What happens in too closed groups. The experience of Therapeutic Communities

Anthony Bateman *

What happens in too closed groups. The experience of Therapeutic Communities.

Lucca; paper presented at the meeting about "Affiliation and detachment from sectariam Communities", 7-8 october 1999.

The term borderline personality disorder (BPD) has emerged from a confluence of psychiatric and psychoanalytic research. From a psychiatric point of view the DSM-IV descriptive criteria for a diagnosis of BPD can be summarised as ‘stable instability’comprising: intense but unstable interpersonal relationships; self-destructiveness; constant efforts to avoid real or imagined abandonment; chronic dysphoria such as anger, boredom or depression; transient psychotic episodes or cognitive distortions; impulsivity; poor social adaptation; and identity disturbance. Not surprisingly such patients present frequently to psychiatrists representing 11% of all out-patients and 19% of inpatients. Medication alone is an inadequate treatment (Soloff 1998) and frequently psychiatrists refer borderline patients to psychotherapists. Borderline patients are a heterogeneous group and vary in the severity of their psychiatric symptoms, their suicide risk, propensity to self-harm, and in their degree of chaotic and impulsive behaviour. Thus classical outpatient psychotherapy may be inadequate and psychotherapeutic treatment may need to take place in an in-patient (Norton and Hinshelwood 1996) or a day patient setting and be followed by out-patient and community treatment.

In this talk I shall describe a psychoanalytically orientated day hospital programme for patients with severe borderline personality disorder whose difficulties make them impossible to treat within a ‘classical’ psychoanalytic model. The day hospital programme is organised around group psychotherapy, takes into account both conflict and deficit models of BPD, and emphasises staff discussion of countertransference response. The programme has been evaluated in a randomised control trial. First I shall outline the results of the randomised controlled trial evaluating outcome. Second I will discuss why some of the patients in the trial may have done less well than others. In essence this was because a subgroup of patients formed the core of an anti-group both within a group and outside the group itself. They became a closed group, unreachable by staff and patients and so did less well over a treatment period of 18 months.

Treatment Outcome

Full details of the design of the study and outcome of treatment may be found elsewhere (Bateman and Fonagy 1999) and only an outline will be given here. In a randomised controlled design borderline patients were assigned either to intensive treatment in the day hospital programme or to treatment as usual in the general psychiatric service for a period of 18-months. Treatment as usual in the general psychiatric service was chosen as the approach to control for spontaneous remission. Patients were able to participate in normal psychiatric treatment seeing their doctors and nurses when necessary and receiving general psychiatric day hospital treatment (72%), outpatient and community follow-up (100%), and in-patient admission (90%) as appropriate. The general psychiatric day hospital treatment involved an occupational therapy programme incorporating art therapy and music therapy but none of the control group received any formal psychotherapy. While the control group cannot be considered to have received comparable amount of professional attention to the day hospital group, the control is valuable from the point of view of the effects of medication as well as spontaneous changes in mental state.

Patients entering the study were assessed at entry with a structured interview. The SCID-1 and SCID-2 were used to assess psychiatric diagnosis, the Diagnostic Interview for Borderlines (DIB) (Gunderson et al 1981) to evaluate borderline function, and a suicide and self-harm inventory administered. A full clinical history, including demographic information, was taken. Patients subjective experience of symptoms was measured using the SCL-90-R . Depression and anxiety were measured using the Beck Depression Inventory (BDI) and the Spielberger State/trait anxiety scale [Spielberger 1962] respectively. In order to ensure assessment of areas targeted by psychoanalytic therapy, social adjustment and interpersonal function were measured pre- and post-trial using the modified Social Adjustment Scale-self-report (SAS-M) and the Inventory of Interpersonal problems- Circumflex version (IIP) . These provide an assessment of an individual's work, spare time activities, and family life and difficulties with interpersonal function. The reliability and validity of all these instruments is well established.

Monitoring of symptoms during treatment was with self-rating questionnaires at 3 monthly intervals on all symptom measures except the SCL-90-R that was given at 6 monthly intervals.

Clinical measures included number of hospital admissions and length of stay and number of suicide attempts and acts of self-mutilation. For all patients, a search of the hospital in-patient database was made to obtain the number of hospital admissions and the length of stay during a period of 6 months before entry into the study. This was cross-checked with the medical notes. Hospital admission and length of stay and psychiatric day hospital programme attendance’s were monitored throughout the study for all patients.

The criteria for suicidal acts were: 1) deliberate; 2) life threatening; 3) had resulted in medical intervention, and; 4) medical assessment was consistent with a suicide attempt. Criteria for acts of self-mutilation were: 1) deliberate; 2) resulting in visible tissue damage; 3) nursing or medical intervention required. A semi-structured interview (Suicide and Self-harm Inventory) was used to obtain details of both suicidal and self-damaging acts for the 6 month period before patients entered the study. This interview asks specific questions not only about numbers of acts but also about dangerousness of acts, i.e. presence or absence of another person, likelihood of being found, preparation, and lethality. Multiple acts over a short period of time, for example a frenzied self-cutting, were counted as a single act. Day hospital patients were monitored carefully with regard to self-destructive acts and control patients were interviewed every 6 months. Self-reports of suicidal and self-mutilatory acts were cross-checked with medical and psychiatric notes.

19 patients with borderline personality disorder were treated with a psychoanalytically orientated day hospital programme and compared with 19 patients treated with standard psychiatric care. In contrast to patients treated within the general psychiatric service who showed little change or deterioration, patients treated with a day hospital programme for 18 months showed significant improvement on both symptomatic and clinical measures. Particularly notable was improvement in depressive symptoms, decrease in suicidal and self-mutilatory acts, reduced inpatient psychiatric days, and better social and interpersonal function which is an area specifically targeted in psychoanalytic psychotherapy. Treatment was effective for both males and females. Improvement in psychiatric symptoms and suicidal acts occurred after 6 months but a reduction in frequency of hospital admission and length of in-patient stay was only clear in the last six months indicating a need for longer term treatment. No patients committed suicide.

It has been mentioned that a major effort is made to ensure that destructive enactments within the transference/countertransference relationship are minimised. The reduction in self-damaging acts suggests that this therapeutic aim was successfully realised. Individual sessions focus on the meaning of such acts as understood within the context of the therapeutic relationship. For example some patients are more likely to harm themselves at the time of a therapists absence. This is interpreted. Further evidence for the success of containing destructive enactments comes from the fact that the programme and staff were effective in retaining patients in treatment. This is in contrast to inpatient and outpatient psychoanalytic treatments for borderline personality disorder . Only 3 out of 25 (12%) patients dropped out of the programme and could not be re-engaged. There are a number of possible reasons for this. Firstly the structured nature of the programme ensures patients are fully aware of the boundaries of treatment. No formal contract was made as experience suggests that borderline patients unwittingly sabotage their treatment. Discharge due to failure to meet stringent attendance requirements is likely traumatically to re-enact the abandonment the borderline patient is both desperate to avoid but simultaneously provokes. Secondly treatment within a day hospital programme simultaneously balances support and treatment with separation and individuation. This mirrors the central conflict of the borderline patient who eschews excessive intimacy and yet fears abandonment. The programme is neither too much nor too little. This contrasts both with inpatient treatment, which may stimulate loss of identity and terror of entrapment, and with outpatient treatment which may evoke overwhelming feelings of abandonment. Thirdly there is active pursuit of non-attendees by phone, letter, and home visit if necessary. Finally the factors leading to non-attendance are worked through within the transference-countertransference matrix of individual therapy and their meaning understood.

It is clear that the psychoanalytically orientated programme shows a treatment effect. It remains unclear which aspects of the programme are essential for progress particularly since the control group received a treatment package that included day hospital care, extensive out-patient support as well as in-patient care. However this lacked coherence and was inconsistently applied particularly at times of crisis. It is likely that a multi-component programme is necessary and that the critical feature is the way its components are brought together. Dismantling studies looking at the necessity of different aspects of a treatment package are possible and it would be of interest to determine whether both interpretative and expressive therapy are necessary or whether one component alone is adequate. But such studies are difficult and it may be more important to determine whether 18 months of treatment is required or if a shorter time is equally beneficial. In my view essential features of an effective programme include a theoretically coherent treatment approach involving a relationship focus which is consistently applied over a period of time. Under these circumstances not only is there a structured treatment plan with clear boundaries but also a focus for the mind of the patient allowing transferences to be concentrated rather than being split between different staff of a large service and between various clinics which encourages disintegration. Similarly there is a structuring of the minds of a team that can formulate the multiple problems of the borderline patient, contain countertransference responses and minimise enactment, thereby improving integration. This allows the development of a capacity to mentalise, stimulates the formation of an increasingly secure attachment, redirects aggression, and improves affect regulation.

But not all patients in treatment did equally well and so we looked at clinical factors that may account for this. We found no obvious factors such as demographic features, levels of symptoms at start of treatment, or severity of suicidal acts. So we reviewed our process notes on each individual. The patients who did less well were all in one small group. This could be a function of the group therapist compared with the other group therapists but it seems more likely to be due to the group process itself. Let me give you some clinical detail.


The group described here - a slow, open group - met in the morning on Tuesdays, Wednesdays and Thursdays. There were eight members and three group facilitators: two nurses, and a psychiatric Registrar. On Mondays and Fridays, the group merged with the other ‘small’ group to form a large group comprised of all the patients in the Day Hospital and all the staff. The key workers for the patients described here facilitated the other small group which ran simultaneously. The patients in the group described were aged between late twenties and mid-forties, and had long psychiatric histories. All but two had histories of, or still engaged in, acts of deliberate self-harm. Extremely disruptive or abusive parenting was the norm. They had all posed severe problems for previous carers. Their diagnoses were Severe Personality Disorder - Borderline type.

There were three characteristics of this group which were disruptive to the formation of a ‘work group’ (Bion, 1961), and created resistance to change. Firstly there was a behavioural and structural challenge to the group. Secondly there was a formation of a triad or closed group of 3 patients and thirdly, an unconscious control of attendances through the threat of violence. This constellation forms an anti-group but the core is the sub-group of individuals who are closed to others.

Structural Challenge

This first characteristic related to the setting out of the chairs in the morning. This job fell to JM – one of the group therapists. After JM had set the chairs out in a circle, they would invariably be rearranged by Louise into a horseshoe formation. Louise had a long history of self-injury and disruptive behaviour, such as setting fire and climbing onto precarious positions on the roof, so that her previous institution had discharged her after only three days. Louise would sit on the floor at the open end of the horseshoe in the corner of the room. As she sat she clasped to herself the pile of socks, jumpers and an old teddy-bear which she took with her everywhere. She would sit silent while rocking anxiously in what appeared to be a world of her own.

Members of the group felt rearranged by her, along with the chairs, but any challenge to Louise was met with sullen silence or a comment by her to the effect that it was unimportant, and that she had to have the chairs like that to give her some space to breathe. Quite soon the group accepted this rearrangement as a fact or characteristic of their group. Within weeks, comments by JM or other staff were felt to be calling attention to a problem which could best be dealt with by ignoring it. Our comments were then experienced as being a problem themselves and met with impatience. In this way, Louise led the group to a pathological solution which involved evacuation of the awareness of a problem by making it a fact of the group life: something that made this group what it was.

The Triad

The second significant characteristic of this group was the formation of a close relationship between three members of the group, Annette, Peter and Paul, whom I refer to as the Triad. Annette was the oldest member of the group, and had been in the Day Hospital longer than anyone else. She was a large, blowsy woman whose erratic moods, ranging from spiteful rage to manic gaiety or occasional tenderness, often dominated the atmosphere of the group very powerfully. Peter and Paul were both men in their early 30s with severe narcissistic borderline features. They both had histories of intense ambivalent relationships with women from whom they strove to separate but to whom they continually returned. Both these men attempted to gain Annette’s favours, feeling at times that their very existence depended upon it. Both Peter and Paul would complain that Annette was ignoring them in the corridor, or deliberately attacking them at some vulnerable moment. Groups were dominated by the intense rage generated by this conflict which often degenerated into insults and abuse.

In one group Peter was talking about how neglected he felt by his girlfriend, saying that as a result he felt depressed. He thought that if she wanted, Annette could make him feel better by being kind and sympathetic to him, but that she probably would not do this and he, therefore, felt there was nothing for him in the group. Annette replied with abuse, calling Peter a ‘wimp’ and saying he should ‘snap out of it’. Peter replied with bitter scorn, calling Annette an ‘old witch’ and the fight escalated. The rest of the group sat and watched. A member of staff commented that through engaging in this battle, depressing and anxious feelings were being avoided in the group. This had the temporary effect of silencing the adversaries, but within minutes the argument escalated once more.

Group Attendance

The third problem was group attendance. This was poor and there were many weeks during which only Annette, Peter, Paul and possibly Louise would attend. The group’s response to this was largely one of resignation. Attempts by staff to discern meaning in the absences in terms of, for example, rebellion or resistance, were met with indifference. Suggestions by staff that group members become active in pressurizing the absentees to come were met with lassitude and helplessness. Peter said that if the staff couldn’t get people to come, he certainly could not. Annette said that when she thought about asking others to come she felt like their mother and then couldn’t. The staff felt exasperated by this apparent refusal to take any responsibility for the group. The absences quickly became an accepted part of the nature of the group as though the group norm was of non-attendance and this characteristic was not to be challenged, just as Louise’s positioning of the chairs and the existence of the Triad were not to be mentioned.

The group was referred to by its members, with some pride, as the ‘bad group’. They compared it with the other small group which was scornfully referred to as the ‘good little school-children’s group’ which members attended regularly. ‘Nothing happened’ in the other group, it was bland, ordinary and unexciting. The problems in the ‘bad group’ were smiled about by its members and it was hinted that in some way the staff secretly enjoyed the struggle, the intransigence of its members and their ‘rebelliousness’. They felt it was more exciting, more challenging, than a group where members would predictably attend and all sit on chairs. Some of this was related in casual comments to staff outside group times, during breaks in the day. Although individuals might occasionally be challenged by other members, the group’s definition of itself remained largely that of a perverse playground of a triad with occasional visitors.

In the following vignette, Rachel, a 28-year-old woman with a history of sexual abuse who cut herself on a regular basis, had recently been discharged from inpatient status. Leonard was a 30-year-old man who attended infrequently and carried with him an atmosphere of sullen hostility which he was reluctant to relinquish. He lived in his mother’s flat following her death five years previously. For the whole of that time he had preserved his mother’s room as it was on the day of her death, and slept in her bed.

Annette, Peter and Paul were in the group as usual, with Louise sitting on the floor. Annette briefly acknowledged Rachael’s discharge from the inpatient ward the day before following an overdose. There was then some discussion of whether staff bothered to visit Day Hospital patients when they were on the wards. There was a general tense and depressed atmosphere and Annette began to describe how she hated her evening job in the local supermarket.

After 10 minutes, Leonard entered the group. He looked calm and apologized for being late. Peter enquired somewhat testily ‘Why are you here?’, a reference no doubt to Leonard’s rare attendance in the group. Leonard said he was preoccupied with thoughts of people dying and that he had had a bad dream which he’d been unable to shake off. He said ‘death is undignified’, referring to his mother’s death. Annette replied ‘death is always undignified’ and mentioned her own mother’s death from breast cancer. Leonard became upset and felt Annette was rubbishing his feelings and he said so. Peter intervened with ‘I don’t think Annette was patronizing you: you always feel you’re worse off than anyone else’. Annette agreed, she had not been patronizing Leonard, but now he had responded in this way she did feel like patronizing him. This led to a rapid increase in tension between Annette and Leonard with vicious accusations of insensitivity and eventual name-calling. Finally, Leonard picked up a chair and advanced menacingly on Annette. A member of staff shouted ‘Stop it!’ Annette leered at Leonard and said ‘Why don’t you finish the job?’ as he towered over her. The same member of staff said that this was going nowhere and Leonard put the chair down and sat down. The mood quietened.


Horwitz (1980) writes:

The central problem for borderline patients is their difficulty with vast amounts of latent destructive energy because of excessive oral frustration. Furthermore, the inability to integrate good and bad internalised objects deprives them of the capacity to neutralise their hostility.

This description recalls Kernberg’s reference to ‘unmetabolized’ internal objects. I think a solution to this ‘vast amount of latent destructive energy’ is for a group to ‘close off’, define itself as admirable and exciting and to form an anti-group. Thus, such destructive enactments as have been described were considered by its members to comprise an active and alive group and were not experienced as destructive to group work and development.

I suggest that in the group that I have discussed that, by establishing such a perverse attitude, the patients avoided the depressive anxieties associated with the group’s own destructiveness; and further, that the problem with the chairs, lack of attendance, and the triad was felt by the patients to be the successful formation of a group which did not present the members with the anxiety of preserving it. This pathological organization of the group we call the ‘anti-group’, extending Nitsun’s (1991) introduction of the term. In Nitsun’s paper the anti-group is a description of a broad set of forces which run counter to the creation and life of the group, for example, hatred, mistrust or envy of the group.

The anti-group is maintained in part by the projection of the faculties of observation and judgement onto the staff. This is evidenced in the tension between Annette and Leonard where the staff exercised a controlling function. On the whole this was felt by the patients within the group to be important in that it provided a sense of safety within which such enactments could occur. As a consequence, however, the staff were seen as strict and unbending, with little warmth or humanity. This projection allowed the destructiveness to be enjoyed without thought, because all concern for its consequences were safely lodged in the minds of the staff. But any serious attempt to invite questioning of the functioning of the Triad in the group, for example, was reacted to with hostility and treated as envious attack. Staff were told that they just wanted to ‘spoil things’ between Annette and Peter or between Annette and Paul, just when they were getting close to each other and starting to understand themselves, and sometimes they would appear to ‘work’ in the group as if to prove the point. The triad was important in fixing the group and preventing progress. Annette, Peter, and Paul met outside the group in the evenings, letting other members know but not inviting them. The three of them decided whether they were attending the group the next day or not.

The persistent absences in the group, Louise’s ruthless indifference to other members, and the abusive arguments oscillating with extreme closeness between the Triad, sometimes felt unbearable because of their endless self-obsessed and indulgent qualities. The frustration thus evoked was expressed particularly by Malcolm, a 27-year-old man from a very bleak and emotionless family, who was known for his barely-controlled rage. He would often shout at the staff telling them to do something. He said he couldn’t stand the group a moment longer; ‘you’re just like my bloody family – you’re worse than a pack of children’. Indeed, continual outbreaks of violence gave the group a nightmarish quality, something of the human dimension of an aggression that finds its limit in exhaustion, and reconciliation seemed almost entirely absent. Bion’s (1961) description of his ‘basic-assumption group’ seems to be describing something of this nature.

Sutherland (1985), in his description of the basic assumption, writes:

‘The more the individual becomes identified with a basic assumption, the more does he get a sense of security and vitality from his fusion with the group, along with the pullback to the shared illusory hopes of magical omnipotent achievement inherent in the phantasies of the assumptions. Form all these sources there is derived what Bion described as a hatred of learning, a profound resistance to staying in the struggle with the reality task until some action gives the experience of mastery of at least part of it. (p. 59).’

I suggest that the experience of ‘mastery’ over the excessive destructive energy within the group is achieved by the idealization of destructive forces to form a pathological organization or anti-group. The core of this anti-group is the triad – a closed group within the anti-group. The anti-group holds the promise of constant excitement and perverse satisfaction without the pain involved in true creativity or, in Bion’s terms, formation of the work group.

A group such as this is clearly unlikely to change by virtue of the satisfactions inherent in the anti-group. Roth (1990) writes movingly of his reactions four months into his experience of an outpatient group comprised, as this one was, solely of borderline patients:

My mood prior to the next group session seemed inexplicably to turn to despair, nearly foreboding. I wrote ‘What have I let myself in for? What am I to do?’ Group themes that once seemed fresh and exciting had become repetitive and unchanging. Affect in the group refused to diminish. Before each session I now felt anxious and powerless. I experienced the group as a bottomless well of affect. (p. 409)

This sense of unrelenting intensity was well known within the Day Hospital and the staff were familiar with the ‘holding on to your seat’ syndrome. We have already outlined the means by which the countertransference responses to this group and to the borderline patients in general were contained within the Day Hospital structure, and the constant effort required to continue to think as Bion has said, ‘under fire’. We believe the maintenance of staff supervision and support groups are essential if the staff are to avoid either being driven to despair, or into a collusion with the anti-group in its perversion of what is good and what is bad. Indeed if staff in such an environment are to be successful in their containing function, it is necessary that they be helped by the structure of the institution in which they work to remain painfully aware if the difference between good and bad and between perversion and truth. Without this the anti-group takes hold and, as we have shown in our research, this leads to poorer outcome of treatment.

HOME PAGE Go at the top of the page

A cura di Geko Sistemi