequences
As we noted in the section on changing the consequences for self-injurious behaviour that is rewarded by attention or tangible events, there can be problems with simply not presenting a reward. The situation is no different for self-injurious behaviour that occurs when there are aversive tasks or events and the reward is escape from these. There is certainly evidence that when self-injurious behaviour occurs during an aversive task or event, and thus the function is escape from the event, that not allowing the person to escape i.e. continuing to present the task or event will, eventually lead to a decrease in a self-injurious behaviour. This procedure is called escape extinction and the same problems that we have previously described above will occur. That is the behaviour will increase dramatically in both frequency and intensity prior to the eventual decrease of self-injury. Once again if this kind of intervention is considered then it is important to think through the safety issues and the likelihood that it is possible to persist with the task demands with significant injury. It is important to seek advice when considering this kind of intervention and it is critical that the intervention is considered in conjunction with those described above in which the nature of the task demand is modified and also with an increase in alternative behaviours being programmed.

For a summary of the main points about selfinjury that occurs when tasks are presented see Box below.

Five important things to know about decreasing self-injury that occurs when tasks are presented
  1. Try not to remove the task when self-injury occurs. If you must, to protect and restrict, come back to the task for a brief time and stop when there has been no self-injury.
  2. Break the task into a series of smaller tasks, give big rewards for completion or tries, give time between prompts and check the prompts are not unpleasant.
  3. Find or teach a communicative response that can tell you the person wants the task to stop and respond to this.
  4. If the communicative response for stop happens too frequently error set a timer with the person and only respond after the timer gone off. Gradually increase the time.
  5. Make sure everyone is doing the same thing.

Increasing alternative behaviours
We noted above that when behaviour is reinforced by others then simply removing that the behaviour from a child or adult’s repertoire will mean that they will then be unable to satisfy their personal needs. The situation is no different for behaviour that is reinforced by escape from aversive events or tasks. Under these circumstances it is equally important that the ways in which a child can express the need for a task to stop or be removed is attended to. Inevitably, this means trying to increase the ability of the child to communicate that his or her experience in that a task as aversive. In Section 7.6 we consider ways in which an appropriate form of communication can be increased
Interactions with setting events
We have previously discussed the way in which what we referred to as “setting events” might interact with demands. Put more simply this means that when we make demands on anybody, the demands will be experienced as more aversive, and thus more likely to lead to self-injurious behaviour, if there is a setting event present such as fatigue, low mood, pain or discomfort. (See Box 7.20).
An interaction between a task demand and a setting event
Mark O’Reilly and colleagues describe how a child with Williams syndrome would show problem behaviour when there was background noise whilst tasks were being presented to him but there was no problem behaviour when there was no background noise with tasks or background noise but no tasks13. When the child was given ear plugs the problem behaviour decreased. This is interesting because it shows how a feature of a syndrome (hyperacusis, or sensitivity to noise, is a feature of Williams syndrome) interacts with an environmental event, tasks, to increase problem behaviour.

The identification of these setting events is helpful for intervention planning and trying to identify behavioral predictors of setting events such as low mood, pain and discomfort and fatigue are important. The reason for this is that when these setting events are evident then clearly it is not a good time to introduce aversive tasks. If these aversive tasks are introduced at this time they will be much more likely to evoke self-injurious behaviour if it is reinforced by escape from these tasks. It is important therefore to have programmes of activities that are flexible and will allow for the opportunity for the child to participate in less aversive tasks if the setting events are evident.

In the longer term it is important to try to modify and eliminate these setting events such that they are not present and thus cannot interact with aversive task demands. Attending to sleep difficulties that may give rise to fatigue, pain and discomfort are good examples of how general interventions may help with self-injurious behaviour. There are a number of methods for dealing with sleep difficulties and information on these can be found in some of the books in the Bibliography. Finally, we would note that the periods after a meal times are perhaps a special case for some children and adults who have Cornelia de Lange Syndrome and are experiencing gastrointestinal reflux. It is certainly true that people who are experiencing reflux do experience pain and discomfort in the oesophagus, upper chest and throat. At these times people may find the presentation of any tasks or other aversive events as even more unpleasant than usual and consequently this might lead to selfinjurious behaviour.

Functional communication training

Since the mid-1980s there have been repeated demonstrations in the research literature that improving the functional communication of children and adults with learning disabilities can lead to a decrease in different forms of challenging behaviour14. However, the research literature is also a very clear about two things. First, functional communication training is only really effective when the assessment of self-injurious behaviour shows that the self-injury is reinforced either by social positive reinforcement of attention or tangible items or when it is reinforced by social negative reinforcement of escape from task demands or other types of social interaction. Second, that it is extremely important that the communication that is taught is matched to the reason that the self-injurious behaviour occurs. In effect this means that when the assessment shows that self-injurious behaviour is maintained by attention or access to more tangible items then the communication must be able to do exactly the same thing. Similarly if the selfinjurious behaviour is maintained by escape from task demands or other aversive events then again, the functional communication must have the same effect. (See Box below).

Saying the right thing at the right time
The first description of Functional Communication Training by Ted Carr and Mark Durand in 1985 was something of a turning point in interventions for problem behaviour because it made us think of these behaviours as being like communication 15. More than this though, they showed that it was not just important to teach communication but it was also important to teach the right type of communication. In their study of five children with disabilities they first found out whether the children showed problem behaviour because it led to attention or escape from task demands. They then taught each child attention or escape gaining responses. When the child was taught the right response (for example, an attention gaining response for an attention maintained problem behaviour) problem behaviour decreased. However, when they taught the wrong response (for example, escape gaining response for an attention maintained problem behaviour) the problem behaviour stayed at the same, high level. So assessment of the reason for self-injury is important as it will tell us the most important kind of response to teach.

These two principles have been repeatedly demonstrated and consequently we would again emphasise the importance of the assessment process that we outlined in Chapter 6.

Precursor behaviours and Functional Communication Training

Before describing the various forms of functional communication that can be taught and some of the principles that should be considered when teaching Functional Communication Training there is some recent research evidence that we think might prove to be important. It is important in Functional Communication Training to pick the right time to prompt the person to show the functionally communicative behaviour.

Obviously, we want to do this at precisely the time that they would usually self-injure in order to gain the reinforcement thus displacing the selfinjury. However, also we do not want to present the reinforcement for the functional communication when the self-injury has occurred as this may reinforce this behaviour. Recently it has been demonstrated that self-injurious behaviour does not always occur in isolation16. Rather, there are some behaviours that will happen just before self-injury, especially when it is reinforced by social processes. This means that just prior to self-injury occurring there are what are called “precursor behaviours” that may often be observed. These behaviours may take the form of attempts to attract someone’s attention, for example, vocalizations or increased signs of agitation or mild forms of self-injury prior to the behaviour being shown in an extreme form. It is important to try to identify these behaviours. If they do exist they will give an important clue as to when to try to prompt someone to show a functionally communicative behaviour. Some of the behaviours that we have seen in children with Cornelia de Lange Syndrome are listed in Box 6.15 and when these behaviours are occurring this may be a very good time to prompt functionally communicative responses.

The effectiveness of forms of Functional Communication Training

In Section 5.3.1 we described how self-injurious behaviour can be reinforced by attention or by escape from task demands. In section 5.4 we described how this means that self-injurious behaviour may be considered to be similar to a form of communication in that it is able to effect the behaviour of others and thus satisfy the needs of the individual. The rationale to Functional Communication Training as an intervention for self-injurious behaviour is that the communicative response that is taught will replace the self-injurious behaviour and by influencing the behaviour of others satisfy the needs of the individual. 

Consequently, the intervention is based on teaching a form of communication that is able to affect the behaviour of other people. This means that what is taught may be different from that which is traditionally taught as part of a speech and language therapy programme. There may be many similarities but it is important to note that the most important aspect of Functional Communication Training is the capacity of the behaviour to affect the behaviour of others.

Some forms of augmentative communication

In order to identify the most appropriate form of functional communication to teach, it is important to assess the child’s existing method of communication. If a signing system is already in place then the task may be more one of increasing the effectiveness of responses to the communications of the child as opposed to replacing the signing system with something else. The intervention in this case consists more of ensuring that the communication is effective, i.e. is responded to by others as opposed to increasing the repertoire. Speech and Language Therapists and Clinical Psychologists can give advice on the most appropriate form of Functional Communication Training for any child or adult. Here we will describe the main types of augmentative communication that can fulfil this purpose.

There are a number of signing systems that can be taught to children and adults who have intellectual disabilities and the most popular of these is Makaton. The signing system predominantly involves the hands and consequently for some children with Cornelia de Lange Syndrome this may be problematic.
Picture Exchange Communication System (PECS). 
PECS is rapidly gaining popularity as a method of communication for children and adults with severe intellectual disability and\or autism17. As the name suggests the method basically rests on the child or adult either pointing to or giving pictures in exchange for a desired item or activity. There are some advantages to this method over manual signing systems. First, the child or adult can point to a picture and this may be easier for children who have upper limb abnormalities than signing. Second, one important aspect of communication is that the communication is effective i.e. that the things which somebody desires are available when they show the communicative response and thus the communicative response is reinforced. With signing it is entirely possible that the person may sign for something that is not available or that cannot immediately be made available for practical reasons. Whilst a response of “wait” or “later” may be made at this time, this may not be understood by the child and consequently the strength of that communicative response is weakened. With PECS it is possible to only make available pictures of items or activities that can be immediately presented for any given period of time. Although this limits what is available to a child or adult for a given period of time it does mean that the pictures available will show things or activities that are possible and can be presented when requested and consequently the communicative response is further strengthened. As with Makaton there is a structured procedure for teaching PECS and recent evidence shows that it is effective for children with autism. It also shows an association with decreases in problem behaviours
(see Box below).
The effect of PECS on challenging behaviour
In 2002 Marjorie Charlop-Christie and her colleagues taught three boys with autism to use the Picture Exchange Communication System18. The results were good and the boys learned to use a number of pictures to make their needs known. Interestingly there was also an increase in the boys’ social communicative behaviours and a decrease in problem behaviours such as tantrums, grabbing, disruption and being out of seat. Across the three boys there was a 70% or greater reduction in 10 out of 12 problem behaviours and four behaviours fell to zero levels.
Electronic devices. 
There has recently been an increase in the availability of electronic devices or “Voice Boxes” that can be activated by children and adults with intellectual disabilities. These devices can be programmed to say anything that would be useful to the person. The child can then be reinforced for activating the Voice Box and thus the Box has the properties of a functional communication system (see Box below).
What happens if you cannot speak?
Building on his previous work Mark Durand conducted further research on Functional Communication Training (FCT) with five children with disabilities and published his results in 199919. This time instead of teaching the children signs or the right phrase to say he taught the to use an Introtalker that can be programmed to ‘say’ phrases using digitised speech at the press of a microswitch. Once again FCT was successful and the challenging behaviour of all five children decreased. More importantly, when the children used the Introtalker with adults who did not know anything about the programme or the children’s challenging behaviour the behaviour stayed at low levels as the adults naturally responded to the requests made by the child via the Introtalker.

One advantage of the system over signing is that everybody can understand what is said by the Voice Box whilst not everybody will understand all Makaton signs. PECS also has this advantage over Makaton as the word is written beneath the picture.

While signing, PECS and electronic devices are the most common forms of functional communication that can be taught to children and adults with intellectual disabilities they are not the only methods. The most important thing is to find a method that will suit the child, has a structured procedure for being taught and will be effective in all environments. We would strongly emphasise this last point. Any communication system is useless if it is not responded to. The effectiveness of Functional Communication Training in decreasing self-injurious behaviour is heavily dependent on how easily the child can reliably effect the behaviour of others in precisely the way that they want to. This key to effective intervention hinges on the idea of “Response Efficiency”.

The importance of response efficiency in Functional Communication Training

The technical term “response efficiency” simply refers to how effective any particular behaviour is in gaining reinforcement or reward from the environment given the cost of showing that behaviour20. It is perhaps self-evident that if we have two ways of influencing the behaviour of other people then we will usually choose the way that gives us a very high return i.e. always affects the behaviour of others, and is very low cost i.e. is easy to do with no negative consequences. So, a functional communication system will generally be effective in reducing self-injury that is maintained by attention or escape from demands if the system has a higher response efficiency than self-injury. This means that when we think about teaching a functional communication system and trying to ensure that it works in the natural environment that there are number of aspects of efficiency that we must consider. This is because when the person has a choice between self-injury and functional communication they will choose the most efficient response
(see Box below).

Responding to disruption and speech that occur for the same reason
We looked at the problem and communicative behaviours of a seven year old boy with Down’s syndrome to try and understand why he might show problem behaviour when he could already communicate with speech, albeit limited 21. We conducted some analogue experiments and watched him in his class. The results were clear. He was frequently disruptive and aggressive when asked to do tasks but would also say ‘no!’ and wave ‘bye-‘bye’. When we looked at how adults responded to these different behaviours they were much more likely to stop asking him to do something when he was aggressive and disruptive than when he said ‘no!’. So his aggression and disruption were much more effective than his communication and consequently he showed these behaviours more often. The lesson here seems to be that even when someone can communicate, if it is not effective then they will show a more effective behaviour, in this case aggression and disruption.

As this is the case we must ensure that the functional communication that the child has is much more efficient than the self-injury. Recent research has shown us that there are a number of aspects of response efficiency that are important to consider.

Low cost. 
Any form of functional communication must be easy to do. If the effort required to show the response is low, then the effectiveness of the system is reduced. It is important to ensure that the response that is taught is as effortless to show as is possible
High return. 
It is also important that the return for any form of communication is very high and by this we mean a much higher than the return that would occur for self-injurious behaviour. This means a number of aspects of return must be attended to. First, the amount of reinforcement for any given communicative response must be high. This means that the response that is given to, for example, a picture request for a period of attention, should be a longer period of attention than would be given in response to self-injury. Second, the response to a communicative act should be fast. This means that when a communicative response is made the reinforcer should be presented more quickly than it is presented following the self-injurious behaviour. Third, the reliability of reinforcement should be very high for the communicative response. What this means is that every communicative response should be responded to and by everybody. Whilst we recognise that it is a tall order to ask that responses to communicative acts are fast, long and reliable by rewarding the communication in this way, the communicative act becomes much more effective than self-injurious behaviour and consequently comes to replace the self-injury.

The high return for communicative acts clearly present some practical difficulties in the longer  term. However, there are ways in which these  may be overcome. First, there is some research  evidence that it is possible to teach tolerance for  reinforcer delay. What this means is that once the  communicative act has been established in a  child’s repertoire then it is possible to signal in  some way that there will be a brief period before  the presentation of that reinforcer. This time period  can be gradually increased from a matter of  seconds to a longer period, thus allowing a more  natural situation to develop. However, it is important  to note that when we introduce this procedure  we are necessarily decreasing one aspect of  high return, that of speed. A second way in which  we can overcome the practical difficulties is with  the PECS system. The picture component of the  system means that we can give the child or adult  pictures of the reinforcing activities that are available  at that time and the pictures will include only  those things or activities which we know we can  present quickly, for reasonably long periods and  reliably. For those activities that cannot be presented  in this way then the pictures can be withheld until it is practical to make them available.

For a summary of the important points about Functional Communication Training see Box Below

Four important things to know about Functional Communication Training (FCT) as an intervention for self-injury

  1. For FCT to succeed you need to teach the right response:
    • Attention getting communication for attention rewarded self-injury;
    • ‘Stop the task’ communication for selfinjury rewarded by escape from tasks.
  2. Communication in FCT can take a number of forms: speech, signs, picture cards, pressing microswitches to activate tapes. The most important thing is that everyone can understand without being told.
  3. Whatever the communicative response is,it should be less effort than self-injury, get more reward than self-injury, be rewarded more frequently than self-injury and get rewards faster than self-injury.
  4. Be careful about communication that cannot be rewarded e.g a sign for ‘park’ when it cannot be delivered. Picture systems may be better as only pictures of things that can be done could be made available.

Facilitated communication

Whilst discussing the importance of functional communication we want to be clear about the difference between the approaches we have described above and Facilitated Communication. Facilitated Communication is a method in which a child’s hand or adult’s hand is guided by a “facilitator” to tap out messages on a keyboard. This method has been widely discredited and should not be confused with Functional Communication Training.
(See Box below).

Facilitated communication
A brief and useful resource for evaluating the effectiveness of Facilitated Communication is a Fact Sheet that can be obtained from the National Autistic Society. In the Fact Sheet there is comment on a review of all the available evidence undertaken by Prof. Pat Howlin.

She reviewed 45 carefully conducted studies of Facilitated Communication and found that of the 350 people who had been involved in trials of Facilitated Communication, only 6% showed any evidence of independent communication and for 90% there was evidence that the responses were unwittingly guided by the facilitators. In the United States five professional bodies have adopted a formal position opposing the acceptance of Facilitated Communication.

Additional strategies

In addition to changing the responses to selfinjurious behaviour and increasing adaptive behaviours, such as communication, there are some general strategies that may be helpful in reducing the number of incidents of self-injurious behaviour and for which there is some research evidence. Parents and clinicians have been aware for a long while that for some individuals the lack of a structured environment can promote brief periods of anxiety in a child or adult and can lead to incidents of self-injurious and other challenging behaviours. For children and adults who find the lack of structure anxiety provoking this can be most evident at times of transition from one environment or activity to another 22. This is especially the case when the child does not know which environment or activity they are going to or if that environment or activity is not the one that they expect to go to.

One way of helping with this difficulty is to try and structure the environment and daily activities using timetables such that predictability is high. This is, of course, more easily said than done. It is not always possible to run a household, a classroom or a day centre in exactly the same way each day. However, it may be possible to increase the predictability even though there is some variability. This can be done by having daily photo timetables or objects of reference available at the beginning of a day presented in the order in which the events or activities may take place. This allows the child to know what is happening throughout the day and as each event or activity becomes imminent the photograph or object of reference may be shown to the child, the child can be taken to the activity, and then the object of reference or photograph is removed. This way the child or adult may move through the day with an element of predictability. It is, of course, important to choose photographs and objects of reference carefully and to ensure that they are always associated with the activity and additionally that they are never associated with any other activity. Consistency is important to help establish the child’s understanding of the relationship between a photograph or object of reference and what is about to happen.

Psychological interventions when the function of self-injurious behaviour cannot be identified

Up to this point we have indicated the kinds of interventions that the research literature indicates are effective when we are able to understand why the self-injurious behaviour is occurring on the basis of assessment and the model that we build. It is, of course, sometimes the case that we cannot work out why self-injurious behaviour is occurring and the results of the assessment may not indicate a clear cause and thus we find it difficult to select our intervention. Under these circumstances there are still interventions that we can try and again there is evidence that they can decrease self-injurious behaviour. We will deal with these interventions in the following sections.

A comment on punishment

The early methods of behaviour modification tended to focus on how self-injurious behaviour could be decreased as quickly as possible. Inevitably, this meant that there was a good deal of research conducted into the use of punishment as a treatment for self-injurious and other behaviour problems. The term punishment used in this context refers to the decrease of behaviour when an aversive stimulus is presented following an instance of the behaviour. Following heated debates in the late 1980s and early 1990s the description of punishment techniques in the research literature has decreased and has been replaced by the approach which we have adopted in this book.
  That is, an approach which tries to uncover the reasons for self-injurious behaviour, replace the behaviour with a more adaptive response and manage the behaviour in a way that would lead to its eventual decrease. 

Whilst this is the approach that is now advocated by many researchers and clinicians in the field, there is still the possibility that punishment can be proposed as a method of control of the self-injurious behaviour of people with intellectual disabilities. In this section we are want to draw attention to the three main issues that are important to consider.

  1. First, punishment tends to give a short term success. There is some evidence in the research literature that the behaviours decrease more quickly when punishment methods are used than when alternative approaches are adopted, although there is some debate about whether this is the case. One factor that may be related to the short term success is that the punitive stimulus delivered, needs to be quite severe or unpleasant in order to suppress the behaviour. Inevitably this may lead to the use of punishers in a way that may be considered inhumane. For example, squirting lemon juice into the mouth, enforcing physical activity and inducing pain. Whilst there is evidence that these procedures may lead to short-term success there is also evidence that when children and adults who have been involved in these treatments are followed up at a later date the self-injury has returned and persists 23.
  1. Second, whilst punishment may decrease selfinjurious behaviour, it does so simply by teaching somebody what not to do. However, if we accept the model that is described in Chapter 5 that the behaviour is functional then if we use punishment alone, clearly we do not teach the person what to do instead of self-injuring. This means that the person will have the same needs but no way of satisfying those needs. If the model is correct then this means the self-injury will occur again at a later stage and the evidence on punishment shows that this is exactly what happens.
  1. The model that we have presented in Chapter 5 shows how self-injurious behaviour can be learned. Punishment as an effective treatment also depends on learning i.e. the person learns not to self-injure because the punishment will follow. If this is the case then there is agreement that the self-injurious behaviour is learned. The question then is how best to affect the unlearning of self-injurious behaviour. Our argument is that if there is agreement that self-injurious behaviour is a learned behaviour then either approach may work but we would argue that using positive methods is preferable to punishment simply because of the short term success issue, the inhumane nature of many punishment methods and the importance of replacing the behaviour with a different behaviour.

This discussion of punishment presents a brief summary of our views and we accept that there are different opinions on the use of punishment as an intervention for self-injurious behaviour. Clearly, it is a personal decision as to whether punishment techniques are used and not all punishment techniques will necessarily be inhumane. If you are considering using punishment techniques to decrease self-injurious behaviour we would strongly advise you to seek advice from a clinical psychologist or behaviour analyst and to keep all of the effects of punishment under close review.

Differential reinforcement

One strategy for decreasing self-injurious behaviour for which there is good evidence is differential reinforcement. This simply means presenting some kind of reward when an undesirable behaviour is not occurring and thus trying to make the consequences of not self-injuring “better” than the consequences of self-injuring. There are a number of ways in which this can carried out and these are described below. However, before describing the various methods of differential reinforcement it is important to know some of the basic principles that underlie how this is thought to work.

For sometime it was believed that differential reinforcement worked because it was rewarding the person for not doing something as opposed to doing something. However, it now seems likely that it may work in a different way and is effective because the reward that usually follows self-injurious behaviour is now presented at a different time and consequently the person showing self-injurious behaviour does not need to show self-injury behaviour in order to gain the reward. If you look back at the model described in section 5.3 and think about a child showing self-injurious behaviour in order to gain attention then that if we present attention at regular intervals then it makes it more likely that the child will not need to selfinjure in order to gain the attention, as the attention is occurring with sufficient regularity.

One of the problems with some forms of differential reinforcement is that, a bit like punishment, it does not necessarily teach the person what to do. So, if we present attention as a reward for not self-injuring then the person may not have shown any particular behaviour that would normally gain attention. All they are doing is not self-injuring. This means that the person has not learned a behaviour that they can use in the future to gain attention. For this reason Functional Communication Training, discussed in section 7.6, tends to be favoured over differential reinforcement techniques. However, when we are unable to identify the reinforcement for selfinjurious behaviour, and thus the function of the behaviour, differential reinforcement can be a useful strategy to try to decrease self-injury.

Important aspects of differential reinforcement
There are four aspects of differential reinforcement that are important to consider before starting any kind of programme.
Reinforcer selection. 
If you are going to present a reward for no self-injury occurring for a given period of time then clearly it is important that the reward you choose is indeed a reward for the person. Different people have very different preferences in terms of what they find rewarding. Before starting the programme it is important to draw up a list of things that the person finds rewarding and to ensure that more than one type of reward is available.
Avoiding satiation.
One of the problems with presenting a reward after a period of no selfinjury when the period is very brief, is that people may very quickly satiate to the reward. What this means is that when you have had too much of a good thing then you do not want any more! Three ways of dealing with this problem are to use very brief amounts of rewards, to use a variety of different rewards and to use symbolic rewards, such as stars on a chart that can be exchanged for an item of the person’s choice.
When using any reward programme it is extremely important that there is a high degree of consistency. This means that when the reward system is set up the reward is always delivered after a period of no self-injury, for example, and this is carried out by everybody in all environments.
Changing the programme over time. 
In the initial stages of any differential reinforcement programme the time period for which a behaviour should not occur needs to be fairly short so that the person has a very good chance of success. One way in which the time period can be calculated is by working out the average period of time that elapses between incidents of self-injurious behaviour and then dividing this by three. So, if self-injurious behaviour is seen on average every 15 minutes, then the initial differential reinforcement period would be 5 minutes and the reinforcement would be delivered if there is no selfinjury for a 5 minute period. Also, in the early stage of the programme the amount of reinforcement that is delivered should be high for a short period e.g. 2 minutes of attention for 10 minutes of no self-injury. Over time, as the programme begins to be effective it is important to increase gradually the amount of time that the person does not show self-injury for (10 minutes, then 15, then 20 etc.) and to gradually decrease the amount of reinforcement that is given (2 minutes of attention then 1, then 30 seconds etc.) in order to avoid satiation which we discussed above.

Types of differential reinforcement
There are three basic types of differential reinforcement.
  1. Differential reinforcement of other behaviour (DRO). DRO means presenting reward when a behaviour that has not occurred for a predetermined period of time. Thus, it does not matter what happens during the period of time, the person can be doing anything except self-injury, what matters is that they did not self-injure. If the person does self-injure then the time period starts again. 
  2. Differential reinforcement of incompatible behaviour (DRI). DRI is similar to DRO but in this case the reward is presented when the person has shown a behaviour that is incompatible with self-injurious behaviour for a given period of time. So, if it is decided that a DRI programme will be implemented for head hitting then a child may be reinforced for playing with toys with his hands for a predetermined period of time because the behaviour of playing with toys with hands will be incompatible with head hitting. If the person does self-injure then the time period starts again. There is some limited evidence that DRI is more effective than DRO (see Box below).

    The importance of rewarding incompatible behaviours
    An old study but an important point. In 1974 James Young and John Wincze described their intervention for head banging and hitting for a 21 year old woman with profound intellectual disability 24. They showed that if they presented a reward if the woman kept her hands on an object then head hitting decreased but head banging increased. This shows us two important things. The first is that rewarding an incompatible behaviour can work, hands on an object is incompatible with head hitting. The second is that if someone has not been taught a behaviour to replace the self-injury, such as communication, then another behaviour may increase.
  3. Differential reinforcement of alternative behaviour. (DRA).
    We have already discussed one form of DRA and that is Functional Communication Training. Differential reinforcement of alternative behaviour means presenting reinforcement for a behaviour that has the same function as self-injurious behaviour. So, if a child shows self-injurious behaviour in response to a task demand then the DRA programme would consist of terminating the task demand when the child shows a behaviour such as making the sign for “break” as opposed to showing self-injurious behaviour. The sign for break would be the alternative communicative behaviour.

Ways in which differential reinforcement can be delivered
There are three basic ways in which the differential reinforcement can be delivered.
  1. Whole Interval Method. This method would be used when running a DRO or DRI programme. The method entails watching the child for the all of the period that you have set in the programme. So if the programme consists of DRO after five minutes, then you would watch the child for five minutes and if the behaviour has not occurred then the reinforcer is presented. If the self-injury does occur then the time starts again. The obvious problem with this method is that it means constant observation of the child in order to ensure that the behaviour has not occurred for five minutes. Inevitably, this has led to problems with implementing this approach but thankfully there are two other ways in which the differential reinforcement programme can be implemented. 1.

    Momentary DRO. This method is much easier to implement than the whole interval method and consists of setting a period and then only observing the child at the end of that period and presenting the reward if the child is not self-injuring when that observation takes place. So, if the programme consists of a Momentary DRO (or DRI) with a ten-minute interval, then running the programme comprises having a watch with a countdown function that will bleep after 10 minutes. When the watch bleeps, the person running the programme will briefly look at the child and if the child is not self-injuring then the reward is presented. If the child is self-injuring then the reward will not be presented and when the selfinjury has stopped the time will be reset and the time period starts again (see Box below).

    Comparing methods of differential reinforcement
    Alan Repp and his colleagues compared two different methods of differential reinforcement to try to decrease the problem behaviour of three 7 year old boys with intellectual disabilities25. They found that the whole interval method (in which a reward was gained for not showing problem behaviour for 5 minutes) was more effective than the momentary method (in which a reward was gained if for a brief period at the end of a 5 minute interval) the child was not showing problem behaviour. Importantly they also found that if the momentary method was used after the whole interval method than it was just as effective.

  2. Noncontingent reinforcement (NCR). This is perhaps the easiest way of presenting reinforcement but it is debatable as to whether this is a method of differential reinforcement or whether it works in a different way. To take the example of a child showing self-injurious behaviour who finds attention reinforcing, noncontingent reinforcement means setting a time period and then presenting a period of attention after that time period has elapsed regardless of whether or not the self-injurious behaviour has occurred. This sounds very curious as we have previously argued that it is important not to present reinforcement when selfinjurious behaviour occurs. However, there is some research evidence that using this method is as effective as differential reinforcement methods for some children (see Box below).

    Noncontingent reinforcement versus DRO
    For many years there was a strongly held view that one of the best ‘positive’ interventions for self-injury was DRO (presenting a reward or reinforcement if self-injury has not occurred for a period of time). In 1993 Timothy Vollmer and his colleagues published a report showing that noncontingent reinforcement was equally effective26. This means that if you know the reinforcer for a behaviour (say attention) then instead of running a DRO programme (where you have to watch for, say, 10 minutes and if the behaviour has not occurred then present the reward of attention) you could try noncontingent reinforcement and present the reward of attention every tenth minute regardless of what is happening. This finding really needs to be replicated but for the time being it shows us that there is a more practical alternative to DRO.

    Noncontingent reinforcement may work by simply ensuring that the child receives attention at regular intervals thus meaning that the child does not have to self-injure in order to gain the reward. If this is the case then it is important that the interval is very brief and that the programme is rigorously maintained. The advantage of this method over the whole interval method is that for the whole interval method it is preferable if there is no response to the self-injurious behaviour. However, this may lead to extinction occurring with the inherent problem of increased injury. Using this method may offset the problem of extinction whilst being able to run a much more manageable programme than that described in whole interval methods.

The above descriptions may appear somewhat bewildering given the variations that could be used. Our advice is as follows. The best alternative would be DRA in the form of Functional Communication Training. However, we recognise that it may not always be possible to identify the function of behaviour consequently it is not easy to put this kind of programme into practice. We cannot teach a behaviour as an alternative way of satisfying a need if we do not know what the need is. So, the next best method would be the whole interval method followed by Momentary DRI with the incompatible behaviour being an adaptive behaviour that is likely to be intrinsically reinforcing for the child. However, in order to offset the problem of extinction it is necessary to ensure that there is a response to the self-injurious behaviour when it occurs that ensures the safety of the child. This will commonly mean some kind of physical restraint by others until the self-injury abates. During this period it is absolutely critical that the response of the person involved in the restraint and the restraint itself is not experienced as reinforcing or rewarding by the child. This means that the response of the adult must be cool and should not include eye-contact or talking to the child as these may be rewarding. How rewarding this is will differ depending on the child and it is important that the response of the child to this approach is monitored and records are kept to evaluate the effectiveness of the approach. Finally, we would strongly advise seeking advice from a clinical psychologist or behaviour analyst before implementing any restraint procedure

The second piece of advice that we would give is that differential reinforcement procedures and NCR should be seen as helping to gain some decrease in the self-injurious behaviour and giving a window of opportunity for increasing the child’s communicative behaviours in the ways that we have described in section 7.6. So, it may not be a procedure that would have to be used for a very long period of time. There is no doubt that the procedure is hard work and requires a good deal of input. This may be made easier by gradually lengthening the period of time between rewards whilst keeping a close eye on the effectiveness of the programme.


Within the last decade there has been a small number of research papers that have described a beneficial effect of physical exercise on self-injurious and stereotyped behaviours. Whilst it is not clear why exercise should decrease self-injurious behaviour the effect on some people is notable. We do not underestimate the difficulty in trying to find physical exercise that people may find enjoyable and would be willing to participate in and, of course, some people will be limited by physical disability.

Nevertheless, for the sake of completeness we should note that there is evidence that this intervention may be effective.

Generalisation and maintenance

There are two issues associated with behavioural interventions for self-injurious behaviour that have caused both clinicians and researchers significant challenges over the past 30 years. First, whilst it is possible to decrease self-injurious behaviour in one environment using a fairly structured programme the gains do not necessarily naturally occur in all environments. In other words the decrease in self-injurious behaviour can often be specific to the place in which the programme was first conducted and the people who were there. The second issue is maintenance. This means trying to ensure that any early gains in decreasing self-injurious behaviour are maintained over time. Whilst reading this chapter you have no doubt been struck by the amount of planning and effort that goes in to running a successful behavioural programme. It is perhaps inevitable that over time it becomes more and more difficult to keep these intensive programmes running. We will deal with the issues of generalisation and maintenance in turn.


One way in which generalisation can be assisted is by carefully programming the gradual implementation of a programme into all environments. So, if a behavioural programme is shown to be successful in one environment then it can be taken into new environments in a systematic way. In practice, this means sharing information with others and may also mean convincing others that the intervention is worth pursuing. This can be helped if there are records kept of the effectiveness of an intervention so that others can see that an approach is successful. Another way in which this kind of generalisation can be promoted is by the person who has been closely involved with the programme in the early stages going into the second environment and working alongside those who will be taking the programme over. This pairing of the person who initially runs the programme and those who will be running the programme in the new environment could also take place where the programme has already been running.

The second way of trying to promote generalisation, and also may be helpful in reducing the onerus nature of a programme, is to use a novel and clear signal that identifies the programme as running and thus indicates to the child or adult that a new learning process is in operation. In practice what this means is that when a programme is first implemented the clear and novel signal is placed in an obvious position and the child is clearly aware that the signal is present. The signal can be anything that is new, and thus will only be associated with the programme, portable, and thus can move with the child wherever the programme and the child need to go, and highly visible to the child. The signal can take any form such as a brightly coloured large card with a unique design on it or an item of clothing that has never been worn before and so is always visible to the child.

The signal should then be shown whenever the programme is running but should never be shown when the programme is not running. This way the person learns to discriminate between those times when the new learning is taking place and those times when it is not. This may be helpful in two ways. First, there are always times when it is simply not possible to run the programme and this gives us a bit of breathing space without losing the effectiveness of the programme.

The second is that when we come to generalise the programme to a new person or a new environment we can take the signal with the child and this will help the child to know that the new learning is now operating with the new person or environment.


Maintenance of programmes over time is particularly problematic. One of the things that we have become aware of is that when the self-injurious behaviour has decreased somewhat and things are not so problematic there is a tendency for the behavioural programmes to be used with less consistency because there is less urgency. The problem then is that the behaviour may return as the new learning becomes undone. This means that ensuring the consistency of programmes over time is extremely important. There are two issues that might help with maintenance. The first is that whatever the programme implemented, it needs to be as minimally demanding of others as is possible whilst still being effective. This means, for example, trying to increase the period of differential reinforcement programmes as rapidly as possible while still maintaining any gains. Secondly, we believe it is extremely important, and there is research evidence to support this, that an adaptive behaviour that can replace the self-injurious behaviour is taught to the child or adult at an early stage in any programme. There is good evidence that Functional Communication Training can maintain overtime because of its inherent capacity to give the child control over their social environment. In other words there is natural maintenance in programmes that manage to build up children’s adaptive behavioural repertoires.

Loss of control

For some children and adults increases in selfinjurious behaviour in both frequency and intensity may be unrelated to environmental events or medical causes such that it appears that they have started to lose control over the behaviour. This idea is somewhat contentious but we feel that there are some aspects of self-injurious behaviour that should lead us to think that the person may not have complete control over the behaviour. The first is the presence of selfrestraint or preference for imposed restraint. We have discussed these in Chapter 5. The second is the association between self-injurious behaviour and compulsive behaviours and, to a lesser extent, hyperactivity and stereotyped behaviour. These associations make us think that the person has difficulty in inhibiting their own behaviour. A lack of behavioural inhibition, this means not being able to prevent a behaviour from starting or not being able to stop a behaviour once it has started, is a feature of hyperactivity and compulsive disorder. As we discussed in Chapter 6 it may well be that self-restraint and preferred imposed restraint are indicators that are the individual is unable to inhibit their behaviour and consequently needs to do so physically.

Using protective devices

When behaviours occur at a higher level of frequency and intensity there is a temptation to use protective devices such as arm splints, gloves, and helmets in order to limit the amount of selfinjury and the consequent damage. There are a number of issues associated with the use of protective devices that it is important to consider. First, using the devices such as padded gloves might help with ignoring the behaviour as part of an extinction programme (see above) because the chance of injury is reduced. However, we have noted the importance of not conducting an extinction programme in isolation because the person might easily develop another form of the  behaviour and then there has been no gain. So, before using protective devices it is extremely important to assess the function of self-injurious behaviour as we have described in Chapter 6.

The second issue is that the use of any protective device can give rise to physical problems such as irritation of the skin and, with straightarm splints, atrophy of the muscles. It is important therefore to ensure that medical advice is sought both prior to and during the use of protective devices. The third issue is that if the person is unable to inhibit their behaviour then whilst protective devices may help them to do so in the short term they may come to rely heavily on the presence of the protective devices in order to control the self-injurious behaviour. Thus, a preference for this imposed restraint may develop and it is important to plan how the restraint may be faded over time in order to avoid a high level of constant restraint.

This kind of intervention i.e. the gradual fading of preferred imposed restraints, can be very successful. There is, curiously, an advantage in an individual having a high preference for imposed restraint. First, it usually means that the behaviour is under control. Second, it can mean that the restrictive nature of the device can be reduced over time without losing the control over the selfinjurious behaviour. We have managed to achieve this in the past in people who have a preference for wearing arm splints to control their headpunching. We were able to gradually introduce more movement into the elbow joint of straightarm splints and also gradually reduce the length of the straight-arm splints, down to a cuff around the wrists, without losing the control of the protective device. (See Box 7.30).

Replacing self-restraint with protective devices that can be faded
We carried out an intervention for Dawn, a 25 year old woman who had profound intellectual disability and vision and hearing impairments27. She punched her head and body and self-restrained by wrapping her hands tightly in clothes and pushed her hands into wooden objects. As can be seen in the graph although she self-restrained she still punched her head and body about 5% of the time. We then introduced arm splints which had an adjustable joint on the elbow. At first we set the degree of movement to about 50% of the full range. During this period Dawn did not self-injure at all as the head and body punching were restricted by the splint. After a time we increased the range of movement to 75% and then 100% (full, normal range of movement) without self-injury occurring. We were able to put the splint under Dawn’s clothing so that it did not appear odd and whilst wearing the splint she was less restricted than when she self-restrained.

This kind of intervention is difficult to implement and we would strongly advise seeking the advice of a clinical psychologist or behaviour analyst before trying to implement the intervention.

The use of protective devices is clearly contentious as it may be seen as a punitive method of intervening with self-injurious behaviour. However, protective devices are often used simply out of the desperation of a parent or carer to protect the person they care for from injury. We would strongly advise that before using protective devices carers should seek the advice of a clinical psychologist or behaviour analyst and the advice of an occupational therapist or physiotherapists to ensure that the device can be faded over time. Our experience is that it may take some time to find the right device for an individual but if the device does gain control over very severe and intense self-injurious behaviour then the fading process can be effective. Finally, we would urge extreme caution in using devices when there is a clear social function to the behaviour. Under these circumstances the effect may be very similar in some ways to that of punishment and extinction in that all we have done is suppress the behaviour of the individual and not given them a different way of meeting their needs. Under these circumstances we would expect the behaviour to return or another form of the behaviour to occur.

We noted in the introduction to this chapter that it is important to ensure that wounds heal as quickly as possible. We do recognise that this is easier said than done, nevertheless it is important to try to pursue any strategy to promote healing. One way in which this can be done is by ensuring that the site of injury is covered. It may take some imagination in order to find the right way of keeping a wound covered but it is worth the effort. This seems to help by promoting healing so getting through the scratch-itch cycle that accompanies healing more quickly and thus avoiding constant scratching that leads to further injury and so on. Additionally, we have noticed that some people who show self-injurious behaviour do deliberately cover up the site of the injury and we believe that this may be a self management strategy that helps people to inhibit the self-injurious response. That is by removing the visual stimulus of the wound, the self-injury seems to occur less. Some of the issues that associate to covering a site of self-injury are those which we have mentioned with regard to protective devices. It does appear that some people come to rely on a wound being covered and will prefer to have bandages, for example, covering their wounds. If this is the case then it may be possible to fade the size of the bandage over time to become a symbolic form of control.

Managing and changing self-restraint

Our past research has shown that self-restraint does appear to be common in children and adults with Cornelia de Lange Syndrome who show self-injurious behaviour. As we have mentioned previously we think this indicates that there may be some lack of behavioural inhibition and thus the behaviour is difficult for the person to control. As we noted in Chapter 6 self-restraint can take many forms and can be restrictive for the individual to a greater or lesser degree. If self-restraint does not appear to prevent the individual from taking part in activities and there are no physical consequences to self-restraint, then it may be it the best strategy to allow the selfrestraint to continue but to ensure that it does not become more restrictive. However, when selfrestraint impairs the individual and may be causing physical harm, for example some people can wind their hands in their clothes so tightly that circulation of the blood is compromised, then it is important to try and reduce the amount of restraint while still maintaining control over the self-injurious behaviour.

The most important thing in changing selfrestraint is not to lose the control over the selfinjurious behaviour. In some ways the selfrestraint is a real asset in that it goes everywhere with the person, it is effective and it is under the person’s control. The trick is to try and reduce the amount of restriction the person is experiencing without losing control. There are a number of research papers and descriptions in the literature of how self-restraint can be decreased to a symbolic level whilst still keeping control over the self-injury28. In order to do this effectively it is important to try to identify what aspect of selfrestraint the person usually prefers. So, for someone who likes to wind their hands in their clothes, it could be that they prefer the tight sensation around the wrists, it could be the precise site of the restraint e.g. around the arm, or the type of restriction that is experienced e.g. total movement restriction or just restriction of the lower arm. Once it has been possible to identify the aspect of self-restraint that a person really needs then other facets of the restraint can be faded i.e. gradually reduced over time whilst leaving the important stimulus intact. (see Box below).

Finding the right form of protective device before fading
In Box Replacing self-restraint with protective devices that can be faded we described how we were able to introduce a new splint for Dawn to control head punching and then gradually increase the amount of flexion at the elbow without the self-injury increasing. However, we did not get it right first time. The first splint we introduced finished at the cuff. When we tried this with Dawn she became distressed and anxious, tried to hit her head and broke the elbow joint within two minutes. The next splint we tried had a band extending from the cuff, over the back of the hand between the thumb and first finger and back to the cuff. Dawn preferred this and was calm when the splint was put on and would hold out her arm to help. In hindsight we should have looked more closely at how Dawn self-restrained. If we had done so we would have seen that she always had either clothes or something else over the back of her hand and running between her thumb and first finger. This was the most important part of the self-restraint for Dawn and we failed to build it into our first splint.

For a summary of the main points about the use of protective devices see this Box.

Four important things about using protective devices
  1. Do not use protective devices as the only form of intervention. They should be combined with a behavioural programme.
  2. Before using devices seek medical and psychological advice. Be aware of irritation to the skin and other problems such as muscle atrophy (wasting).
  3. It is possible that people will become ‘addicted’ to their protective device. Before introducing the device develop a plan about how the device can be faded.  
  4. If people begin to like their devices (hold their arms out, ‘ask’ for the device) they can reward self-injury if they are used after selfinjury has occurred.


It is beyond the scope of this book for us to discuss medications that might be used in order to decrease self-injurious behaviour but we can give some indication of the current thinking on this topic. You will remember that in Chapter 3 we discussed disorders of neurotransmitters that might be related to self-injurious behaviour. The three types of neurotransmitters were opiatergic, serotonergic and dopaminergic. Generally speaking medication that has been associated with decreases in self-injurious behaviour tends to target disorders in these neurotransmitters.

However, we should note that in the research literature that there is very little evidence that medications that have these actions are effective for people with intellectual disability.

There have been very few large scale trials and the evidence that exists tends to be descriptions of single cases. Many authors writing on the topic of medication for self-injurious behaviour conclude that the evidence is more suggestive than conclusive and we would agree with this position. At present therefore, it cannot be said that there is a medication available that can be used for all people with self-injurious behaviour. Further information on possible medications that can be used for self-injurious behaviour is given in Dr. Tom Gaultieri’s book that is listed in the Bibliography and discussion of medications can be found on the Cornelia de Lange Syndrome, UK and Ireland website.


Interventions for self-injury should be selected on the basis of the cause of self-injury. Medical interventions for conditions that give rise to pain and discomfort should be implemented first. For self-injury that occurs because of the resultant stimulation, intervention comprises both reducing the stimulation and presenting competing stimulation. For self-injury that occurs because of social rewards, the intervention should comprise modifying the response to self-injury and teaching alternative responses to displace the selfinjury, Functional Communication Training is the favoured approach. For self-injury that does not appear to be maintained by sensory or social rewards, differential reinforcement may be effective. Protective devices should be used as a last resort with thought given to how they will eventually be reduced. Any intervention should be evaluated with records kept of the levels of selfinjury before the intervention and whilst it is being conducted. For any intervention, advice from a Clinical Psychologist or behaviour analyst should be sought.


  1.   Dossetor, D. R., Coureyer, S., & Nichol, A. R. (1991). Massage for very severe self-injurious behaviour in a girl with Cornelia de Lange syndrome. Developmental Medicine and Child Neurology, 33, 636-644

  2.   Berkson, G. & Tupa, M. (2000). Early Development of Stereotyped and Self-Injurious Behaviors. Journal of Early Intervention, 23, 1-19.

  3.   Favell, J.E., McGimsey, J.F. and Schell, R.M. (1982). Treatment of self-injury by providing alternate sensory activities. Analysis and Intervention in Developmental Disorders, 2, 83-104.

  4.   Murphy, G.H., (1982). Sensory reinforcement in the mentally handicapped and autistic child: a review. Journal of Autism and Developmental Disorders, 12, 265-278.

  5.   Rincover, A. and Devany, J. (1982). The application of sensory extinction procedures to self-injury. Analysis and Intervention in Developmental Disabilities, 2, 67-81.

  6.   Iwata, B.A., Pace, G.M., Dorsey, M.F., Zarcone, J.R., Vollmer, T.R., Smith, R.G., Rodgers, T.A., Lerman, D.C., Shore, B.A., Mazaleski, J.L., Goh, H., Cowdery, G.E., Kalsher, M.J., McCosh, K.C., & Willis, K. (1994). The functions of self-injurious behaviour: An experimental-epidemiological analysis. Journal of Applied Behavior Analysis 27, 215-240.

  7.   Lovaas, I. & Simmons, J. Q. (1969). Manipulation of self-destructive behavior in three retarded children. Journal of Applied Behavior Analysis, 2, 143-157.

  8.   Lerman, D.C., Iwata, B.A., & Wallace, M.D.(1999). Side effects of extinction: Prevalence of bursting and aggression during the treatment of self-injurious behavior. Journal of Applied Behavior Analysis, 32, 1-8.

  9.   Oliver, C., Oxener, G., Hearn, M., & Hall, S. (2001). Effects of Social Proximity on Multiple Aggressive Behaviors. Journal Of Applied Behavior Analysis, 34, 85-88.

  10.   Steege, M.W., Wacker, D.P., Berg, W.K., Cigrand, K. K., & Cooper L. J. (1989). The use of behavioral assessment to prescribe and evaluate treatments for severely handicapped children. Journal of Applied Behavior Analysis, 20, 23-33.

  11.   Gaylord-Ross, R.J. (1982). Curricular considerations in treating behaviour problems of severely handicapped students. In: K.D. Gadow, and I. Butler (Eds.). Advances in Learning and Behavioural Disabilities, Vol. 1, 193-224.

  12.   Mace, F. C, & Belfiore, P. (1990). Behavioral momentum in the treatment of escape motivated stereotypy, Journal of Applied Behavior Analysis, 23, 507-514.

  13.   O’Reilly, M. (1995). Functional analysis and treatment of escape maintained aggression correlated with sleep deprivation. Journal of Applied Behavior Analysis, 28, 225-226.

  14.   Carr, E.G. and Durand, V.M. (1985). The socialcommunicative basis of severe behaviour problems in children In: Reiss, S. and Bootzin, R. (Eds.). Theoretical Issues in Behavior Therapy. New York: Academic Press.

  15.   Carr, E.G. and Durand, V.M. (1985)Reducing behaviour problems through functional communication training. Journal of Applied Behavior Analysis, 18, 111-126.

  16.   Smith, R. G., & Churchill, R. M. (2002). Identification of environmental determinants of behavior disorders through functional analysis of precursor behaviors. Journal of Applied Behavior Analysis, 35, 125-136

  17.   Charlop-Christy, M. H., Carpenter, M., Le, L., LeBlanc, L. A., & Kellet, K. (2002). Using the picture exchange communication system (PECS) with children with autism: Assessment of pecs acquisition, speech, social- communicative behavior, and problem behavior. Journal of Applied Behavior Analysis, 35, 213-231.

  18.   Bondy, A., & Frost, L. (1994). The picture exchange communication system. Focus on Autistic Behavior, 9, 1–19.

  19.   Durand, V. M. (1999). Functional communication training using assistive devices: Recruiting natural communities of reinforcement. Journal of Applied Behavior Analysis, 32, 247-268.

  20.   Sprague, J. R., & Horner, R.H. (1992). Covariation within functional response classes: Implications for treatment of severe problem behavior. Journal of Applied Behavior Analysis, 22, 23-33.

  21.   Oliver, C., Hall, S., & Nixon, J. (1999). A Molecular to Molar Analysis of Communicative and Problem Behaviors. Research in Developmental Disabilities, 20, 197-213.

  22.   Mccord, B. E., Thomson, R. J., & Iwata, B. A. (2001). Functional analysis and treatment of selfinjury associated with transitions. Journal Of Applied Behavior Analysis, 34, 195-210.

  23.   Murphy, G., & Wilson, B. (1982). Long term outcome of contingent shock treatment for self-injurious behavior. In P. Mittler. (Ed). Frontiers of knowledge in mental retardation (Volume 2). Baltimore: University Park Press.

  24.   Young, J. A., & Wincze, J. P. (1974). The effects of the reinforcement of compatible and incompatible alternative behaviours on the self-injurious and related behaviours of a profoundly retarded female adult. Behaviour Therapy, 9, 688.

  25.   Repp, A. C., Barton, L. E., & Brlle. (1983). A comparison of two procedures for programming the differential reinforcement of other behaviors. Journal of Applied Behavior Analysis, 16, 435-445.

  26.   Vollmer, T. r., Iwata, B. A., Zarcone, J. R., Smith, R. G., & mazaleski, J. L. (1993). The role of attention maintained self-injurious behavior: Non-contingent reinforcement and differential reinforcement of other behaviour. Journal of Applied Behavior Analysis, 26, 9-21.

  27.   Oliver, C., Hall, S., Hales, J., Murphy, G. H., & Watts, D. (1998). The treatment of severe self-injurious behavior by the systematic fading of restraints: Effects on self-injury, self-restraint, adaptive behavior, and behavioral correlates of affect. Research in Developmental Disabilities, 19, 143-165.

  28.   Foxx, R. M., & Defrense, D. (1984). ‘Harry’: The use of physical restraint as a reinforcer, timeout from restraint and fading restraint in treating a selfinjurious man. Analysis and Intervention in Developmental Disabilities, 4, 1-13.

Chris Oliver

by Chris Oliver, Jo Moss, Jane Petty, Kate Arron, Jenny Sloneem, Scott Hall

Historial de la página
Última modificación por Gerritjan Koekkoek el 2021/05/25 22:24
Creado por Gerritjan Koekkoek el 2021/05/24 20:38



Sobre el contenido del sitio web

Toda la información contenida en este sitio web tiene únicamente fines educativos. El lugar para obtener consejos médicos específicos, diagnósticos y tratamientos es su médico. El uso de este sitio es estrictamente bajo su propio riesgo. Si encuentra algo que cree que necesita ser corregido o aclarado, por favor háganoslo saber en: 

Envíe un correo electrónico: