From: Haddock, G. & Slade, P.D. (Eds.) Cognitive-Behavioural interventions with Psychotic disorders, London: Routledge, ps. 86-102

Chapter 5

COPING STRATEGY ENHANCEMENT FOR PERSISTENT HALLUCINATIONS AND DELUSIONS

Lawrence Yusupoff and Nicholas Tarrier

INTRODUCTION

The Coping Strategy Enhancement (CSE) approach for persistent hallucinations and delusions is characterised by its emphasis on an individually formulated training programme to enhance symptom self-management skills. In practice, however, a range of cognitive-behavioural methods are applied flexibly to match the diversity of clinical presentations of individuals who experience these psychotic phenomena. This chapter begins with a brief background to the approach and basic conceptual model, folk>wed by a summary of research findings from an initial controlled trial of CSE and a description of a second, larger scale investigation currently ongoing. Specific therapeutic methods and mechanisms are described with some illustrative case material.

BACKGROUND

Experience of persistent positive psychotic symptoms is often associated with marked psychological distress, as indexed by measures of anxiety and depression (Breier and Strauss, 1983; Tarrier, 1987; Hustig and Hafner, 1990). By implication, one might expect that the presence of such negative affective states would give rise to the development of naturalistic coping strategies amongst these individuals. Coping in this sense refers to the active self-generation of cognitive and behavioural procedures either to impact on the symptom directly or to minimise the resultant distress. A series of studies confirm that the majority of patients (range: 67-100 per cent) do indeed use active coping strategies with varying degrees of success (Falloon and Talbot, 1981; Breier and Strauss, 1983; Cohen and Berk, 1985; Tarrier, 1987; Carr, 1988).

In the Tarrier (1987) study, twenty-five patients were identified who were experiencing hallucinations and/or delusions during a nine month period after discharge, from a population of eighty-five patients with a diagnosis of schizophrenia. All patients were maintained on neuroleptic medication. Using a detailed semi-structured interview, the Antecedent and Coping Interview (ACI; Tarrier, 1992), the individual's subjective account of their psychotic experiences was elicited as were the conditions under which these were experienced and their idiosyncratic emotional reaction and subsequent use of coping strategies. Seventeen patients (68 per cent) experienced true hallucinations, two (6 per cent) experienced pseudo-hallucinations and nineteen (76 per cent) experienced coherently expressed delusions. Thirteen patients (32 per cent) were able to identify antecedents to either symptom onset or symptom exacerbation. These varied from specific stimuli (e.g. traffic noise, television; 16 per cent), social situations (20 per cent), internal states (e.g. feeling anxious; 4 per cent), being unoccupied or alone (24 per cent), or to specific times of day (16 per cent). All patients were able to identify symptom consequences, although two patients explained that they no longer experienced the severe anxiety that previously accompanied their symptoms. Seventy-two per cent reported that they experienced distress and 36 per cent that the experience of symptoms disrupted ongoing behaviour, with 12 per cent disclosing severe social changes such as arrest or attempted suicide.

Eighteen patients (72 per cent) were able to describe symptom-related coping strategies. Cognitive strategies were used by 40 per cent, which included distraction (12 per cent), attention narrowing (12 per cent) and self-instruction (16 per cent). Thirty-six per cent used behavioural strategies, 16 per cent increased or initiated social interactions, 8 per cent used an increase in solitary activity, and 12 per cent withdrew from social interaction. Sixteen per cent reported coping by increasing sensory stimulation such as turning on the radio to drown out voices. The use of relaxation or breathing exercises were used by 8 per cent to decrease levels of physiological arousal. Of the patients who used active coping strategies, 72 per cent reported at least one of them as being moderately successful or better in controlling symptoms and 75 per cent of all strategies used were rated as being at least moderately successful. Eight patients (32 per cent) reported multiple strategy use which was significantly associated with strategy effectiveness.

On the basis of these findings, Tarrier (1987) concluded that using programmes to train patients in appropriate and effective coping skills was a viable proposition, given that determinants of these symptoms could be identified. It was envisaged that such an approach would utilise and extend coping methods already established in the patient's behavioural repertoire and would thus have high face validity which would facilitate a collaborative therapeutic alliance. Furthermore, the assessment procedure itself, with its emphasis on accurate and detailed cognitive-behavioural analyses of symptoms, would contribute to the psychological reattribution process and also generate additional therapeutic interventions to modify beliefs, which would result in reduced distress. Coping strategies would be practised in vivo where possible and this would be achieved by encouraging the patient to simulate or actually activate the symptom during sessions and to practise coping strategies as homework by entering situations in which symptoms are likely to occur. Finally, training in coping would not be restricted to the application of a single technique, but would include combinations or sequences of individual strategies.

CONCEPTUAL FRAMEWORK

Basic model

The CSE approach is a pragmatic one and a theoretical position regarding symptom origins, biological or otherwise, is not crucial to effective clinical practice. The model is one of symptom maintenance. Hallucinations and delusional beliefs occur in subjective and social contexts and thus symptoms assume significance and meaning only when there is an accompanying emotional reaction (with cognitive, physiological, behavioural and affective components). A psychotic phenomenon is unlikely to develop to a clinically significant degree without the support of these psychological parameters, i.e. a (persistent) symptom cannot exist in isolation from normal evaluative and motivational processes. The emotional reaction both defines the symptom and at the same time creates conditions which result in its maintenance. For example, a powerful physiological reaction may represent confirmation of a delusional belief and the threat associated with it. Even minor changes in behaviour, such as stooping in response to unpleasant voices, may provide additional grounding of the beliefs associated with them; for example, a patient's post hoc rationalisation of his behaviour following symptom onset might include a statement such as "My body stoops, therefore the voices must be more powerful than me and dangerous."

Thus, the CSE approach relies on eliciting a detailed account of the components of the emotional reaction and their reciprocal relationships. These can then be manipulated with the patient, on a collaborative basis, by way of traditional cognitive change methods, including behavioural experiments, reality testing tasks and verbal techniques.

Where coping strategies are present an analysis of their effectiveness, consistency of use and limitations can then be conducted and a systematic training provided to be applied in such a way that is consistent with the cognitive-behavioural formulation developed jointly with the patient. These strategies might include the relabelling of the symptom as self-generated, attentional strategies, changing activity levels, or physiological self-regulation. Some naturalistic strategies may be ostensibly effective but may inadvertently result in long-term symptom maintenance. A maladaptive strategy is defined as one which is used by the individual as a protection from imminent danger or unacceptable personal risk. In other words, the symptom is resisted at all costs and therefore the patient is never in a position to assess the true risk. A patient who always shouts back at his voices, as if their accusations and directives have true potential to threaten his psychological integrity or cause real physical harm, may experience temporary relief by counter-attacking in this way, especially if his response is associated with the symptom stopping. The therapeutic intervention might involve encouragement and training of non-responding to facilitate reality testing, an approach which parallels cognitive-behavioural therapy for obsessive-compulsive patients (Salkovskis, 1989). Other maladaptive strategies which would be targeted for change would include substance abuse and medication misuse to control symptoms.

Symptom content

The basic model described above provides a framework for clinical intervention, but does not address the specificity of individuals' experiences. There are essentially two positions with regard to symptom content relevance. First, one might argue that psychotic experiences are by definition discontinuous with reality agreed by consensus (unlike the neurotic disorders which are merely exaggerated normal responses) and therefore the content of hallucinations and delusions are not amenable to meaningful psychological analysis. Second, symptom content is not arbitrary and may represent personally significant material. Thus, taking the first position, it does not matter that an individual is deluded about his brother organising a plot to kill him versus some impersonal secret organisation trying to do the same; both beliefs would be labelled as a persecutory delusion. Alternatively, one might hypothesise that the relationship with his brother may well be pertinent. The CSE approach does not necessarily argue for the primacy of psychological events as ultimately causing psychoses, but that there may well be therapeutic advantages for acknowledging the potential meaning of symptoms. Clinical experience suggests that therapeutic change is indeed facilitated by acknowledging the psychological relevance of psychotic symptoms by linking them with key aspects of the patient's historical narrative. The more powerful the psychological formulation for the patient, the greater the likelihood of therapeutic compliance; the degree of conviction with which a patient may hold a delusional belief may be such that shifting away from his view of reality will be too threatening unless legitimised in some way. Kingdon and Turkington (see Chapter 6) provide such a legitimisation by using a normalising rationale, and Haddock, Bentall and Slade (see Chapter 3) offer hallucinating patients an opportunity to compare the content of their voices with the content of their own thought processes/current concerns to reconsider their misattributions. Romme and Escher (see Chapter 8) also detail and differentiate between different types of meaning that the voice content may represent.

The CSE approach to symptom content is perhaps best illustrated by a case example. A patient, in his late fifties, frequently experienced the belief that strangers were talking about him. He believed that they were aware of his schizophrenia diagnosis and in particular that he was a 'dangerous schizophrenic' and that they were likely to persecute him because of this. During the first cognitive-behavioural interview the dangerousness issue was examined further and cognitively he disclosed a strong association with a past event; the key historical link was his account of the onset of his psychosis thirty years previously, when in the Navy. He described having violently assaulted a colleague because he thought the man had hinted that he had had an affair with the patient's girlfriend; the seriousness of the offence was such that he was subsequently admitted for psychiatric care and discharged from the Navy. The dangerousness had continued to be an important issue for him, as he had always remained unclear about his potential to lose control in such a way again. Going back even further he recalled the terror of living with a violent father who would frequently accuse his mother of having affairs and would attack her physically. These links facilitated rapid clinical changes thereafter such that the patient was more than willing to engage in reality testing tasks to establish to what extent his interpretation of interpersonal events may have been influenced by memories of past events being reactivated. Another symptom for this patient was the daily experience of non-verbal Morse code hallucinations. His associations were again illuminating; he had assumed that his successful brother who had remained in the Navy was sending coded messages to undermine him. Towards the end of therapy, he disclosed that he had always resented his brother for abandoning the family and escaping to the Navy. He also accepted that the Morse was a daily reminder of his regret and sadness of a lost career.

Motivation

The third element of the current CSE model is that of motivation and its role in modulating therapeutic engagement and outcome. Motivation, in this case, does not refer to the amotivational state associated with the negative symptoms of schizophrenia, but rather whether there is sufficient incentive to achieve therapeutic goals. Motivational considerations are pertinent to therapeutic interventions for any clinical problem, although there may be specific issues with regard to the loss or change of persistent hallucinations and delusions.

The experience of distress in itself may not be a sufficient condition for therapeutic engagement. There may be both perceived advantages and disadvantages associated with psychotic phenomena, which will be reflected in individuals' mixed attitudes. Miller et al. (1993) investigated the relationship between the attitudes of fifty psychiatric inpatients towards their hallucinations and the outcome at discharge. The majority of subjects (98 per cent) confirmed that their hallucinations were associated with undesirable effects, but 52 per cent reported that their hallucinations were beneficial in some way. Advantages included feelings of relaxation, a sense of companionship, protective functions and self-concept enhancement. Inpatient treatment involved a combination of pharmacotherapy and a range of psychosocial interventions. Patients who had disclosed some symptom benefit were significantly more likely to continue hallucinating post-treatment. The implication of the study was that symptoms persisted because they were valued, but as the authors themselves note, positive attitudes towards symptoms may represent an adaptive response, rationalising the experience of chronic therapy resistant phenomena. The CSE approach supports the view that attitudes to symptoms should be clarified at the outset of therapy where possible and addressed where appropriate.

A separate, but related, issue involves patients' expectations regarding the potential loss or change of symptoms. There may be feared practical consequences, for example the loss of social security payments, which if unaddressed may compromise outcome. A change of symptom status may be associated with an alteration of role; an example here would be the individual who believed that the supportive network provided by psychiatric services would be available only to patients who were clearly symptomatic. Also, the chronicity of some individuals' persistent symptoms may be such that a life without them cannot be imagined; assisting the patient to consider and rehearse the potential positive effects of a change in symptoms may be useful at a preliminary stage, to generate sufficient incentive to remain engaged in the therapeutic process.

The motivational factors described above may be readily accessible by the use of appropriate questioning. It is, however, also possible that symptom change may be associated with consequences that are not immediately apparent, but operate at an implicational level to undermine good outcome. Cognitive-behavioural methods are potentially powerful and rapid clinical change may be seen after a short number of sessions, even with psychotic patients (Tarrier et al., 1990). The potentially sudden realisation that one's experiences may not have a basis in reality may have deleterious effects such as the patient disempowering the therapist (and his cognitive-behavioural logic) by including him in the delusional system, especially where the implications of symptom loss are threatening. Psychological reattribution, for example, may result in individuals becoming acutely aware of their 'lost years' spent as a psychiatric patient, with regrets about disrupted career and relationship goals. The shift in locus of responsibility for past events might also result in the experience of guilt feelings for actions/behaviours which were consistent with the delusional material at the time, but not acceptable in retrospect. What may be required is a relegitimisation of the individual's psychotic experiences, by enabling the patient to appreciate that he or she was not in a position to choose their earlier history and other factors which may have contributed to the development and maintenance of the psychosis.

Perhaps the clearest example of disadvantages associated with symptom loss is with individuals whose symptoms are enmeshed with beliefs about the self. Religious and grandiose delusions potentially fall in this category. There is some evidence that individuals with chronic, systematised delusions of this type have higher levels of perceived purpose and meaning in life and lower levels of depression and suicidal ideas compared to individuals who have 'recovered' from their delusional beliefs (Roberts, 1991 ). This would suggest that it may be more appropriate to assist these patients by working within the delusional framework to enhance coping strategy effectiveness, rather than attempting to change the delusional beliefs themselves.

Although some of the notions described above regarding motivation are speculative, and based mainly on clinical experience, the application of the CSE approach may require sensitive modifications during therapy to maximise good outcomes.

CSE STUDY 1

The first controlled trial was conducted with patients with a diagnosis of schizophrenia, living in the community, who continued to experience hallucinations and delusions despite anti-psychotic medication (Tarrier et al., 1993a). The CSE approach was compared with a problem solving (PS) therapy which was selected as a control treatment since this is an established cognitive-behavioural treatment method (D'Zurilla and Goldfried, 1971; Hawton and Kirk, 1989) with wide applicability and was considered to be a credible control treatment with potential benefits in terms of increased social functioning. It was predicted that a reduction in hallucinations and delusions would be obtained only amongst those patients allocated to the CSE condition, and that these patients would also improve in their social functioning, as a by-product of decreased impairment resulting from the experience of the positive psychotic symptoms. Forty-nine suitable referrals were received for the project, but the attrition rate was high; 45 per cent either refused to participate or dropped out prior to the start of treatment. Twenty-seven continued through treatment and twenty-three were assessed at six month follow-up. Methodologically, the difficulties in recruiting and retaining subjects may have biased outcomes in favour of positive results, given the potentially self-selected population of treatment acceptors. Given the heterogeneity of individuals who receive a diagnosis of schizophrenia, much stands to be gained from careful descriptions of populations, methods of recruitment and details of attrition amongst published studies of cognitive-behavioural treatments for hallucinations and delusions.

Subjects were randomly allocated to one of the two treatment conditions, both of which involved attending for ten one hour sessions. Approximately half of the patients were assigned to a waiting list control group. Overall, the results demonstrated that patients in both treatment conditions obtained significant improvements in positive psychotic symptoms compared to subjects in the waiting list control group, who did not demonstrate any change in symptom severity scores. Comparing the results of the CSE and PS therapy was less clear-cut. Significant improvements were obtained from inclusion in either condition and there was some evidence that CSE was superior in terms of reducing delusional beliefs. Treatment gains were maintained at six month follow-up. To establish whether improvements were at a clinically significant level, a criterion of a 50 per cent or greater decrease in hallucinations and/or delusions was used. Sixty per cent of the CSE patients and 25 per cent of those who received PS achieved this criterion at post-treatment assessment; the difference between the groups approached significance. At six months, these figures fell to 42 per cent and 36 per cent respectively. Overall, at this stage, five patients (22 per cent of the follow-up sample) were completely free of positive psychotic symptoms. Significant changes were not noted for negative symptoms and social functioning in either of the conditions.

Pre- and post-treatment coping and problem solving skills were also assessed; these data were rated blind to both treatment condition and time of assessment (Tarrier et al., 1993b ). Multiple strategy use was associated with increased efficacy, as had been found previously. There appeared to be a linear relationship between the number of coping strategies used and the rated efficacy of each strategy .Subjects who received CSE significantly improved their coping skills and improvements in coping were significantly related to decreases in hallucinations and delusions. Conversely, patients who received PS did not show improvements in coping skills. All patients improved their problem solving skills, whether they received CSE or PS. Improvements in problem solving skills, unlike improvements in coping skills, were not related to symptomatic improvement.

CSE STUDY 2

The CSE approach is currently being investigated further, in a second, larger scale treatment trial funded by the Wellcome Trust over five years. The first study demonstrated that the severity of positive psychotic symptoms could be reduced by a clinically significant margin with ten sessions of therapy and that these changes were maintained at six months. Using coping skills as a process measure was suggestive of the therapeutic specificity of the clinical procedures employed, however, the unexpected positive outcomes obtained with PS highlighted the possibility that changes were a function of non-specific factors or at least hitherto unspecified mechanisms. An extended package has been developed for the present study, which combines both the coping strategy and problem solving procedures and formal relapse prevention training has also been introduced. Subjects are seen for a total of twenty sessions, over a ten week period, followed by four, monthly booster sessions. This cognitive-behavioural treatment is being compared to a supportive counselling condition (thus controlling for therapist contact factors) and a routine psychiatric care condition. Individuals are followed up at one and two years and the impact of the interventions on the long-term course of psychotic illness, including relapse rates, can thus be assessed. Subjects are now also independently assessed, the rater being blind to allocation, which was not the case in the first study. The random allocation process includes stratification for gender and symptom severity scores, since a more favourable prognosis has been noted for women with a diagnosis of schizophrenia (Shepherd et at., 1989) and symptomatic improvement following therapy may be a function of higher initial severity scores (Tarrier et at., 1993a).

In the first study, symptomatic subjects were recruited from a community population who had been free of an acute episode for at least six months; this is also the case with the current study, but recruitment is now being extended to acute episode patients, who have been stabilised on medication and are due for discharge, but continue to experience hallucinations and/or delusions. Favourable outcomes for this latter group would support the future provision of psychological intervention, on a routine basis, to those whose recovery from an acute episode is incomplete. Improvements in social functioning were not noted in the first study; the brevity of the interventions and the abrupt cessation of therapy after ten sessions may have accounted for this. The assessment procedure in the present investigation includes the selection of six deficits perceived by the patient as personally significant; these are operationally defined (for a range of good and poor outcomes) and three are selected randomly as goals to be addressed with the therapist during the problem solving component of treatment. Outcomes will be compared to the three functioning deficits not selected as goals, thus assessing for treatment generalisation.

A wide range of pre-treatment measures have been included to access the potentially diverse factors associated with differential therapy response and drop-out. These include illness attributions, cognitive deficit measures, current and pre-morbid intellect estimates, expectation of the positive and negative impact of symptom loss, suspiciousness, affect, hopelessness, negative symptomatology, sociodemographic and other illness variables. Subjective reasons for non-continuation of therapy will be assessed by means of a drop-out questionnaire.

SPECIFIC THERAPEUTIC METHODS

The following account represents a selection of therapeutic methods as opposed to a comprehensive treatment protocol.

The cognitive-behavioural interview and engagement

Prior to cognitive-behavioural interviewing, a systematic assessment of all positive psychotic symptoms is desirable, using the Present State Examination (Wing et al., 1974), or a similarly reliable equivalent. Establishing the presence of other non-psychotic symptoms is also recommended; the presence of these may well affect clinical decision making. A recent case may highlight this point. A man with a lengthy history of persistent auditory hallucinations and thought echo, also described frequent episodes of distressing, vivid, intrusive imagery , the contents of which appeared to support the nature and content of his voices. This phenomenon was given clinical priority and successful intervention secondarily resulted in a reduction in the frequency and severity of the voices, without directly addressing these. Cognitive-behavioural analyses of each psychotic symptom then follows and for this purpose the Antecedent and Coping Interview (Tarrier, 1992) is used as the basic format to determine symptom contexts, emotional reactions, symptom-related beliefs and meanings as well as adaptive and maladaptive coping.

Usually, a disproportionate amount of time is spent obtaining a detailed phenomenological account of each symptom. Even at this stage, where sufficient rapport between therapist and patient has been established, information gathering is a highly interactive process, such that statements made by the individual about their beliefs or experiences are not taken at face value and recorded, but are enquired about beyond what is usually available to clinicians and perhaps the patient himself. For example, where the patient describes a plot against him, the therapist may ask about the organisation's motivations, the reasons why the persecutory agents have failed to date, their source of funding, power hierarchies within the organisation, mechanisms by which unusual skills are achieved by the group, etc. This process can be conducted respectfully, and is designed to disrupt the patient's habitual processing of the delusional material, to determine the rules which govern what is possible and not possible in the patient's alternative reality and to access the patient's 'free behaviours' (Brehm, 1976) which might be of use therapeutically. Asking the patient to come up with new material in such a way, thus extending the number of descriptive statements, is potentially useful later on in therapy when reviewing the evidence for and against beliefs, and eventually highlighting inconsistencies within the account (Lowe and Chadwick, 1990).

In order to elicit the emotional reactions which accompany symptoms, the patient is asked to generate the phenomenon in vivo, and where this is not possible, the therapist may use guided imagery , symbolically reactivating a recent symptom occurrence. Under such circumstances, 'hot cognitions' are more readily available and also the therapist is in a position to 'track' the patient's affective state, observable physiological changes and minor behavioural and postural alterations. At this stage in the interviewing, initial hypotheses about the personal significance and meanings associated with symptom content might guide subsequent questioning.

There may well be good opportunities in the initial interview(s) to conduct brief behavioural experiments as part of the engagement process. This would involve the manipulation of any or all aspects of the emotional reaction in relation to the experience of the symptom (see case study at the end of this chapter).

At some point during initial sessions symptom attitudes and expectations regarding symptom change or loss should be established and addressed. It may also be pertinent to enquire about the patient's perceptions of the likely reaction of significant others in the event of a change in symptoms and the implications for therapy.

Treatment methods

These vary considerably and are dependent on whether the phenomena are delusional, hallucinatory (or both) and whether the therapeutic goal is purely to reduce distress or to change fundamental aspects of the symptom and its maintaining factors. There may well be a range of delusional beliefs and hallucinations in a number of modalities, but target symptom selection can usually be achieved relatively easily by establishing the degree of distress experienced, the discreteness of occurrence and a jointly agreed account about the relationship ( or non-relationship) between symptoms. A coping strategy or other intervention is selected initially on the basis that there is a high probability of demonstrating, even temporarily, a successful reduction in distress.

The intervention, whether this involves coping strategy training, manipulation of the emotional reaction or response prevention is conducted under salient training conditions. This might involve requesting that symptoms are elicited by the patient during sessions, especially where auditory hallucinations are experienced (Fowler and Morley, 1989). An alternative method involves symptom simulations; for example, once an account is obtained of the typical content of voices and their physical characteristics these can be recorded onto a cassette tape (by either patient or therapist or both) and played back for a more detailed assessment or to experiment with and rehearse alternative responses. The benefit of this method is that different parameters can be manipulated (such as voice location in three-dimensional space and loudness) as part of a graded approach with regard to task difficulty, as the patient progresses. Additionally, the role-playing of symptoms, for example, the therapist taking the part of typical auditory hallucinations or presenting a summary statement reflecting delusional beliefs ( and related cognitions), is a method which can be applied flexibly with roles reversed as necessary .Often, even under such simulated conditions, emotional reactions may be powerful and the modelling of alternative responses may be insufficient to disrupt the automaticity of the patient's response. One solution would be to request that the patient exaggerates his response in a dramatic fashion, a technique adapted from Gestalt Therapy (Van De Riet et al., 1985) and repeating this several times, to bring the reaction under conscious control which may temporarily change the meanings ascribed to symptoms, thus allowing further training to take place.

The development of a shared psychological formulation is a key aspect of this approach. The degree of complexity, the timing with regard to communicating this to the patient, or whether the account is developed jointly will also vary depending on the nature of the case. In general, a formulation might include an individualised cognitive-behavioural model of maintenance and educational aspects regarding psychological processes such as the 'confirmatory bias' (Chadwick, 1992). A further component might be to offer symptom explanations consistent with the patient's historical narrative, such as previous relationship conflicts, psychological traumas, bereavement and significant life events around the time of the first onset of psychosis. Symptom maintenance can also be defined in terms of what the symptom may represent with regard to the future functioning and quality of life of the individual compared to prior life goals; the symptom may be associated with fleeting imagery of the individual finally being overcome by the symptom some time in the future, and, by implication, life having become meaningless.

CASE STUDY

Mrs M, a 48-year-old married woman, was a subject in the first CSE investigation. She was referred to the project nine months following her first psychotic episode. Her core delusional belief was that a malevolent sex machine had invaded her abdomen and was forcing her to have sexual intercourse with it. She was convinced that other individuals were similarly afflicted, but did not admit to the experience. During her previous inpatient admission a high dose of neuroleptic was used, but had little impact on her delusional beliefs; she was discharged back to the family home and continued to take anti-psychotic medication.

Independent pre-treatment assessment revealed the presence of five psychotic symptoms, which included a delusion of thoughts being read, delusional misinterpretation, delusion of grandiose ability, delusional explanation, and the sexual delusion. Thematically, all the symptoms were related. She scored maximally on measures of delusional conviction, preoccupation and interference. She was also assessed with regard to her cognitive adaptability in relation to her abnormal beliefs; this test is known as the Reaction to Hypothetical Contradiction (Brett-Jones et al., 1987); this involves presenting the subject with a hypothetical scenario which would represent powerful evidence against the abnormal beliefs being true, and in Mrs M's case she indicated that she would reject the evidence and would continue to be thoroughly convinced of her material.

Despite her initial reluctance to pursue therapy, given the degree of her distress and her severely curtailed functioning, she agreed to attend. The first session was spent obtaining a more detailed account of her symptoms. She believed that a group of individuals, several miles away, operated a computer to control powerful vibrations produced by the sex machine and that they had designed the apparatus such that she had been wired up to a video which recorded her responses. She explained that communication between the group and herself occurred telepathically, by thought waves, and that they were able to read her thoughts at any time. She was similarly aware of their intentions, although she was not able to hear what was being said. When asked about the individuals' motivation she said that four of them were annoyed with her, because of her refusal to have sex with them, but one of the group tended to support her. At this point in the interview, some of her associations were elicited. She admitted that she had always wondered what sex would be like with other men as she had only been intimate with her husband. A historical link was established fairly quickly, she recalled that her parents divorced when she was 15 years of age and she described having been distressed at the time and was aware that both her parents were having affairs.

Physically, she described the sex machine as 'something like a magnet' but that at other times it took the form of a penis. More often than not, however, it took the form of sensations, rather than an object. When the machine was active, she experienced somatic hallucinations in her arms, abdomen, vagina and lower back. She explained that actual intercourse with the machine was less common now, but that she continued to spontaneously climax, frequently during the day (up to fifteen times per day), whenever the 'vibrations' were present. She found this aspect particularly embarrassing, from a social point of view, although she did admit to enjoying the experience. Mrs M pointed to her abdomen at one stage during the interview to indicate when the vibrations had started and these were indeed clearly visible. She reported that these occurred with very high frequency, every few minutes, throughout the day, although she admitted to some variability in intensity (and preoccupation) when she was able to keep herself busy.

In terms of her emotional reaction, a good account was obtained during the session as she was clearly experiencing her symptoms at the time; she was agitated, being unable to sit down for more than a few minutes and experienced a range of autonomic symptoms of anxiety as well as marked muscular tension. When she experienced the involuntary movements of her abdomen, she either crossed her arms over her abdomen or attempted to 'force the machine out' or at least suppress the sensations by voluntarily tensing her abdominal muscles. During such an episode, the other delusional material was also activated but her attention was largely taken up by her sensations. She believed that she had no control over the experience and 'wished it would go away'.

In terms of her coping strategies, she described a partially effective strategy for her delusion of thoughts being read, for which she used attention switching 50 per cent of the time, although her strategies for the sexual vibrations were largely non-effective.

The long-term consequences and impact of symptoms on her life included increased sexual contact with her husband (because of the general increase in her sexual arousal), significant social anxiety, especially with groups of men, such that she was unable to pursue her usual leisure activities out of the home and most of her household chores could not be completed. She had been unable to return to work and she was realistically concerned about being made redundant.

The misattribution of somatic sensations, i.e. 'the vibrations', suggested that Mrs M would require a powerful face valid alternative explanation and that this might be available from her medical history .She recalled her 'sterilisation' following a biopsy at the age of 37 or 38; it was noted that she used the term sterilisation as opposed to 'hysterectomy' and she confirmed that she was unhappy with her loss. She received a course of hormone replacement tablets at the time. She was also diagnosed as having arthritis of the lower back, the first symptoms being noted approximately eight years prior to therapy. She was also asked whether she had a history of bowel problems. She confirmed this and described a series of investigations for rectal bleeding. She also reported that she had been advised to start a high fibre diet approximately six years ago, but had stopped this two to three years ago. She currently complained of diarrhoea alternating with normal bowel habit and she was asked whether she had every received a diagnosis of ‘irritable bowel syndrome' and she recalled that the term had indeed been used. Her medical notes were subsequently checked to confirm her account.

Towards the end of the first session a number of simple trial interventions were attempted. She was asked, non-confrontationally, whether her vibrations could possibly be misinterpreted symptoms of irritable bowel syndrome. She was dubious of this explanation, but agreed that it would be worthwhile, as an experiment, to return to a high fibre diet. Also, a brief behavioural experiment was conducted by asking her to reverse her protective behaviours (i.e. crossing her arms over her abdomen and the voluntary tension other abdominal muscles) as well as the manipulation of her attention from her abdomen to external stimuli. Ratings of preoccupation and distress were obtained during this experiment and when directed to behave as she normally did in response to the vibrations. Significant reductions in the levels of preoccupation and distress were demonstrated and she expressed relief that some control was available to her. She agreed to try the strategies as homework prior to the next session and symptom self-monitoring was also started.

At the second session, she indicated that her ability to cope with her experiences had been enhanced by reversing her habitual responses and 'keeping busy'. She had been compliant with her high fibre diet and experienced less abdominal discomfort. By session three, she reported continued improvement and had noted that the vibrations were less frequent. There had been an associated reduction in preoccupation, and she had stopped seeking reassurance from her family about the delusion, at the therapist's request. She remained deluded, but her conviction rating had reduced such that she was beginning to accept the possibility of an alternative explanation. A full psychological formulation was introduced at this point incorporating notions of misattribution of somatic sensations, a physiological arousal-attentional vicious cycle mechanism and the role of behavioural maintaining factors such as her protective behaviours and reassurance seeking. The sexual content of the delusions were explained in terms of the historical context of her parents' extra-marital affairs coupled with aspects of her own sexual relationship with her husband, prior to the onset of her psychosis. Some explanation was also offered for the timing of her first onset of psychosis and included a number of key stressors, such as the death of her mother two years prior to onset, the start of heavy alcohol use associated with the bereavement and also a relationship 'crisis' with her own daughters around this period. Symptoms were also legitimised by engaging the patient in a discussion about the 'complexity of the human mind' and how unusual phenomena could be accounted for in psychological terms, such as vivid dreams, visual illusions and hypnotically induced analgesia. A significant shift in her conviction rating was noted such that her symptoms were now a 'mystery' to her and this change was consolidated by further verbal cognitive therapy techniques. She continued to experience some muscular tension and autonomic symptoms of anxiety when sensations were present and these were addressed by introducing formal progressive muscle relaxation training. As therapy progressed, further reductions in conviction and preoccupation were noted and a series of behavioural targets were introduced to improve her ability to complete daily chores and reinstigate her leisure pursuits which she had dropped because of her abnormal beliefs. She also received training to employ covert self-instructional statements to activate her coping behaviours whenever the sensations and/or beliefs were experienced.

At session ten, a belief conviction rating of 5 per cent was noted for the sexual delusion and her other related delusions had dissipated completely. At the start of therapy, her self-monitoring data suggested that she had experienced vibrations for several hours per day and by the tenth session this had reduced to approximately forty-five minutes per day. Her husband confirmed that she had improved significantly, in terms of her level of agitation and daily functioning, but that she was still relatively amotivated compared to her pre-morbid state.

At six month follow-up, Mrs M was asymptomatic with regard to her positive symptoms. She declined an offer of a formal interview at thirty month follow-up because she wanted to minimise her contact with services, but agreed to a brief assessment over the telephone. She reported that there had not been a recurrence of her sexual (or other) delusions and that she had continued to pursue an active social life. She admitted to the experience of occasional auditory hallucinations (which had been intermittently present in the past), which she described as a faint voice, approximately once or twice per month for a few seconds, to which she paid little attention.

REFERENCES

Birchwood, M. and Tarrier, N. (eds) (1992) Innovations in the Psychological Management of Schizophrenia, Chichester: Wiley.

Brehm, S. S. (1976) The Application of Social Psychology to Clinical Practice, New York: Wiley.

Breier, A. and Strauss, I. S. (1983) 'Self-control in psychotic disorders', Archives of General Psychiatry 40: 1141-5.

Brett-Jones, I., Garety, P. A. and Remsley, D. R. (1987) 'Measuring delusional experiences: A method and its application', British Journal of Clinical Psychology 26: 257-65.

Carr, V. (1988) 'Patients' techniques for coping with schizophrenia: An exploratory study', British Journal of Medical Psychology 61: 339-52.

Chadwick, P. (1992) Borderline: A Psychological Study of Paranoia and Delusional Thinking, London: Routledge.

Cohen, C. I. and Berk, B. S. (1985) 'Personal coping styles of schizophrenic out-patients', Hospital and Community Medicine 36: 407-10.

D'Zurilla, T. I. and Goldfried, M. (1971) 'Problem solving and behaviour modification', Journal of Abnormal Psychology 78: 107-26.

Falloon, I. R. R. and Talbot, R. E. (1981) 'Persistent auditory hallucinations: Coping mechanisms and implications for management', Psychological Medicine 11: 329-39.

Fowler, D. and Morley, S. (1989) 'The cognitive-behavioural treatment of hallucinations and delusions: A preliminary study', Behavioural Psychotherapy 17: 267-82.

Rawton, K. and Kirk, I. (1989) 'Problem solving', in K. Rawton, P. M. Salkovskis, J. Kirk, and D. Clark (eds) Cognitive Behaviour Therapy for Psychiatric Problems, Oxford: Oxford Medical Publications.

Hawton, K., Salkovskis, P. M., Kirk, J. and Clark, D. (eds) (1989) Cognitive Behaviour Therapy for Psychiatric Problems, Oxford: Oxford Medical Publications.

Hustig, H. H. and Hafner, J. (1990) 'Persistent auditory hallucinations and their relationship to delusions and mood', The Journal of Nervous and Mental Disease 178: 264-7.

Lowe, C. F. and Chadwick, P. D. J. (1990) 'Verbal control of delusions', Behaviour Therapy, 21: 461-79.

Miller, L. J., O'Connor, E. and DiPasquale, T. (1993) 'Patients' attitudes toward hallucinations', American Journal of Psychiatry 150: 584-8.

Roberts, G. (1991) 'Delusional belief systems and meaning in life: A preferred reality?', British Journal of Psychiatry 159 (supplement 14): 19-28.

Salkovskis, P. M. (1989) 'Obsessions and compulsions', in J. Scott, J. M. G. Williams and A. T. Beck ( eds ) Cognitive Therapy in Clinical Practice: An Illustrative Casebook, London: Routledge.

Scott, J., Williams, J. M. G. and Beck, A. T. (eds) (1989) Cognitive Therapy in Clinical Practice: An Illustrative Casebook, London: Routledge.

Shepherd, M., Watt, D., Falloon, I. and Smeeton, N. (1989) 'The natural history of schizophrenia: A five-year follow-up study of outcome and prediction in a representative sample of schizophrenics', Psychological Medicine, Monograph Supplement 15: 1-46.

Tarrier, N. (1987) 'An investigation of residual psychotic symptoms in discharged schizophrenic patients', British Journal of Clinical Psychology 26: 141-3.

-(1992) 'Management and modification of residual psychotic symptoms', in M. Birchwood and N. Tarrier ( eds ) Innovations in the Psychological Management of Schizophrenia, Chichester: Wiley.

Tarrier, N., Harwood, S., Yusupoff, L., Beckett, R. and Baker, A. (1990) 'Coping strategy enhancement (CSE): A method of treating residual schizophrenic symptoms', Behavioural Psychotherapy 18: 283-93.

Tarrier, N., Beckett, R., Harwood, S., Baker, A., Yusupoff, L. and Ugarteburu, I. (1993a) , A trial of two cognitive-behavioural methods of treating drug- resistant residual psychotic symptoms in schizophrenic patients: I. Outcome', British Journal of Psychiatry 162: 524-32.

Tarrier, N., Sharpe, L., Beckett, R., Harwood, S., Baker, A. and Yusupoff, L. (1993b) , A trial of two cognitive-behavioural methods of treating drug-resistant residual psychotic symptoms in schizophrenic patients: II. Treatment specific changes in coping and problem-solving skills', Social Psychiatry & Psychiatric Epidemiology 28: 5-10.

Van De Riet, V., Korb, M. P. and Gorrel, J. J. (1985) Gestalt Therapy: An Introduction, New York: Pergamon Press.

Wing, J. K., Cooper, J. E. and Sartorius, N. (1974) Measurement and Classification of Psychiatric Symptoms: An Instruction Manual for the PSE and Catego Programme, Cambridge: Cambridge University Press.