Admitting offenders with antisocial personality disorder to a medium secure unit: a qualitative examination of multidisciplinary team decision-making.

This paper reports on the results of a qualitative study funded by the Economic and Social Research Council (ESRC) looking at multidisciplinary team decisions to admit sentenced offenders with antisocial personality disorder to a medium secure unit. The aim of the study was to examine admission decision-making from a multidisciplinary perspective, and to explore the interprofessional dynamics and contextual pressures informing those decisions. The primary method of data collection was 12 semi-structured interviews with a convenience sample of various multidisciplinary staff involved in pre-admission assessment and post-assessment decision-making. Data was then coded according to the dialectic of competitive and cooperative goal seeking within groups. The findings suggest that, whilst both forms of goal seeking inform admission decisions, the presence of significant resource pressures will lead to decisional solidarity among the multidisciplinary team. When minor professional disagreements arise, they are resolved by the group leader, the Responsible Clinician, in order to maximise group productivity. It is argued that the discursive-limiting effect of resource pressures on group decision-making may weaken the morale of certain front line staff, if not undermine institutional purpose.

Introduction

Forensic medium secure units are often called upon by referrers to admit sentenced offenders with antisocial personality disorder (ASPD) in need of specialist care and treatment (Grounds, Melzer, Fryers, & Brugha, 2004a). Whilst we now know a great deal about the therapeutic challenges that such offenders can present to professionals once admitted to specialist services, we know little about the discretionary professional judgments that lead to an offer of admission.

NICE Guidelines (2010) recommend that only offenders who are ‘seeking treatment’ be considered for hospital admission; however, the motivation of the offender is likely to form only part of the complex story of admission decision-making. It is well known, for instance, that offenders serving fixed sentences are occasionally transferred to medium secure units at the behest of the Secretary of State upon reaching their earliest date of release (Morris Gibbon, & Duggan, 2007). There are also various contextual and relational factors that may impact on admission decisions, over which the willing offender has little or no control. These factors include the availability of beds, operating exclusion criteria, quality of relationship with referrers, unit ethos (Grounds et al., 2004b) and the less empirically tested phenomenon of ‘different attitudes and perceptions’ among multidisciplinary team staff (Ministry of Justice, 2009, p. 146). Whilst only Psychiatrists have the power to make admission recommendations following the diagnosis of a mental disorder (section 12(2) of the Mental Health Act 1983), the admission decision itself may be informed by complex interactions between multidisciplinary staff involved in pre-admission assessment and post-assessment deliberations.

The probable complexity of admission decision-making is reflected in a vast and important sociological literature, much of which is beyond the scope of this paper. For current purposes, this literature can effectively be summarised into two separate models. The first model is a development of continental philosophy post-Marx, and posits that group decisions should be seen as a product of the potential for individuals within those groups to positively or negatively influence one another. Within groups acting on competitive professional strategies, decision-making power is not held by any individual within the group, such the Responsible Clinician (RC); rather, ‘[p]ower only exists when it is put into action’ (Foucault, 1982, p. 219). The impact of this on multidisciplinary working is often defensiveness, unclear professional boundaries and responsibilities, and lack of solidarity (Glover-Thomas, 2007).

The second model, deriving from the Anglo-American sociological tradition, posits that group members are primarily motivated by the achievement of collective goals consistent with their group mandate. For instance, members of the multidisciplinary forensic team may have the goal of admitting suitable offenders with ASPD for treatment to reduce reoffending rates. To increase the chances of goal success, medical decision-making groups will recognise the primacy of solidarity over self-interest (Nugus, Greenfield, Travaglia, Westbrook, & Braithwaite, 2010; Parsons, 1989). If, or when, low-level decisional conflict threatens to undermine the group's productivity, the group leader will intervene to direct its decision. Talcott Parsons (1963) posits four means by which the leader will guide her team to a collective decision:

Persuasion (intentional channel, positive sanction): why it would be a ‘good thing’ to agree with the group leader;

Activation of commitments (intentional channel, negative sanction): why it would be ‘wrong’ for individual members to disagree with the group leader;

Inducement (situational channel, positive sanction): the offering of ‘situation advantages’, such as money; and

Deterrence (situational channel, negative sanction): the use of negative sanctions.

Parsons (1965) reminds us that the usual leader of psychiatric decision-making teams is the Psychiatrist, for it is she who historically became the locus of care-based decisions when an individual's mental difficulties extended beyond the capabilities of the family. Historical legitimacy is now reflected in the preponderance of Psychiatrists in charge of the patient's overall care (the RC role), and her authority to make medical recommendations under the Mental Health Act 1983. It is also reflected in the preference of empirical researchers to examine psychiatric admission decision-making from the perspective of the Psychiatrist's decision frame alone. Most notable among those researchers are Grounds et al. (2004b).

Covering 98% of the medium secure bed estate, grounds and colleagues found that ‘a complex range of contextual factors impinge on admission decisions’ (p. 48). One of these, Psychiatrists stressed, was that the multidisciplinary team must achieve ‘a clear and shared view of what [they are] trying to achieve’ before an individual is admitted (p. 40). The implication that it would be the Psychiatrist (group leader) who would resolve group conflict in this situation has been confirmed elsewhere:

I think you could direct [the decision] one way or another … I could … in quite an influential way, because of the position that I have, not because of who I am but of what I am. (Psychiatrist in Grounds Howes, & Gelsthorpe, 2003, p. 129)

The problem with this view is that it presupposes that prototypical leadership qualities are more influential than other complex relational phenomena, such as the use of subversive strategies by group members (competition), in explaining admission decisions. Tickle counters that the potential for staff to ‘disagree when making such decisions’ means that our current understanding of the decisional processes underlying prison transfers is inadequate. Indeed, Grounds et al. conclude that further research with a ‘more varied sample base, particularly including other medium secure unit staff, would be merited’ (2004b, p. 48).

The current study responds to this call by examining decisions to admit offenders with ASPD from various feeder prisons to a medium secure unit for treatment (under section 47 of the Mental Health Act 1983) from a variety of professional perspectives. Aims of the study were to determine whether cooperative or competitive strategies characterised multidisciplinary team decision-making, and to generate moderatum generalisations for use by future researchers wishing to explore the complexity of forensic decisions.

Working method

The study took place on a specialist ward for male personality disordered offenders in a medium secure unit located within the auspices of the Nottinghamshire Healthcare NHS Trust. The decision to conduct the study in the field of personality disorder was mediated by two factors. First, the uncertainties surrounding the responsiveness of ASPD to treatment, and the potential for fractious behaviours by those with the disorder, has been shown to complicate admission decision-making (Grounds et al., 2004b). Second, a complimentary aim of the study was to examine the rehabilitative effects of treatment, if any, once admission had taken place. The results of this inquiry are presented in a separate paper, and require an alternative theoretical framework.

To explore the identified themes of cooperation, competition and leadership in the decisional forum, multiple methods of data collection were employed. The primary method was semi-structured interviews with 12 members of the multidisciplinary team: Psychiatrists (n = 2; cited as 1a and 2a, respectively), Nurses (n = 6; 1b to 6b), Psychologists (n = 2; 1c and 2c), Occupational Therapists (n = 2; 1d and 2d). Each interview took place in a private room, and lasted between 20 and 75 min (providing over 10 h of data).

The essential prerequisites for inclusion in the study were that the professional be involved in pre-admission assessment and post-assessment deliberations. To avoid what Becker (1967) has described as the assumption of a ‘hierarchy of credibility’ within ‘ranked groups’, all members of the multidisciplinary team conducting pre-admission assessments and post-assessment deliberations were invited to participate, irrespective of seniority or grade. Since no generalisations were apparent on the basis of professional seniority, the author only refers the reader to the profession of the respondent.

The aims of the semi-structured interviews were, first, to identify the impact of contextual factors on discretionary decisions taken by multidisciplinary staff to admit offenders with ASPD for treatment; and, second, to pursue the more sensitive inquiry of whether members of the team, including the RC (Psychiatrist), were generating cooperation in decision-making through the use of strategies. Whilst the RC, as group leader, may secure the cooperation of his or her team through the use of explicit, Parsonian-type strategies, it was anticipated that non-leaders would be more likely to rely to subversive strategies to secure compliance in the event of disagreement. Clearly, this sort of inquiry would not have been usefully served by participant observation or ethnographic methods.

Staff were selected for participation on the basis of availability (convenience sampling), and the giving of informed consent. All respondents gave their consent for interviews to be audio taped. Interview content was then transcribed, and a thematic content analysis of the anonymised data was undertaken. Consideration was given to the use of a statistical package (such as NVivo), but manual analysis was deemed to be sufficiently robust for the sample size. Moreover, computer-assisted analysis is not free of bias, because ‘themes have to be coded in the first place’ (Grounds et al., 2004b).

A further method of data collection was systematic and detailed analysis of 34 patient medical records (for which patient consent was received). The records provided useful qualitative data on pre-admission assessments and written recommendations by staff of the perceived suitability of respective patients for admission and treatment. Further, Supplementary, data was achieved by three methods. First, informal conversations with four members of staff (three Nurses and one Health Support Worker); second, formal observation of a preadmission assessment interview with an offender in prison, conducted by the

RC (Psychiatrist), a Nurse and an Occupational Therapist; and, third, formal observation of a pre-admission assessment exercise with the aforementioned offender. In all cases, contemporaneous ‘field’ notes were taken. The decision not to use an audio tape for these observations was taken to reduce the chances of contrived group dynamics (Pope & Mays, 1995).

The study's working method was approved by an NHS ethics committee and the researcher's academic institution. Remaining limitations may include the study's single site design and relatively small sample size. Melzer, Tom, Brugha, and Fryers, contend that individual unit characteristics can have ‘powerful effects on admissions, and thus generalising from local studies can be hazardous’ (2004, p. 8). However, it is submitted that it is more hazardous to divorce decisional phenomena from their interrelational context in an effort to create generalisable data about discretionary decisions. Complex relational interactions are only ever intelligible from the standpoint of a specific location. Furthermore, researchers who pursue data collection beyond the point of data saturation inevitably encounter ‘repetitive discussions’ of decisional dynamics taking place within that institution (Lyall & Bartlett, 2010, p. 891). The findings below demonstrate that single site research has an important role to play in the development of moderatum generalisations (those linked to the broad features of the research) of use for future researchers pursuing (multisite) research.

Pre-admission assessment: the operation of exclusion criteria

Once an offender has been identified through referral, it is commonplace for him to receive a full multidisciplinary assessment in prison. The first assessment is usually undertaken by a section 12 (of the Mental Health Act 1983) approved Psychiatrist with experience in the diagnosis of mental disorder. The outcome of this assessment normally determines whether further assessments take place:

If I thought that someone [with ASPD] wasn't suitable, I wouldn't ask the rest of the team to see them … And there are usually very good reasons when someone is excluded right at the beginning … But most get the full assessment. (Psychiatrist 1a)

One ‘very good reason’ to suspend further assessment is if the offender says ‘there's nothing about them which needs to change’ (Nurse 6b). Unmotivated patients are a drain on professional resources and a disturbance to the therapeutic ethos of the ward. A further exclusion criterion is applied in respect of offenders who receive a comorbid diagnosis of ASPD and schizophrenia. Even if the psychosis is in remission, most psychotic patients find ‘the stress of the treatment is actually quite damaging’ (Psychiatrist 2a). This particular finding compares with previous research (Grounds et al., 2004b), in which it was highlighted that many national units had effectively become psychosis only services. The reason for the discrepancy is likely to be different treatment emphases between the two disorders: psychotherapeutic approaches used for the treatment of ASPD may be incongruent with the ‘fundamental and characteristic distortions of thinking and perception, and by inappropriate or blunted affect’ common to sufferers of schizophrenia (WHO, 1992, p. 54).

A third criterion operates to exclude offenders displaying hostile or assaultive behaviour. By comparison, if the offender had a mental illness, hostility would usually be ‘one reason why you would consider admitting [them]’ (Psychiatrist in Grounds et al., 2004b, p. 37). The different approaches can, again, be explained by divergent treatment approaches between disorders: psychotherapeutic treatment aimed at reducing antisocial behaviours can only be effective with the cooperation of the patient, whereas the absence of cooperation is no bar to the compulsory administration of psychotropic medication for the treatment of symptoms related to a patient's mental illness (section 63 of the Mental Health Act 1983; B v Croydon Health Authority (1995) Fam 133).

In most cases, however, the offender will not present as manifestly aggressive during initial interview. Therefore, to reach informed decisions about his suitability for medium security, further multidisciplinary assessment will be required. One important indicator of aggressive tendencies will be the presence of severe psychopathic traits (Hart, 1998). These will be identified through qualitative assessment by the Nursing team, and a Psychologist applying the Hare's Psychopathy Checklist (PCL-R). A formal score of 25 or above (out of 40) on the PCL-R would be cause for concern. A Nurse (1b) explains:

Psychopaths set other people up as well; they're behind the scenes … And they're quite charismatic, you see … And they cause a lot of difficulties within the staff team because they'll target people, and they'll befriend people, and people will feel quite positive about them, and others feel quite negative. And they actually create that scenario. And it often results in violence …

A final exclusion criterion is the presence of learning disabilities. The extent of any functional difficulties is assessed by a Psychologist using the Wechsler Adult Intelligence Scale (WAIS). The WAIS defines intelligence or I.Q., as ‘The global capacity of a person to act purposefully, to think rationally, and to deal effectively with his/her environment’ (Wechsler, 1939, p. 229). An I.Q. score of below 75 would not satisfactorily reflect these qualities for the purposes of admission and treatment. Excluding those with low I.Q. helps maintain therapeutic conditions on the ward, and ensure the safety and well-being of vulnerable individuals:

[They're] quite vulnerable, and it's frustrating for other patients … They often take more training, but they get it in the end … They can frustrate the patient group … [T]here needs to be a facility for people who are borderline with similar programmes, but with slower, more individual practising. (Psychiatrist 1a)

A further justification for applying threshold scores is that they help engender consistency and predictability into the decisional forum. Consider, for instance:

I don't think [the Responsible Clinician] would let someone come in who was really highly dangerous, posing lots of risks, high PCL-R, no motivation. I don't think he'd entertain them, and neither would we. (Nurse 2b)

What this does not mean, however, is that the multidisciplinary team always agrees with the RC. Whilst collaborative processes characterised many, if not all, of the decisions reached by the team, the findings suggest that admission decisions taken by the RC can be heavily influenced by the use of subversive strategies by members of her team. This confirms that discretionary decisions explaining the admission of offenders with ASPD to hospital are best explored from a multidisciplinary perspective.

The pre-admission meeting: collaborative decision-making?

Following pre-admission assessment, ‘[we] come back [to the ward] and do the pre-admission meeting, when the team decide whether the person can be managed in a medium secure unit (Nurse 3b). Within this relational context, members of the multidisciplinary team provided two general accounts of decision-making. The first was that admission decisions took place within a collaborative framework, but were ultimately the province of the RC:

I think as far as the various disciplines are concerned, everyone has an opinion. But, ultimately, it will be guided by what the RC will say. We do our best by saying, we think this person should be brought in, or this person shouldn't be brought in, but if you do bring him in, then these are the things you need to watch out for. I think that's the best we can do in that situation. (Psychologist 1c)

‘The Psychiatrists just make that decision, I think. But you can put your recommendations in your report, and that's all you can do, really’. (Nurse 3b)

The second, more nuanced, view was that decisions were taken by the RC, but, in the event of substantial disagreement, members of the team involved in the front line patient management cast the deciding vote:

I think if there's very substantial disagreement between the various professions involved, then Nursing generally trumps the others. And I think it ought to trump the others, because the Nurses have to manage the patient on a day–by–day basis … It hasn't happened very often; it's happened, in fact, very infrequently when there's a stand-up fight about the issue, but when that has occurred in the past, I certainly have given way to the Nursing views. (Psychiatrist 2a)

[A]t the end of the day, if the Nurses have got to manage these people 24 h a day, you know, they're there at 9 o'clock at night, not a Psychiatrist, when someone is shouting down a corridor and, you know, threatening people. So, that usually did win the war, basically. (Nurse 1b)

This latter quote indicates either an inversion of the Parsonian account – that group members, not group leaders, have the ‘rights to the action of others’ (Parsons, 1989, p. 121) – or that Nurses’ reliance on their professional currency in the decision-making forum is a situational advantage, or strategy, used to determine admission decisions. The latter account should be preferred for three reasons. First, Nurses do not have the legal ‘right’ to authorise admissions under the Mental Health Act 1983: psychiatric recommendations are the ‘cornerstone of the admission procedure’ (Bean, 1986, p. 35). Second, the Psychiatrist does not always follow the Nursing view when conflict emerges. And, third, agreement with the decision of the RC may conceal the fact that a strategy has been used by the Nurse to frustrate admission:

… we'd say, right, this is what they've got to achieve in the next 3 or 6 months, or whatever, and we'll keep in contact … [W]hen you do do something like that, very often, the patient can't comply with it, and that's why they don't get admitted, because they perhaps don't choose to come off their Valium, which they're addicted to over a three month period, and they're not prepared to sign up and reduce it …’ (Nurse 1b)

On the other hand, if the offender is admitted contrary to the wishes of the Nursing team, his stay may only be temporary:

I think one goes against the Nursing view at one's peril … I think with PD it's an interaction, and, if people have a particular view at the beginning that someone is unsuitable, then they will need a considerable amount of persuading that that view is wrong, unfortunately … (Psychiatrist 2a)

These examples suggest that the centrality of Nursing practice in ongoing patient management has the potential to act as a hierarchical structure which replaces the uncertainty of antagonistic professional relations. They also indicate that there is no indefeasible link between (legal) authority and decision-making power: decisions reached within group fora are ‘pervasive, complex and often disguised’ (French & Raven, 1968, p. 150). However, the problem with this moderatum generalisation is that absent evidence to unpick the

‘pervasive’ nature of power, ‘people may agree about the facts [of decision making], but disagree about where power exists’ (Lukes, 2005, p. 63). This empirical limitation will strengthen the presumption of RC-led decision-making in forensic practice.

To address this limitation, there are several potential reasons why respondents may not have provided further examples of strategy used to frustrate the decision of the RC. One such reason is the perception among forensic professionals that manipulative behaviour is evidence of psychopathology Another reason is the relatively small sample size employed by the study; for instance, the number of Nurses in the sample was not quite proportional to their representation in the multidisciplinary team. It was also possible that the multidisciplinary team simply had ‘confidence, trust, and faith’ in the ability of the RC to reach appropriate decisions, and that this reduced incidences of decisional conflict (Parsons, 1989, p. 55). To explore this possibility, and to account for the absence of generalisable conflict in the majority of team decisions, the data were analysed for contextual factors that might be impinging upon the free play of competitive strategies within the decisional forum.

The informal context to admissions

Formal changes to pre-admission assessment

It transpired there were two contrasting accounts of assessment and post-assessment. The first was that decision-making was the product of highly purposive assessment exercises used to guard against admitting unsuitable offenders for treatment:

The Nursing assessment is a lot more formal than it was before … [they do] the family history, index offences, whether there's been any adjudications, whether they were on medication, whether they've got issues of smoking, anything that affects us, risks, you know, any risks they might present on the ward. (Nurse 4b)

The second, prominent account was that pre-admission assessment had become increasingly more informal in recent years. For example:

I've been on the [risk] assessment training and people have said: ‘What d'you mean you're only asking these questions – what about the HCR-20? What about the in depth questions? You should be sticking to a more formal assessment … (Nurse 2b)

The effect of informality on the depth and breadth of pre-admission deliberations was reported to be considerable. An Occupational Therapist (1d) stated: ‘We used to have something that wasn't tagged on to the end of a [meeting]. We'd have an hour, two hour slot, where we'd discuss things in much more depth’. One relatively inexperienced Nurse (5b) even declared that pre-admission assessment was ‘just to get an idea of what [they're] like’: in the event that the patient disengaged with treatment – or otherwise proved difficult to manage – upon admission, he could be sent back to prison, either indefinitely or for a ‘period of reflection’.

The importance of ‘a clear plan in place to respond to any difficulties we might face with [those] admissions’ (Occupational Therapist 2d) was confirmed by all staff. Some, however, expressed concern that the option of prison re-admission increased the chances of unsuitable offenders being admitted in lieu of thorough assessment and post-assessment discussion. When the unsuitable patient was inevitably conveyed back to prison, they would become ‘doubly stigmatised, because a specialist system couldn't help them’ (Psychologist 1c). Others pointed out that prison re-admission is complicated by factors such as high staff turnover and prison bureaucracy: ‘We've got to show we've attempted everything before we send [them] back’ (Nurse 2b).

Since both of these negative outcomes are the direct result of team decisions, they seem to undermine the assumption that group solidarity in admission decision-making is apt to increase productivity, and question whether the reduction in incidences of conflict due to leadership (the RC) is really the product of long-standing trust relations. Alternatively, if it is argued that group decision-making is mediated by co-dependency of leadership, trust and solidarity, it must be because the degree to which this leads to productivity – and, indeed, the definition of productivity – is presumed to be highly sensitive to the context in which group decisions are taking place.

The ‘bed situation’

Reasonable demand in relation to bed capacity is one form of evidence for the cost effectiveness of healthcare interventions for people with personality disorder. Low demand implies inefficient use of available healthcare resources, and may lead to the suggestion that public money is best invested in other mental health services (Bartlett, 2010, p. 25). Typically, researchers have found that medium secure units are inundated with admission requests from referrers: ‘Not only [here], but anywhere, we can't even find beds nationally most of the time’ (Psychiatrist in Grounds et al., 2004b, pp. 41–42). However, the situation in the current study was quite different:

… ideally, we should have a full ward. And we have had in the past a dozen people, and half of them have been assessed and are just waiting for a space and half of them we could assess in the future. But now it seems we're always down to ten [out of twelve] beds, and there's [sic] three or four people who need to be assessed. (Nurse 2b)

This, in turn, has a disarming effect on the quality of preadmission assessment:

[It] was very formal. We were in a position where things were different anyway, and we could cherry pick. Bed situation and, obviously, commissioners are questioning if you're sitting there with empty beds and pressure's on, and then people start to question the viability of that service:

Do we need that service if we can't fill beds? (Nurse 1b)

As would be expected, an even greater impact was felt on post-assessment deliberations. The need to ensure the viability of the service assuages the usefulness of conflict between group members:

I think one of the unfortunate things is we are governed by our bed situation …, and I think it has influenced some of our decision making, that perhaps when we may have said no …, we have said yes. (Occupational Therapist 1d)

Shaw posits that the ‘pressures towards uniformity’ exerted by contextual factors cause individuals to conform to that of the ‘modal group member’ (1976, p. 165). The notion that groups in conflict can reach a consensus without the input of a group leader is supported by the data:

There is disagreement …, but I get the sense … sometimes that there's an underlying understanding from people who are expressing concern and disagreement that we've got to fill the bed, you know, we can't create too much disagreement because, you know, that bed has got to be filled. (Nurse 4b)

However, a major limitation of a Shaw's analysis (conformity linked to the modal, rather than lead, group member) is that it implicitly suggests that group conflict is both functional and prosaic (we can't create too much disagreement). But, this is inconsistent with the force of the statement – ‘There is disagreement …, but I get the sense’ – which indicates that functional conflict between group members had become subordinate to implicit assumptions favouring the admission of the offender (I get the sense’ (Nurse 4b)). Since implicit assumptions must be rooted in explicit norms, two functions of a group leader can be proposed. First, she may bring important norms to the attention of the group; and, second, she may engage in strategy to resolve conflict that arises, or pre-empt conflict irresolution. As well as preventing the escalation of conflict, her use of strategy will reinforce the generally robust implicit assumptions about group decisions. To put this another way: when resource pressures act to constraint the use of discretion, the leadership model of group decision-making becomes compelling.

A useful example of conflict resolution by the RC derives from a pre-admission assessment conducted by a Nurse, Occupational Therapist and Psychiatrist (RC) with an offender in prison. Following the offender's assess-ment, two staff members (the Nurse and Occupational Therapist) were adamant that he should not be admitted. The RC (2a) responded, pragmatically: ‘But we need to fill beds’. The effect of the intentional channel, negative sanction (activation of commitments; Parsons, 1963) was that the RC was able to secure group support for the offender's admission. The RC writes:

Both my nursing and Occupational Therapy colleagues were less than enthusiastic about admitting [him] … I, on the other hand, took a different view … After discussion today, it was agreed that he ought to be admitted for a further assessment. In the event, however, of him refusing to engage on any terms other than his own; then we would wish to return him to prison. (Taken from a letter written to the prison healthcare team)

It could be argued that the RC's reference to the prison re-admission safety net is evidence of another Parsonian strategy, namely: persuasion (intentional channel, positive sanction). But, again, any patient who it transpires is unsuited to the ward regime may be sent back to prison; the probable aim of forewarning prison staff of this fact was to mitigate any bureaucratic hurdles acting against this eventuality. As such, the admission decision will further justify the team's underlying trust in the ability of the RC to make productive admission decisions in difficult times.

The limitations of exclusionary criteria

The context in which trust in the RC appeared to be more precarious was if members of the multidisciplinary team (usually Nursing staff) perceived that the admission decision ran counter to an offender's therapeutic best interests. An example was the admission of offenders serving fixed term sentences whose earliest date of release (EDR) would expire before the completion of treatment (approximately two years):

You know, we've had someone who's only got nine month left on their sentence. What can we do for this person in nine months? Again, there's a sense that we're setting people up [to fail] … (Nurse 4b)

However, other members of the multidisciplinary team took the view that the two-year time frame for treatment was arbitrary. For instance:

… I think six months is sometimes enough. I think sometimes people come, there's a particular skill, they get skilled up in it, and that to me is success … Some people would perhaps need longer here … Some people it's bang on. Some people it's too long. Some people it's not long enough. I guess sort of shifting the way people think and function is success, however long that may be. (Occupational Therapist 1d)

It was also noted that the respective offender would have the opportunity of remaining on the ward upon reaching his EDR:

[W]e point out to them that, once they do pass their EDR, if they then choose not to stay, certainly my practice is that I would discharge them, irrespective of whatever risk they posed. And, in fact, there is [a patient] in exactly that position at the moment, who has about 9 months to serve. He certainly needs about two years, and we've offered him the option of coming in and staying in beyond his 9 months, but it's not absolutely required that he does so. But if at the time of his 9 months he then decides to leave, I would discharge him. (Psychiatrist 2a)

The promise of discharging (risky) patients, even when the presence of empty beds favours continued admission, adds weight to the generalisation that ‘Clinicians resist pressures they perceive to be in conflict with a primary therapeutic purpose for their services’ (Grounds et al., 2004b, p. 48). Whilst, therefore, the decision to admit an offender with less than two years remaining of his fixed term sentence represents a shift in ward policy, trust in the RC is maintained by the general perception that her decisions have therapeutic intent.

A similar, ostensibly negative, trust-context coupling operates in respect of the admission of highly psychopathic offenders and those with low I.Q. scores:

To be honest, there's people I think are totally inappropriate. There was a certain level on I.Q. to come here …, because you needed that level of understanding to engage with group work … And we've had people [with I.Q.'s] in the 60s, and they're not successful. I mean [one patient] is hanging on by his fingernails at the moment, and it's a shame. It's one of the things I struggle with … (Nurse 4b)

[W]e've never been very successful with people with high PCL-R scores [over 30]. I don't know of one high PCL-R patient who's survived the whole programme. (Nurse 2b)

At first blush, these remarks are surprising. For it has always been accepted that the relationship between preadmission assessment scores and potential treatment benefit is imprecise. With regards I.Q., an offender with a WAIS score of below 83 would have once been refused admission to the ward; however, the cut-off score was reduced to 75, because it was found that even low scorers could ‘respond well to the type of structure the PDU offers’ (Psychologist 2c). If following admission, the programme proved to be too intellectually challenging, staff would, in line with evidence-based guidelines (NICE, 2010), ‘tailor [the programme] for that person’ (Nurse 1b). The result for the patient in question (I.Q. Of 67) was unambiguous: ‘very positive treatment period since his last CPA review’ (Nursing report, four months into treatment). Nevertheless, the discrepancy between staff views may indicate that those with a negative view at the pre-admission stage do, indeed, need ‘considerable amount of persuading that that view is wrong’ during treatment (Psychiatrist 2a).

Like I.Q. score, the decision to refuse entry to those with PCL-R scores over 25 has no concrete therapeutic justification: ‘[t]here's no good evidence that treating people with high PCL-R scores is good or bad – we simply don't know’ (Psychiatrist 2a). The problem is that the PCL-R assessment does not provide an empirically reliable means of identifying those who are unsuitable for current interventions intended to reduce criminal re-offending (Blackburn, 2007). Even if it did, gaining access to the necessary prison information cannot be guaranteed:

… you'd want access to healthcare records; prison records; their custody files; their psychology file; their probation file. So, what you're actually saying is … we're coming in; we need a prisoner for a whole day; we need you to help us move around; we need access to every file you have. And of course it doesn't happen. (Psychologist 1c)

It is not without some justification, therefore, that psychologists carrying out PCL-R assessments believed that psychopathy should not act to automatically exclude the individual from treatment. Indeed, and notwithstanding the ‘bed situation’, there is a belief that general forensic mental health services are now admitting ‘more and more severely affected people’ (Psychiatrist 1a). Presumably, by bringing this shift in forensic practice to the attention of the multidisciplinary team, the RC is capable of encouraging support for subtle changes made to admissions’ policies, without provoking fatal distrust or conflict, both of which would severely undermine the group's resilience to resource pressures.

Group strategies, individual consequences

Parsons describes the strategic phenomenon linked to policy changes as the inevitable consequence of the ‘imperfect integration’ of ‘value conflicts and role conflicts’ (1989, p. 128 – presently: exclusivity vs. inclusivity of admissions, amid the ‘bed situation’; and self-interest vs. group interest). Without the imperfect integration of these integrative factors, the leader would never be required to implement strategies to direct her group towards a particular decision. However, Coleman (1963) points out that, if leadership is truly predicated on trust, strategy should not be necessary to resolve disagreements arising through value conflict. Another problem in Parsons’ account is his unarticulated assumption that socialisation factors enable group members to be a priori conversant with group core values and professional role orientations. Other commentators, such as Foucault, contend that values and role orientations in society are often incomprehensible to the population: ‘the logic is perfectly clear’, he argues, ‘the aims decipherable, and yet it is often the case that no one is there to have invented them, and few who can be said to have formulated them’ (1981, p. 95). If the population is, indeed, only dimly aware of societal norms and role expectations, one crucial question that arises in respect of resource pressures is: why do group members invest so much in the potential productivity of the group, if they believe the result is inappropriate admission decisions? For example:

Before Christmas, we had about three people on here with high PCL-R scores, and it was quite chaotic on here. There was [sic] a number of staff who burned out. (Nurse 4b)

It is likely that those who disagree but abide by admission decisions do so in pursuit of other socialised need dispositions, such as wealth, identity and future professional status. These tangential need dispositions can only be gratified if team members remain ‘responsive to the expectations of others’ (Parsons, 1989, p. 127). By appeal to the findings, then, it appears that that collective interest and self-interest remains intimately linked to admission decision-making, even when resource pressures seem to indicate preference for group solidarity. But, why might this matter?

Individuals who satisfy need dispositions within groups facing resource pressures are more likely to reach admission decisions that, paradoxically, undermine their long-term best interests. Factors such as staff illness and absenteeism, ‘divisive staff attitudes’ and the belief that non-patient-centred decisions are being reached, may lead some staff to perceive the ward as dysfunctional (Grounds et al., 2004b, p. 40). If those professionals are unable to resolve the irreconcilability of role expectations and performance of roles, she may leave for other, more effective, social systems (Parsons, 1989). For those who wish specialist personality disorder services to demonstrate to staff that treatment is possible (Crawford & Rutter, 2007), this would be an unfortunate limitation.

Conclusion

Admission decision-making is a complex relational phenomenon. The findings of the study indicate that the general view that professionals pursue both competitive and cooperative strategies in group situations (Napier & Gershenfeld, 1973) is applicable to the forensic admissions’ context. The findings also strongly suggest that professional preference between the two types of strategies is strongly influenced by bed availability. When referrals are plentiful, and beds are filled, the RC (group leader) is more likely to authorise admission decisions encouraged through the use of subversive strategies by members of her team. Researchers who choose not to examine admission decision-making from a multidisciplinary perspective will inevitably suffer the problem of false ascription between professionals.

By comparison, the presence of empty beds on the ward encourages informality in pre-admission assessment; weakens post-assessment deliberation; and, ultimately, results in RC-led decision-making. Whilst this helps to ensure the viability of the service, the reduced opportunity for discussion of ‘really complex people’ (Psychiatrist 1a) and perceived inability to exercise control over daily decisions may cause disenfranchisement among some team members (particularly Nurses), if not burnout (Ministry of Justice, 2011). To improve staff motivation and well-being, and reinforce institutional purpose, training may be needed to clarify definitional concepts, such as psychopathy and personality disorder (Huband & Duggan, 2007), and to identify research on effective interventions for these groups (Kurtz, 2005). If this does not provide the collateral benefit of reducing professional disagreement on an individual's suitability for admission, the RC should be willing to ‘articulate clear organizational values to which practitioners can feel committed’ (Ministry of Justice, 2011, p. 63). This study has shown that this is not the same thing as securing collaboration through activation of commitments (Parsons, 1963).

Alternatively, and perhaps more importantly, there needs to be a forum for open discussion of differences of opinion arising between members of staff (Kurtz & Turner, 2007). Daykin and Gordon (2011) explore this possibility in a high secure hospital. One form of support they advocate is multidisciplinary team supervision. This forum is designed to provide staff from different disciplines with the opportunity to discuss issues such as current approaches to clinical practice, the nature of team tasks and working relationships. All meetings are convened by an ‘external facilitator’ whose role it is to create a space for the expression of ‘problematic feelings’ and expose group members to ‘the perspectives and approaches of other disciplines’. According to the authors, this form of supervision has ‘reduced professional isolation’ where it has been tested. However, it is unclear what impact, if any, provision has had on actual decision-making processes and working relationships. Moreover, the authors opine, confidentiality issues and uncertain boundaries between personal and professional realms are unlikely to be overcome by a one-size-fits-all supervisory framework.

We must also acknowledge that the tension between competitive and cooperative tendencies in multidisciplinary teams is not only a consequence of the interaction between managerial, supportive and educational needs. Kurtz, for instance, observes that the limitations of treatment can cause professionals to refocus their frustrations onto ‘external, concrete issues’ within the institution (2005, p. 417). Whatever the extent of this truth, it is certainly worth imagining what impact improved confidence in treatment might have on reducing incidences of team conflict.