An Introduction to Cognitive Behaviour Therapy, 2 edition


Anxiety Disorders: Specific Models and Treatment Protocols


Many of the anxiety symptoms that our clients experience will fall into diagnostic categories, for which there are specific cognitive models and treatment protocols evolved from clinical trials. Major references for key models and protocols are summarised in Table 14.1. The publication dates for some of these references will show you that many of these models are long established.

In this chapter, we will introduce the cognitive model for each of these disorders along with the associated treatment guidelines. You will probably notice similarities across models, but it is important to note the sometimes subtle differences between them. Empirically, the subtle differences matter, as you will see later when we review treatment protocols that are based on the different models.

Table 14.1    Key models and protocols for anxiety disorders

Anxiety disorder


Specific phobia

Kirk & Rouf (2004)

Panic disorder and agoraphobia

Clark (1986, 1999); Wells (1997)

Health anxiety

Salkovskis & Warwick (1986); Warwick & Salkovskis (1989)

Social anxiety

Clark (2002); Clark & Wells (1995); Wells (1997)

Generalised anxiety disorder (GAD)

Borkovec & Newman (1999); Borkovec et al. (2002); Wells (1997, 2000)

Obsessive-compulsive disorder (OCD)

Salkovskis (1985, 1999); Wells (1997)

Post-traumatic stress disorder (PTSD)

Ehlers & Clark (2000)



Specific phobia

There is, as yet, no evaluated ‘cognitive model’ of specific phobias, although a preliminary model has been proposed by Kirk and Rouf (2004). In brief, they suggest that clients with specific phobias (for example, of a particular animal or situation or of blood) are hypervigilant for threat cues. Thus, the cycle begins with:


  • focusing on the perceived threat with selective attention for fear cues. This increases the likelihood of –
  • perceiving a threat, whether this is actually what they fear (e.g. a spider or blood) or a misinterpretation (e.g. a piece of fluff on the carpet or a splash of tomato ketchup). This triggers the phobic response, which has both psychological and physiological elements. This, in turn, reinforces –
  • over-estimating the probability of harm and under-estimating ability to cope (Beck, Emery & Greenberg, 1985), which then maintains the fear that drives hypervigilance.

The primary cognitions (fear of an object or a situation) exacerbate:


  • physiological arousal, which can be further interpreted as threatening; and
  • safety behaviours, such as overtly avoiding certain places (e.g. shops, zoos) or situations (e.g. writing in public) or covertly avoiding feared situations (e.g. wearing excessive amounts of insect repellent to ward off spiders). These then prevent anxious predictions from being disconfirmed, the fear remains unchallenged and the sufferer remains hypervigilant.

Beliefs about the meaning of the phobia (secondary cognitions) can also heighten anxiety, with beliefs such as ‘I am foolish’ or ‘I am going mad’ (see Figure 14.1).

Figure 14.1    A cognitive model of specific phobia

Karen had always been fearful of wasps. The thought of them made her shudder, and the sight of one triggered panic. If she saw one, she could not think straight and would run away – recently she had left her youngest child outside a shop when she ran for cover. She coped by doing all she could to avoid wasps: never going into the garden during the summer months; not allowing her children to eat sweets outdoors in case this attracted wasps; keeping the doors and windows closed in her home. It was hard for her to verbalise what made her so afraid, but she had an image of being unable to escape angry wasps caught in her hair.

Helping a client overcome a specific phobia involves:


  • Exposure: a person can only overcome a fear by facing it. This can be done in imagination as well as in reality – although exposure in reality is usually more effective. Facing fears is typically carried out in a graded way so that the client is challenged but not overwhelmed by the task, and in CBT, exposure is followed by debriefing so that the client and therapist can make the most of the cognitive and emotional opportunity for change which exposure affords.
  • Decreasing focusing on the perceived threat: this can be achieved by using distraction and also by setting up behavioural experiments (see Chapter 9) to evaluate the consequences of reducing the amount of time spent checking or anticipating the worst. Many clients will hold beliefs such as ‘If I don’t keep a look out for spiders, I’ll be taken by surprise by one’ and very often the thought continues in a catastrophic direction: ‘I won’t cope’. They can be encouraged to test this prediction using BEs and in this way can often gain confidence that it is unnecessary to remain hypervigilant.
  • Reducing safety behaviours: this can also be achieved through behavioural experiments (often graded) to test the predictions of harm.
  • Addressing misinterpretationsby teaching decentring and cognitive reappraisal of situations. This is relevant to both primary and secondary cognitions.

Karen wanted to tackle her phobia head on and was prepared to try to confront a wasp. She and her therapist devised a hierarchy of graded behavioural experiments and made predictions about her probable responses to increasingly challenging wasp-related tasks. The experiments began with looking at photographs of wasps and progressed through to releasing a wasp from a jar into the garden. When she got as far as releasing wasps, she realised that they flew from rather than to her: her images of the wasps caught in her hair then diminished. This success gave her courage to work towards dropping her safety behaviours. Each lent itself to behavioural testing, and, through this, she discovered that her anxiety was heightened by her excessive hypervigilance. She also learned that if she left windows open, or let her children eat sweets outdoors, she encountered wasps, but far fewer than she had predicted, and that she coped better than she had anticipated.

Blood and injection phobias are different from other phobias as they have a different physiological maintaining pattern. As blood pressure drops in both these disorders, the client experiences unpleasant bodily and mental sensations and the prospect of fainting is real. These phobias are dealt with in Chapter 10.

Panic disorder

A prominent cognitive model of panic disorder is that of Clark (1986), which identifies the maintaining factors as:


  • Catastrophic misinterpretation of bodily sensations (particularly those associated with anxiety) as indicative of impending mental or physical harm, such as an imminent stroke or heart attack.
  • Safety behaviours employed in order to reduce the likelihood of catastrophe. These include frank avoidance, such as not going to certain places or events, and subtle avoidance, such as holding on to someone to avoid collapse or sucking ginger to avoid vomiting.
  • Selective attention as sufferers become highly sensitised to ‘dangerous’ sensations or situations, and their attention becomes biased towards them.

Figure 14.2    A cognitive model of panic disorder

When Wendy had a panic attack, her chest tightened, she fought for breath and she trembled. She felt pain in her chest and arms and experienced tunnel vision. She thought she was having a heart attack and could die. She avoided any situation where she feared she might exert herself as she feared provoking a heart attack. For example, she no longer did the weekly supermarket shop, nor took her children to the park. She knew that she was becoming physically unfit, and this heightened her fears.

The management of panic disorder typically involves:


  • Generating less catastrophic explanationsfor the origin of feared symptoms and less catastrophic predictions of the consequences – for example, attributing chest pains or a racing heart to anxiety, which is not harmful.
  • Setting up behavioural experiments(i) to discover the benign origin of an unpleasant sensation: for example, asking a client to exert himself to trigger feared sensations such as muscular pains or palpitations; and (ii) to test the validity of the new perceptions that have been generated through cognitive challenging – such as ‘I’m feeling the symptoms of anxiety and this will pass’.
  • Reducing safety behaviours: this can be achieved through cognitive and behavioural work. Cognitive interventions can be used to generate and explore new explanations for client coping (i.e. explanations other than the ones that drive safety behaviours). These, in turn, can be reinforced through behavioural testing. For example, through preliminary cognitive work, a client began to consider it possible that she could walk around the supermarket without leaning on a trolley for safety. However, shopping without the trolley actually consolidated her confidence.

Wendy’s therapist raised the question of whether her muscular pain and difficulty breathing could result from the muscular tension associated with her state of heightened anxiety. Wendy was eventually convinced of this when she agreed to exercise in the session – even though the prospect had frightened her and had initially provoked the predicted symptoms. Once she had exercised in the session and discovered that she was a bit breathless but otherwise okay, her belief changed to: ‘I am not having a heart attack but I am highly anxious and it will pass’. Once she had gained this new perspective, she engaged in increasingly physically demanding activity. She began by exercising in the session and then between sessions, and grew confident that a racing heart or muscular tension was not harmful to her. Ultimately, she attended a gym regularly and ceased avoiding situations that she feared would over-exert her.

Hypochondriasis or health anxiety

The cognitive understanding of health anxiety centres on the enduring catastrophic predictions concerning future health issues and preoccupation with physical symptoms (i.e. focus of attention towards the perceived threat). In itself, the fear of developing physical illness can exacerbate alarming physical symptoms, for which there is also likely to be selective awareness. This leads to high levels of anxiety. Sufferers of health anxiety tend to engage either in reassurance-seeking behaviours or in avoidance of situations they predict will heighten their anxiety (safety behaviours).

Reassurance-seeking is ineffective in changing health anxiety as it reflects a reliance on external support. Sufferers fail to learn to assure themselves, and health concerns remain intact; there is always room for doubt – ‘Maybe I didn’t fully explain my symptoms to my doctor …’, ‘Maybe I didn’t fully understand what he said – maybe he said that I could have cancer!’ Furthermore, it is not unusual for a person who repeatedly complains of potentially harmful physical symptoms to be subjected to various tests, which can be interpreted as proof of real ill-health.

Maintaining cycles can take several forms:


  • Avoidance of situations that trigger health fears: which means that the sufferer does not learn that there is no need to be over-concerned, that such situations are tolerable.
  • Turning to others, such as medical specialists or family members, for comforting reassurances (safety-seeking).
  • Scanning: focusing on the perceived threatwith hyper-awareness of physical sensations such as heart rate, numbness, pain, etc. This means that benign sensations can be misinterpreted and can feed the health anxiety.
  • Checking: this can be related to the sufferer’s body (looking for moles, lumps, etc.) or to external information (reading medical literature, for example). Either way, it is then all too very easy to discover things that can fuel alarm.

Tina woke each morning with thoughts that she might have breast cancer. She tried unsuccessfully to avoid the media and, instead, noticed every article about cancer. Each day she felt compelled to check her breasts, armpits and neck for signs of lumps or enlarged glands. She believed that it would be dangerous not to check, as a missed tumour could become malignant. She always found something that gave her concern, so she persuaded her partner to ‘double-check’ for her. The relief each time she was reassured was exhilarating, although short-lived.

As with GAD, the interpretation of the preoccupation needs to be explored too, as some clients hold beliefs about their continued concern, such as ‘If I am vigilant for signs of illness I will be OK’ or ‘If I think about my illness I will bring it on’.

Figure 14.3 illustrates how health anxiety can be maintained through:


  1. Avoidance
  2. Reassurance-seeking
  3. Scanning.

The treatment approaches for health anxiety reflect these maintaining cycles and incorporate:


  • Defining and challenging the content of the catastrophic prediction: as therapist, you will need to tease out the anticipated worst outcome for your client: for example, abandonment or protracted physical or mental torment. For some, the worst scenario is not simply illness or even death, but the nature and consequences of the illness or of death – so it is important to understand what illness or death means to your client. For example, he may not be worried about dying from a heart attack (perceived as rapid and dignified) but may be preoccupied with the fear of dying slowly from a neurological disorder, needy and incontinent, or he may not be worried about dying from a heart attack as much as he is worried about surviving a heart attack only to be left disabled.
  • It is also wise to explore possible superstitious thinking (meta-cognition) such as, ‘If I don’t think/do think about the illness, I will be protected from it’.

Figure 14.3    A cognitive model of health anxiety


  • Testing unhelpful health-related beliefs, such as: ‘Chest pain means my heart is weak’, or ‘Every worrying symptom has to be checked by my doctor’. This can be achieved through cognitive interventions and behavioural experiments.
  • Reducing safety behaviours(reassurance-seeking, scanning and avoidance). Sometimes, an explanation of the role of these behaviours allows the client to reduce them; in other instances, it is necessary to challenge beliefs concerning safety behaviours, possibly via behavioural experiments. This can also be relevant to carers who collude with unhelpful behaviours.
  • Theory A vs. Theory B, is a useful approach to help clients gain an alternative perspective (see Chapters 9and 13).

Tina’s most awful thought was that she would die a lingering death that would torment both her and her loved ones. With help, she was able to review this prediction, but her relief stemmed less from the statistical data about her risk of dying of cancer than from her re-appraisal of her coping resources. Once she believed that she could tolerate a protracted death (however undesirable this might be), she focused less on her health and her preoccupation diminished.

She also believed that if she did not get reassurance, she would be unable to withstand the uncertainty, and this would undermine her ability to function. She refined her definition of ‘to function’, specifying what activities would be impaired and how. She then conducted behavioural experiments to test the validity of her predictions and learnt that she could use distraction to help her with any tasks that she needed to complete and that she often functioned better when she kept her health worries at bay. She then built on this series of behavioural experiments by testing her prediction that she would become uncontrollably preoccupied by thoughts of cancer if she read or watched the media. She discovered that she could cope.

In addition, Tina’s partner agreed to stop giving her reassurance and, although Tina initially felt uncomfortable about this, she quickly learnt to assure herself.

At her first review session, Tina reflected on her beliefs in Theory A (that she would get breast cancer and would be unable to cope) and Theory B (that her preoccupation and safety-seeking kept health concerns foremost in her mind). She concluded that she now felt that Theory A was unlikely and she believed 80% in Theory B.

Social anxiety

As we said in Chapter 13, social anxiety can be in the more severe form of social phobia or the milder ‘shyness’. Some social anxieties are quite specific, only arising when meeting a new person or an attractive person, for example, or only when carrying out particular tasks such as writing or eating in public. Sometimes the anxiety is more pervasive.

Cognitive models for social phobia have been developed (see Figure 14.4), most notably by Clark and Wells (1995), which, suggest Butler and Hackmann (2004), are readily applicable to ‘shyness’. Models of social phobia incorporate the following:


  • Perceived social danger.Typical assumptions and predictions of the socially anxious person are: ‘If I talk to them they will find me boring and reject me’, or ‘If I don’t get this just right, I will be humiliated’. Essentially, these are fears centring on being negatively judged and on not coping.
  • Self-focused attention.The social anxiety cycle is propelled by intense self-awareness, which can also manifest as self-referent imagery (Hackmann, 1998). This heightened self-consciousness is distracting and, thus, disabling as it is not possible to properly review a situation and engage in productive problem-solving. For example, someone who is preoccupied with their ability to perform in front of friends might not be able to deal with a minor crisis, as all their attention is taken up in self-evaluation. Being so focused on self also prevents the sufferer from reviewing situations objectively, and then it is all too easy to (negatively) misinterpret the reactions of others.

Figure 14.4    A cognitive model of social anxiety


  • Emotional reasoning.The intense introspection about the sensations of anxiety renders the sufferer acutely aware of symptoms such as shaking and blushing. This heightens his self-awareness and he feels increasingly self-conscious. Because he feels self-conscious, he assumes others can see his symptoms as clearly as he feels them – and he assumes that others will judge him negatively.
  • Safety behaviours.Understandably, the socially anxious person will attempt to avoid predicted humiliation or embarrassment by avoiding social contact – for example, focusing on a task such as helping out in the kitchen during a party or avoiding eye contact during conversation. Of course, in doing so the social fear is not addressed and remains intact, ready for the next social challenge. In some instances, the safety behaviour is doubly counter-productive. For example, spending the evening in the kitchen at a party or avoiding eye-contact might well give the impression that a person israther odd.

Bette anticipated rejection. Her prediction in social situations was that others would realise that she had nothing to offer and would not want to know her. If someone did show interest, she discounted it: ‘They don’t know the real me’, or ‘They were just being polite’. As far as possible, she avoided social gatherings, and when she attended them, avoided eye contact but could ‘feel’ the critical gaze of others. She tended to busy herself attending to the practical needs of guests. If she became involved in conversation, the intensity of her negative intrusive thoughts rendered her unable to chat.

Interventions for social phobia involve:


  • Re-focusing attention away from introspection. This strategy was particularly elaborated and evaluated by Wells and Mathews (1994) and involves switching attention between different sources of sensory information (auditory, visual, sensory and so on). This is first practised in sessions, then between sessions until the client becomes adept at shifting his attention away from himself.
  • Developing an assertive or compassionate inner voiceto combat the harsh criticism that sufferers predict from others (Padesky, 1997; Gilbert, 2000). Thus, you would encourage your clients to address themselves in an empathic and understanding way, for example: ‘It is understandable that I feel like this, I am not very confident yet. It’s okay to pace myself – I don’t have to be able to tackle the most difficult situations right now.’
  • Cognitive re-evaluationof the cognitions relating to perceived social danger and emotional reasoning, including behavioural experiments. Particularly useful are (i) the use of videoed sessions, which allow clients to evaluate the severity of their overt anxiety symptoms, and (ii) modelling of feared consequences by the therapist. The latter means that you might have to appear to have blushed, sweated or even wet yourself in a public place – but be comforted that this seems to go unjudged by the general public.
  • There is also a useful series of questions often used to help the socially anxious client gain a more balanced perspective on problems like blushing, shaking, stammering etc:


o    Are the symptoms you fear actually as likely to happen as you fear?

o    Even if they do happen, will it actually be as severe as you imagine?

o    Even if it is, will other people actually notice?

o    Even if they do notice, will they interpret it in the way that you fear they will?

o    Even if they do see it that way, so what? Is it possible you can survive it and get on with your life?

Bette learnt several strategies to combat her social anxiety. First, she described her worst-case scenario and challenged her predictions that (i) she would almost certainly be criticised and ostracised and (ii) that she would not cope with criticism but would accept it and become deeply depressed. Cognitive restructuring, developing a strong but caring inner voice and role play with her therapist helped her to conclude that she was unlikely to be openly criticised, but even if she were, she could stand up for herself and not spiral into despair. She also learnt strategies to refocus attention from her negative, self-referent thoughts. In addition, she carried out behavioural tests: she allowed her therapist to log the number of people who looked at her critically when they attended a social gathering – to her surprise, the therapist noted no one. Finally, she engaged in a series of assignments focused on not tending to guests’ practical needs (i.e. dropping her major safety behaviour). As she progressed through the hierarchy of assignments, she built her confidence that she could mix socially.

Generalised anxiety disorder (GAD)

As described in the previous chapter, GAD is defined as chronic, excessive anxiety and worry pertaining to a number of events or activities (DSM-IV-TR; APA, 2000).

Sam was 64 years old and felt he should be looking forward to his retirement – his wife certainly was. However, as usual, he was beset with worries about it: What if he and his wife did not get on? What if their financial planning had been insufficient? He found his worrying shameful but familiar, and he could not remember a time when he had been free of it, just times when it was slightly better or worse.

Cognitive models of GAD give prominence to worry as a key cognitive factor. There are several possible mechanisms for persistent worry:


  • Focusing attention towards the perceived threat can be an attempt to avoid addressing a more distressing fear, as ‘What if?’ statements are superficial to the real concern. The real concern would be revealed by answering the ‘What if’ question (Borkovec & Newman, 1999).
  • It can also reflect an attempt to avoid facing uncertainty which is felt to be intolerable, even in small amounts (Ladouceur et al., 2000; Dugas, Buhr & Ladouceur, 2004).
  • The meaningof the worry itself (the meta-cognition) can perpetuate worry (Wells 1997, 2000). Wells calls this Type II worry to distinguish it from worries about everyday concerns (Type I worry). Type I worries, such as ‘What if I don’t have enough money!’ can be perpetuated by Type II worries. Type II worries can be positive, such as the superstitious: (‘If I worry, bad things won’t happen’) or a misconception about worry (‘If I am worrying, I am doing something useful’) or the concerns can be negative (‘All this worry will drive me crazy!’). Such beliefs would increase the likelihood of engaging in worry and continuing to worry could become a safety behaviour. In Sam’s case, he believed that worry would better prepare him to cope – that he would not be ‘caught out’ by bad luck. So, uncomfortable as it was, he felt compelled to worry. Type II worries can also be negative and alarming (e.g. ‘I will go crazy’), which can trigger more worry (Wells, 1997, 2000). Again, in Sam’s case he was ashamed by his tendency to worry, he felt that it was weak and unmanly. This, therefore worried him, raised his anxiety levels and increased his vulnerability to worry. A very vicious cycle.
  • Worry undermines the ability to problem-solve, so that a person loses confidence in their problem-solving ability, which supports further worry (Dugas et al., 2004).

Treatment for GAD focuses on breaking the worry cycle by understanding and eliminating unhelpful worry and then helping your client address the underlying fears.

Figure 14.5    A cognitive model of GAD

The required steps include:


  • Normalising worry and worry awareness training. The former means helping your client understand that a degree of worry is normal and even helpful but that their level of worry has probably become unhelpfully exaggerated. Worry awareness training is particularly useful for those for whom worrying has become habitual. It simply involves self-monitoring so that frequencies, triggers and patterns of worry become clear.
  • Overcoming avoidanceby encouraging articulation of the fear (e.g. personal harm or harm to loved ones) beneath the ‘What if?’ question, and then helping your client address these fears rather than avoid facing them.
  • Accepting uncertainty.Butler and Rouf (2004) recommend focusing on challenging the cognitions that reflect intolerance of uncertainty rather than trying to review the likelihood of the feared events actually happening. Therefore they stress the importance of helping clients simply accept uncertainty. This means that, as a therapist, you focus on clarifying the alarming answer to ‘What if?’ rather than debating the probability of the worst-case scenario. Behavioural experiments involving clients testing out their negative predictions about enduring uncertainty can then be used to help clients accept doubt or ambiguity (see Butler & Rouf, 2004).
  • Identifying and testing unhelpful cognitionsconcerning worry. This would involve first identifying meta-cognitions, such as ‘Worrying will damage my heart’, or ‘I must worry so that I am never caught unprepared’, followed where appropriate by behavioural testing. The aim of the behavioural experiments is not to control the worry but to change beliefs about it.
  • Teaching alternative strategiesto worrying, such as problem-solving or distraction, or limiting the time permitted for worrying, where the aim is to help your client have the experience of disengaging from worry and learning that all is well. For example, a client who typically spent five hours in the evening worrying about events in the next day might develop a plan to worry only during the 30 minutes between getting home and supper, with other worrying thoughts ‘held over’ until the following day. Distraction could be a useful strategy to help her resist engaging in worrying, and thus she would have the experience of learning that she could get through the evening without being so caught up with her concerns. Another useful strategy for helping clients ‘put worries aside’ is writing worries on a paper which is then destroyed (Butler & Rouf, 2004).

Sam was encouraged to articulate his fears: he said that he was terrified that his wife would realise he had nothing to offer her and would leave him; that they would run out of money and be unable to afford decent health care; that his wrong decision in taking retirement would cost him his marriage, his security and, worst of all, would prove that he was useless. By using cognitive strategies, he was able to both de-catastrophise specific negative thoughts, but also to appreciate his resilience (he had managed many personal and business crises in the past) and that enabled him to tolerate the uncertainty of his future. Throughout, there was a theme of shame, worthlessness and responsibility, and these general negative themes were also reappraised.

Sam carried out a behavioural experiment comparing his ability to problem-solve when he focused on his worries with his ability to do so when he distracted himself from them. He learned that worrying was counterproductive and recognised that it had become both a habit and a source of comfort, as he believed that he could ward off bad luck by worrying. Once he understood this, he readily distracted himself from worrying, thus breaking an unhelpful pattern.

Obsessive-compulsive disorder (OCD)

Unwanted intrusive thoughts, in the form of words, images or impulses, are not in themselves pathological (Rachman & de Silva, 1978), so we would not attempt to challenge them, but the response to them can be unhelpful. Cognitive models of OCD share the basic premise that intrusive thoughts are in themselves normal but become a problem when they are interpreted as indicating that something bad might happen and that the sufferer is responsible for preventing it. To manage this fear, the sufferer engages in safety behaviours (avoidance, reassurance-seeking and cognitive or motor rituals), which prevent him from learning that his worries are not accurate or that his anxiety will actually decline without his performing rituals. The aim of CBT is for the client to learn that such intrusive thoughts do not indicate a need for action and can safely be ignored.

The most common obsessional worries relate to:


  • fears of contamination, such as infection from touching a dirty cloth or surface, leading to washing or cleaning rituals;
  • fears of missing something potentially dangerous (such as electric switches which have been left on, or an unlocked front door) leading to checking and/or repeating rituals;
  • over-concern with orderliness and perfection, leading to repeating actions until things’ feel right’;
  • fears of uncontrollable and inappropriate actions such as swearing in public, or sexual or aggressive behaviour, leading to unhelpful attempts to control thoughts.

The most common safety behaviours are:


  • motor rituals: e.g. cleaning, checking and repeating actions;
  • cognitive rituals: neutralising ‘bad’ thoughts by thinking other thoughts (e.g. prayers or ‘safe’ incantations, or other ‘good’ thoughts);
  • avoidance of situations, people or objects that trigger the obsessional worries;
  • seeking reassurance about the worries from family, doctors or others;
  • thought suppression.

Most OCD sufferers have motor rituals, but some have predominantly cognitive rituals with few if any motor rituals (so-called ‘pure obsessions’ – a presentation that may be harder to treat).

Vince had always been very cautious and was proud of his high standards for safety. However, since a promotion (with responsibility for ensuring departmental security), his safety checks had become exaggerated and he was now struggling to leave the building at night. He often returned five or six times to recheck – occasionally driving in from his home. He tried unsuccessfully to put the worrying thoughts out of his mind. His fear was that insufficient caution would result in a catastrophe for which he would shoulder the blame. He thought that the shame of this would destroy him.

It has been suggested by the Obsessive-Compulsive Cognitions Working Group (1997), that the key cognitions in OCD are:


  • thought–action fusion: the idea that having a ‘bad’ thought can result in ‘bad’ consequences (e.g. if I think about harm coming to someone, that may make that harm happen in reality); or that having a thought about something ‘bad’ is morally just as bad as carrying out a bad action;
  • inflated responsibility: an assumption that one has the power and the obligation to prevent bad things from happening;
  • beliefs about the controllability of thoughts: e.g. the belief that one ought to be able to control ‘bad’ thoughts;
  • perfection: the dichotomous assumption that only the best is effective or acceptable;
  • overestimation of threat, which is often related to;
  • intolerance of uncertainty: a belief that things can and must be certain, e.g. I ought to be able to be surethat an action is safe.

As with other anxiety disorders, thoughts about negative thoughts (e.g. ‘There must be something fundamentally wrong with me for having such thoughts’) can heighten anxiety (Wells, 2000). Emotional reasoning (the assumption that feelings are a reliable source of information about a situation – for example, ‘I feel anxious, therefore this must be a dangerous situation’) is also common amongst sufferers of OCD (Emmelkamp & Aardema, 1999) (see Figure 14.6).

Figure 14.6    A cognitive model of OCD

Interventions for OCD incorporate:


  • Exposure and response prevention (ERP).This is the best-established intervention for OCD. The aim is for the sufferer to expose himself to the feared situation (e.g. something ‘contaminated’) without engaging in his usual safety behaviour (e.g. washing). Although originally conceived as a behavioural intervention, ERP is readily adapted to a cognitive model in which it is seen as a BE through which the client learns that his obsessional predictions of disaster are not justified and that he can tolerate distress. Rituals are essentially safety behaviours and therefore ERP is exactly analogous to reducing safety behaviours in any other anxiety disorder. It is possible that family, friends or professionals collude or assist with the safety behaviours, and it may therefore be necessary to involve them in this aspect of therapy.
  • Testing unhelpful thoughts and beliefs related to the intrusions, such as ‘If I think it, it will happen’, or ‘I am responsible for the welfare of others’, by using the cognitive and behavioural strategies described in earlier chapters. The continuum method, or scaling, can be particularly helpful in addressing the extreme perspective of the perfectionist (see Chapters 13and 17). In OCD, the intrusive thoughts themselves are not directly challenged in this way (since they are seen as normal phenomena) – it is the negative thoughts about intrusions that need testing.
  • Theory A vs. Theory B.As with health anxiety, this intervention can be useful in highlighting a benign perspective. The aim is to learn that OCD is not about the need to prevent a real threat but rather about being excessively worried about such a threat.

Vince’s most unhelpful belief was: ‘I am wholly responsible for any crisis that arises at work’. He challenged this by recognising the cognitive biases in his thinking and by constructing a ‘responsibility pie’ (see Chapter 17), which helped him to apportion responsibility more realistically. However, he also had to work on the dichotomous thinking that underpinned his unrealistically high standards: continuum work (see Chapter 17) helped him to become more flexible. He addressed other key beliefs such as ‘I will be destroyed if I am to blame’ using standard cognitive interventions.

Feeling more confident that he could tolerate the worst-case scenario, he agreed to a programme to reduce his safety behaviours. This incorporated an agreement that his wife would not reassure him when he was at home feeling uneasy about his department’s security. He struggled initially with refocusing his thoughts away from catastrophic possibilities. He kept (meticulous) diaries of his experiences, and these showed clearly that he felt less anxious and more content on the days that he reduced safety behaviours and catastrophic thinking. He also recognised that catastrophes never ensued at these times, giving him evidence that his safety behaviours were not necessary.


Several cognitive models for PTSD have been developed over the past decade but the most prominent is probably that of Ehlers and Clark (2000). Cognitive models tend to put an emphasis on:


  • the emotion of fear as the main affective component of the model, although it is increasingly recognised that emotions such as guilt, grief and shame can be prominent too;
  • vivid memories that promote the experience of the danger still being current;
  • these memories being disconnected from an intellectual understanding of the trauma which might otherwise put them into perspective and enable the sufferer to better tolerate them;
  • memories being experienced very visually, although recollections of a traumatic event can also be experienced in other sensory modalities (as sounds, physical sensations and smells, for example);
  • memories also being experienced as nightmares or as the very vivid flashback.

These memories are emotionally highly provocative and remain so for several reasons:


  • Safety behaviours: in an attempt to manage high anxiety, PTSD sufferers often use behavioural and mental avoidance to inhibit the memories.
  • This prevents processing of the memory(i.e. reviewing the traumatic content so that it can be linked with information about time, place and outcome and therefore put in the past), so the memory remains a disconnected, emotionally charged recollection which, in itself, triggers high levels of distress and arousal.
  • Misinterpretations: unhelpful appraisals of the traumatic experience (e.g. ‘This proves that no man can be trusted’; ‘I brought this on myself through carelessness’) or of the PTSD symptoms (e.g. ‘I am weak’, ‘I am going crazy’) can further worsen the distress associated with the intrusions and can therefore lead to increased arousal and more safety behaviours.

In addition, the cycle of PTSD can be maintained by:


  • spontaneous and vivid imagery which can be powerfully aversive: thus, sufferers will try to avoid situations which trigger them;
  • selective memory processes, which distort recollections such that they are biased towards negative aspects of the trauma – thus heightening distress;
  • overestimation of danger: it is not unusual for trauma victims to overestimate current threats to their safety – in turn further promoting both heightened arousal and safety behaviours.

See Figure 14.7.

Figure 14.7    A cognitive model of PTSD

Alistair had been involved in a road-traffic accident when his car tyre exploded at high speed. He had had a lucky escape. Eight months later, he still experienced vivid memories of his car flipping over, memories of the sights, sounds and smells as if it were happening again. For him, everything went black again, he heard metal scraping along the road, sounding like an old train braking; he felt as if he were hanging upside down and each time he again had the thought: ‘I’m going to die’. He was particularly likely to have flashbacks when he smelled petrol or when he returned to the area where the accident took place. Therefore, although he still drove his car, his partner would always refuel, and he never drove near the site of the crash.

Treatment for PTSD based on the cognitive models involves addressing:


  • Spontaneous imagery.Strategies are introduced to diminish the very high levels of arousal associated with the images, so that they can be processed and contextualised. This means that they form memories which are subject to rational appraisal. The sense of present threat is then eliminated, by placing the image in context of time, place and longer-term outcome. This is often achieved by using cognitive restructuring while ‘reliving’ the trauma (Grey, Young & Holmes, 2002) or cognitive processing therapy (Resick & Schnicke, 1993; Ehlers et al., 2003) where the client writes a detailed account of the traumatic experience for cognitive review. Along with constructing more helpful interpretations of the traumatic memory, clients are encouraged to expose themselves to real-life situations linked to the trauma so that they might challenge their anxiety-provoking cognitions in vivo.
  • Safety behaviourscan be reduced by reviewing unhelpful beliefs and field-testing new possibilities, as with other anxiety disorders.
  • Misinterpretationscan be reappraised by reviewing these conclusions and generating plausible alternatives, again using ‘standard’ CBT interventions.
  • Selective memory processescan be usefully addressed, as can all of the cognitive biases, by teaching the client the technique of decentring, or standing back and viewing the cognitions at a distance. This allows the sufferer to gain a wider, more balanced perspective, which is less distressing.
  • Reclaiming your life’.Many trauma victims neglect activities that contributed to their well-being pre-trauma – activities such as socialising or exercising. Thus, assignments are negotiated which are designed to re-engage survivors with meaningful activities – the aim of this is to improve the quality of their lives and their mood as well as helping them re-establish a more normal life style.

Alistair’s intrusive memories responded well to cognitive restructuring. Ultimately, his therapist helped him to talk through his experience as if it were currently happening, pausing at the emotional ‘hot spots’ to review his cognitions in the light of what he now knew. By doing this, Alistair was able to challenge his most salient thought: ‘I am about to die’. He was able to remind himself that he got out of the crashed car with few injuries and, in doing so, he reduced the intensity of the flashbacks. He was also able to revise a shameful belief which had developed after the accident, namely that he was responsible for the incident. This further reduced the anxiety that the memories evoked.

He was gradually able to revisit the site of the crash – first with his partner and later alone – and to talk and read about accidents. His predictions that he would have flashbacks were not borne out, and his confidence returned. His avoidance of smelling petrol was more difficult to tackle, as fears associated with smells are particularly resilient, but because he had learnt to be less afraid of flashbacks, he tolerated the occasional vivid memory without undue distress, and he no longer felt the need to avoid the smell of petrol.

The precipitants of PTSD may be impersonal, such as a natural disaster, or perceived as highly personal, for example when a person has been assaulted in some way. In cases of personal attack, there is likely to be a need for greater sensitivity to interpersonal relationships both within and outside the therapy setting. Some clients will have experienced sexual assault and, clearly, there needs to be sensitivity when discussing this and other sexual relationships.


Anxiety disorders can present as discrete problems, in combination with other anxiety disorders or as co-morbid with other problems – for example, the high standards of the person with OCD might predispose her to an eating disorder, the chronicity of an anxiety problem might give rise to depressed mood, while coping strategies such as comfort eating or drinking can develop into difficulties in their own right. Remember to take this into account during your assessment, and throughout treatment remain aware that other problems might exist.


The previous chapter reviewed a generic understanding of anxiety disorders, while this chapter has focused on specific models and the treatment approaches linked to them. In your practice, you will need to be aware of both the generic and specific approaches so that you can be flexible and responsive to your client’s needs. The models give an elegant and invaluable understanding of particular anxiety disorders, while the generic overview provides you with the ‘first principles’ that you can fall back on if the models and protocols do not meet your client’s needs.

Potential problems when working with specific models and treatment protocols

Assuming the validity of a diagnosis without carrying out a full assessment and then adhering to a treatment protocol

Although many of your clients will fulfil criteria for a particular diagnostic group and will benefit from a protocol-driven approach, do not presume this without carrying out a proper assessment. There will be times when the referrer’s diagnosis or your first impressions are wrong.

Trying to force a client’s experiences into a specific model

Keep a curious and open mind during your assessments. If your client’s presentation does not fit neatly into the model that you anticipate being relevant, perhaps the model is invalid for this person. In such instances, a generic formulation (Beck et al., 1979) will be appropriate and this will guide you in deciding which interventions might be best.

Sticking too rigidly to a protocol when the client is not responding well

While it is important to follow a protocol, there will be individual differences amongst your clients and aspects of their presentation which, at some point, may not fit with the protocol. In some instances, the deviation will be sufficiently marked for you to have to reassess your client and consider if the protocol offers the optimum approach. At other times, staying with the protocol will be in your client’s best interest, but you may need to adapt it slightly – for example, introducing a session on specific skills training (assertiveness, time management and so on) or temporarily diverting to tackle an issue that appears to obstruct progress, issues such as excessive anger, unresolved grief or flashbacks.




  • In general, there are well-established and very specific models for understanding the different presentations of anxiety disorders and there are tried and tested protocols for managing them.
  • Where possible, the use of protocols should be the first choice of intervention. Therefore you need to be familiar with them.
  • You also need to be familiar with the ‘first principles’ of anxiety. This means that you need to have

i    a generic understanding of the factors that contribute to, and maintain, anxiety-related difficulties so that you can formulate the problems of your clients who do not fit the standard models and,

ii   an appreciation of the range of management strategies you can apply to breaking the cycles that maintain anxiety disorders so that you can be flexible in your approach when this is necessary.

Learning exercises

Review and reflection:


  • What is your reaction to seeing the models underpinning anxiety disorders? Do they fit with your experience? Do they make sense for you? If not, what is the difficulty you have in linking this theory to your clinical experience?
  • What about the protocols? Again, do they make sense to you? Can you think of clients who might benefit from your applying them more vigorously? Do you have problems in seeing how they could work for your clients? If so, what difficulties do you foresee?

Taking it forward:


  • Read more about models and protocols – go back to source materials to ensure that you properly understand them.
  • Review your anxious clients in the light of understanding more about the cognitive models that underpin anxiety disorders. Review your formulations of their difficulties.
  • If your client fits a model, check that you are making the most of there being a protocol to guide you.
  • If you are not used to closely following a protocol, try out a ‘test case’. Really familiarise yourself with the relevant intervention and stick as closely to the protocol as possible. Discover how this works for your client and for you. If you have difficulties adhering to a protocol that is working for your client, then explore your own assumption about models and protocols.
  • If your client does not fit a model, consider how you will formulate, and help your client manage the problem, given what you understand about the nature of anxiety disorders and the strategies available to you.
  • Use your supervisor’s support and, if necessary, ring-fence time for discussing your questions and difficulties in working with your anxious clients.

Further reading

Wells, A. (1997). Cognitive therapy of anxiety disorders: a practical guide. London: Wiley Blackwell.

A well-established ‘basic’ text, which addresses the various anxiety disorders in a practical and informative way. It is comprehensive and successfully marries theory and practice.

Butler, G., Fennell, M., & Hackmann, A. (2008). Cognitive-behavioural therapy for anxiety disorders: mastering clinical challenges. New York: Guilford Press.

A wise and excellent formulation-based approach to treating anxiety disorders which remains evidence-based in its recommendations for intervention. It is extremely readable yet avoids oversimplifying the management of anxiety disorders.