opinion & debate Forensic psychiatry and general psychiatry: re-examining the relationship {

Something is going wrong with forensic psychiatry, as a concept and as a service. Beds in medium secure units are logjammed, and relations with general adult services increasingly fraught with disputes over resources and responsibilities. Despite a remarkable investment in buildings, and the 300% growth of the forensic specialty (Goldberg, 2006), offending behaviour by individuals with mental illness shows no sign of decline, either in terms of prison numbers (at record high levels in the UK) or the countless demands for risk assessment (Duggan, 1997; Moon, 2000). Indeed, a working party of the Royal College of Psychiatrists reporting on the forensic and general adult interface in July 2003 had difficulty in even defining the specialty, apart from forensic psychiatry dealing with ‘patients and problems at the interface of law and psychiatry’. But most importantly it had to accept that there is no such thing as a forensic patient and that none of the skills of forensic psychiatrists is their exclusive prerogative. Aftercare outcomes for discharged patients from medium secure units show no difference, in terms of reoffending and hospitalisation, between specialist forensic and general adult services (Coid et al, 2007). Viewed historically, forensic psychiatrists appear to be the vanguard of an insidious reinstitutionalisation, high priests of the new religion of risk assessment, and thus primary targets for a scapegoating government agenda. This paper considers the growth of forensic psychiatry, and the unhappy result of its evolution into a separate specialist domain. It argues that forensic services fail to reflect an important advance in our understanding of the risk posed by people with mental illness who offend, namely that such risk is dynamic; in any one individual, risk constantly changes as a result of now wellknown factors. Instead of providing responsive rungs of security that reflect this changing pattern, forensic services comprise a series of rigid, antagonistic subdivisions, and have become hostage to a debate between liberalism and coercion that shows no sign of resolution. To these difficulties we offer various remedies, which centre upon the need for forensic psychiatry to adopt a stance in favour of a unified therapeutic enterprise, care rather than coercion, and the use of its criminological insights to re-engage with the whole patient journey.

Something is going wrong with forensic psychiatry, as a concept and as a service. Beds in medium secure units are logjammed, and relations with general adult services increasingly fraught with disputes over resources and responsibilities. Despite a remarkable investment in buildings, and the 300% growth of the forensic specialty (Goldberg, 2006), offending behaviour by individuals with mental illness shows no sign of decline, either in terms of prison numbers (at record high levels in the UK) or the countless demands for risk assessment (Duggan, 1997;Moon, 2000).
Indeed, a working party of the Royal College of Psychiatrists reporting on the forensic and general adult interface in July 2003 had difficulty in even defining the specialty, apart from forensic psychiatry dealing with 'patients and problems at the interface of law and psychiatry'. But most importantly it had to accept that there is no such thing as a forensic patient and that none of the skills of forensic psychiatrists is their exclusive prerogative. Aftercare outcomes for discharged patients from medium secure units show no difference, in terms of reoffending and hospitalisation, between specialist forensic and general adult services (Coid et al, 2007). Viewed historically, forensic psychiatrists appear to be the vanguard of an insidious reinstitutionalisation, high priests of the new religion of risk assessment, and thus primary targets for a scapegoating government agenda. This paper considers the growth of forensic psychiatry, and the unhappy result of its evolution into a separate specialist domain. It argues that forensic services fail to reflect an important advance in our understanding of the risk posed by people with mental illness who offend, namely that such risk is dynamic; in any one individual, risk constantly changes as a result of now wellknown factors. Instead of providing responsive rungs of security that reflect this changing pattern, forensic services comprise a series of rigid, antagonistic subdivisions, and have become hostage to a debate between liberalism and coercion that shows no sign of resolution. To these difficulties we offer various remedies, which centre upon the need for forensic psychiatry to adopt a stance in favour of a unified therapeutic enterprise, care rather than coercion, and the use of its criminological insights to re-engage with the whole patient journey.

Rise of forensic psychiatry
In 1970, anyone asked to identify the specialist forensic services would have found 2 professors and 18 consultants confined to working in a few grim special hospitals. These forbidding constructions (now called high secure units) were largely custodial and therapeutically arid. But since inception as a recognised specialty in 1973, forensic services have changed out of all recognition. They now possess state of the art buildings adorned with electronic gates, smart wire fences and CCTV. They have developed a wide locality base, a high academic profile, multidisciplinary healthcare staff and, sweetly enough, a range of sub-specialties. The term forensic has come to embrace something much wider than its definition in the Oxford English Dictionary 'pertaining to, connected with or using courts of law'.
It is not clear what stimulated these developments, although there has been strong governmental support for this new domain, reflecting the big brother shadow of the Home Office over the Department of Health. Energetic individual champions clearly played a part, combining academic rigour with political nous. But the key factors relate to deinstitutionalisation, our improved knowledge of the link between offending and mental disorder, and the rise of a legalistic culture funded to attribute blame.
The closure of the asylums in the second half of the 20th century (Barham, 1992;Jones, 1993) and the shift of mental health services into the community created gaps in provision. The few remaining in-patient beds, whether sited in district general hospitals or stand-alone units, were unsuitable for the transfer of patients either from higher security units or, often, from prison. This community orientation meant that nothing was available for mentally disordered offenders who were seen as requiring something less than high security but not an acute open ward. In need of at least a period of compulsory in-patient treatment with rehabilitation, they were by default either remaining in a high-security hospital or, if in prison, they were not transferred out (Birmingham, 1999). The private sector swiftly exploited these concerns for more effective treatment, and the ever-rising cost of placement in private hospitals, rather than a sound evidence base, has been a major influence in the expansion of medium secure units in the NHS. Doubling their formal admissions, from 814 to 1629 a year between 1994/1995 and 2004/2005, the independent hospitals now take some 24% of court and prison disposals (405 out of 1664) compared with some 6% (123 out of 2111) in 1994/1995(Information Centre, 2006. Furthermore the special hospitals had fallen foul of deprivations common to most total institutions (for example, see Goffman, 1961). They needed modernisation and contraction, all three in England (Broadmoor, Ashworth and Rampton) having attracted embarrassing inquiries into a range of inadequacies of care as well as abusive practice (for example see Dyer, 2003). Rampton alone has attracted three such inquiries, and even threats of closure. Out of necessity, move-on capacity had to be created, and most importantly, the patients in these institutions deserved proper rehabilitation and a chance of progress to lower levels of security.
The response to this problem came in the influential Butler Report of 1975 (Home Office & Department of Health and Social Security, 1975), which set out the need for services along a 'ladder' of security, from smaller special hospitals, through a new tier of medium secure units, to specialist teams providing community monitoring and support. Asked to service the general community, special hospital and prison populations, forensic services were caught in a dilemma of care versus containment. Answerable to not one, but two government departments, and dogged by the same lack of resources that has always bedevilled psychiatry, the next 30 years saw the uncoordinated evolution of a highly heterogeneous service. Reviewing the state of services in the late 1990s, Coid et al (2001) found that the term forensic embraced an assortment of services, in which the needs of the prisons took primacy; support for the corresponding local adult general service was best where overall forensic demand was lowest. By the early 2000s, therefore, the medium secure units had become new monoliths, disconnected from the very communities they were partly intended to serve.

The inexorable rise of 'risk'
Common sense has long held that there is a relationship between mental illness and violent behaviour. This is now established beyond doubt (Monahan et al, 2001); among mentally ill populations, active psychotic illness, personality disorder and comorbid substance misuse carry clearly increased correlations with violence (for example Mullen, 2006). The identification of specific clinical and historical factors has also led to ways of measuring and managing the risk posed at a given time. These advances suggest an ostensible foundation on which to build a distinct medical specialty, but such a response may not be sensible. Many non-psychiatric variables, especially a combination of youth, male gender, substance use and low socio-economic status, reveal a far greater association with violence.
Mental illness is only a modest risk factor for the occurrence of violence, and in the case of psychosis, only individuals with current psychotic symptoms carry significantly increased risk (Link et al, 1991). Those with only historical psychotic symptoms carry a much lower risk of violence. Furthermore, among the population with mental illness as a whole, by far the majority appear to pose little or no risk of violence at all (Appleby et al, 2006). A rational response would therefore involve the distribution of resources on the basis of overall need, rather than on the primary basis of risk (Szmukler, 2001). Where risk is concerned, however, our contemporary response is highly irrational.
Since the 1960s, our culture has arguably come to value the rights and comforts of the individual at the expense of the traditional values of civic responsibility. With increasing social atomisation there has come a decline in deference to authority, and a belief instead that the problems that have troubled humankind since antiquity will eventually yield to the systematic application of knowledge, derived from evidence. This is reflected in the plethora of inquiries, guidelines and targets that are nowadays used to address an ever-increasing number of perceived risks to the individual. Risk has become a central feature of modern life; a veritable industry has grown up around its detection, assessment and management. The risk posed by the fraction of people with mental illness who offend has always generated concern (BMJ, 1895), but as care for those with mental illness has moved out of institutions into the gaze of an increasingly risk-obsessed public, the intensity of the reaction that it provokes has grown out of all proportion to the actual risk involved (Ward, 1997).
Likewise, demands for reinstitutionalisation have grown. Lurid media reports of crimes by people with mental disorders have become a principal source of information by which the public form their views about mental illness (Philo et al, 1994). The same preoccupation has filtered into government policy; the proposed revision of mental health legislation sought to reassure that the new law would be 'safe, sound and supportive' (Department of Health, 1998). It is unsurprising that an apparently qualified enthusiastic body of expertsforensic psychiatrists -equipped with a well-funded armoury of seemingly specialised tools and techniques, working in state-of-the-art premises, should be an enticing proposition for a public that has become obsessed with risk to the individual. Whether or not this body can actually deliver what is expected of it remains to be seen.

End of honeymoon for forensic psychiatry
The problems that have arisen from this enterprise could not have been foreseen in the 1970s, the era of Turner & Salter Re-examining forensic and general psychiatry opinion & debate antipsychiatry, when 'greater demands for forensic skill' were the basis for forming the forensic section of the College. At that time, this skill seemed definable: the successful management of the risk posed by dangerous individuals with mental illness. While forensic capacity was still expanding, these skills appeared effective. Wellpublicised court diversion schemes, prison in-reach programmes, and in some areas, even liaison with the local community services, seemed to rise to the challenge posed by this small but difficult population. Those admitted to these well-resourced new units, staffed by professionals displaying the enthusiasm common to novelty, responded well to intensive treatment; symptoms, and risk, subsided. In retrospect, however, this may have been a phoney period. As soon as the system reached capacity, and the need for move-on became explicit, there was less enthusiasm for returning forensic cases to the community, where presumably they should continue to benefit from specialist forensic techniques.
Doubts emerged over the usefulness of the methods by which forensic specialists measure risk (Szmukler, 2001). Among those who feel that this debate has moved on, the question of just who should be making routine use of these tools remains (Maden, 2005). A further question relates to the timescale over which forensic operations are conducted. The average forensic length of stay is many times longer than its adult general counterpart, often by years. Aside from the striking similarity between forensic lengths of stay and prison sentences (for similar index offences among the 'well' population) it is not unlikely that anyone isolated for so long from the outside world would undergo enduring changes of belief and conduct, whatever the intervention. Forensic risk management has come to resemble the search for nuclear fusion: something done by experts deep in bunkers, on irrelevant timescales and at great expense, with the allure of a unique benefit for the world at an indefinable point in the future. In fact, it is little more than ordinary general psychiatry, practised indoors, with ample resources, on a completely different timescale.
Given the substantial diversion of resources away from the vast majority of the population with mental illness who do not pose a significant risk (and the relative neglect of the 'harmless', bedsit-bound chronically ill patient), the onus is upon forensic psychiatry to demonstrate that their specialist techniques are effective beyond their bunker. Other specialised services, for which optimistic claims have also been made, are now reaching more sober conclusions about outcome (Killaspy et al, 2006) We predict that forensic community psychiatry will reach similar conclusions over the next few years; but what will differ significantly will be the response of an increasingly illiberal public.

Problems with definition
The question of true difference between forensic and adult general psychiatry is not confined to tools, therapies or outcomes, but even to definitions of 'the forensic case' itself. What is the criterion for admission to a medium secure unit beyond the phrase 'requires medium security'? The definition does not seem discernible in terms of diagnosis, duration of illness, clinical course, or even response to treatment. Adult general psychiatry carries large (sometimes by tenfold compared with forensic) high-risk case-loads that are a mix of psychosis, personality disorder and substance misuse. When forensic and adult general specialists compare case-loads, there is an extraordinary overlap (Dowsett, 2005). Index offences among the forensic group may be more serious, but the offending is usually remote; current risk is another matter.
Why therefore keep someone in a forensic setting? The explanation is that forensic services, as they currently stand, primarily exist to fulfil political demands for a visible and coercive response to risk. Effective treatment of illness appears to be only a secondary consideration. Meanwhile the poorly adherent, treatment-resistant patients, with as often as not dual diagnosis, who have constant offending histories and poor impulse control, come in and out of general acute wards, frequently abusing and hitting staff on their way, with the police often unwilling to prosecute given the pressures on the prison system (Tuddenham & Hunter, 2005). If there is no such defined entity as the 'forensic patient' then how can we say that forensic psychiatry as it has developed in the UK is a genuine specialty?
In fact, the defining feature of a forensic case is a retrospective view of the concern provoked by an event that, by definition, has already occurred. The forensic response -incarceration -leads to a paradox: high risk is low risk. Time passes; the concern engendered by the event diminishes, and passage to a lower level of security becomes feasible. But, because the forensic services have evolved as a stand-alone service, the facilities for such a flexible response to the dynamics of risk, which requires an integration with adult general services, are commonly inadequate, and in some areas, virtually non-existent (Turner & Salter, 2005).
A typical collision point between the forensic and adult general philosophies is the local psychiatric intensive care unit (PICU). This is seen as a 'low' secure facility by forensic specialists, and therefore a convenient place to put a mentally disordered offender pending evaluation at the court's direction, often for months at a time. From the generalist viewpoint, however, PICUs are intended as a brief intensive care resource, for the most disturbed patients on a general unit, enabling general adult psychiatrists to have genuinely 'open' wards, and a therapeutic rather than custodial ambience.
The reluctance of adult general psychiatrists to accept low secure cases into this environment is often perceived as obstructive by forensic specialists, and many forensic specialists view their generalist colleagues as unwilling to reaccept many patients even when risk is demonstrably low. General psychiatrists, in return, see forensic units as awash with resources and spoilt by the luxury of selectivity based on a specious definition of caseness. Any attempt to address these problems, and so Turner & Salter Re-examining forensic and general psychiatry opinion & debate provide a better service to our patients, will need to consider this difficult relationship.

What can be done?
There are several ways to move beyond the present status quo, some of which are simpler than others. What they have in common, however, is the need for psychiatry to consider a shift -in either direction -along the spectrum between containment and care. Some ways of achieving this would attract less political opprobrium than others; all require a dismantling of the border between the generalist and the forensic perspectives.
The most drastic solution would be to disband forensic psychiatric services as an element of the health services altogether, redistributing the resources so released to provide care for the majority rather than a minority of patients. Management of people with mental illness who offend should be relocated to improved healthcare sections of the prison environment, where it might properly reflect the containment philosophy that currently defines forensic psychiatry. Such a move would sit comfortably with public perceptions of mental illness, and could also carry positive implications for the humanitarian problems that confront the prison services. Regular support at police stations, to evaluate their often disturbed clientele, would also strengthen a preventive and therapeutic role.
Another option would be to withdraw exclusive admitting rights to medium secure units from the forensic specialists. This would bypass the conundrum of forensic definition, yet leave the forensic specialists with a clear professional base. Such a model is well established in mainland Europe; indeed, the intensity of the general/ forensic debate and, more generally, concern over the risk posed by people with mental illness appears to be a curiously British phenomenon. Informal discussion with Scandinavian colleagues, for example, suggests that medium secure units have simply not undergone the dramatic expansion seen in the UK.
Another international perspective for change derives from the USA. Psychiatric practice in America differs from that in the UK in many ways, but one aspect is relevant to this problem. In the USA the organisation equivalent to the Faculty of Forensic Psychiatry of the Royal College of Psychiatrists is the American Academy of Psychiatry and the Law (AAPL). This has a similar relationship to the American Psychiatric Association as the Faculty has to our College, but there are significant differences. The AAPL concentrates on the practice of psychiatry in the courtroom, whether civil or criminal, and is in effect a gathering of expert witnesses. By adopting this role as central to their work, British forensic psychiatrists could avoid the common conflict of interest between the patient and the legal process.
Another response would involve large-scale expansion of low secure, as opposed to PICU, facilities, for which both adult general and forensic teams would have carefully shared responsibilities. This could dovetail neatly into a further improvement, namely dedicated forensic input to existing community teams, set at, or above, a statutorily agreed minimum level. This would end the troublesome heterogeneity seen at its worst in the inner cities and, from a forensic perspective, could avoid the possibly daunting task of demonstrating greater effectiveness of dedicated community forensic teams.
Other changes could be implemented at a conceptual rather than structural level. One would involve attempting to disentangle the unhelpful notion that risk assessment and management is somehow the exclusive prerogative of the forensic services. This would carry the advantage of placing risk assessment back where it belongs, woven into the warp and weft of all routine clinical practice, rather than left to forensic 'risk gurus', whose very existence presently serves to deskill and demoralise other apparently less qualified workers. Why should the forensic imprimatur be a prerequisite for action given that the bulk of clinical risk 'management', and initiation of Mental Health Act detentions, happens in the community?
Whatever the uncertainty of the outcome of this debate, it is certainly time to retreat from the artificial boundaries that create barriers to good-quality care. Psychiatry is more intellectually challenging if various tasks are undertaken, and all psychiatrists should have a thorough understanding of risk, safety and security issues. But in the end it is expert diagnosis and treatment, rather than knowing how many fences to erect, that makes for a good psychiatrist. Of course, it takes a certain skill to manage newsworthy patients who have committed a particularly sensitive crime, but surely this calls for experience and consultant teamwork rather than a specialty?
It is not that long ago that consultant posts in the high secure hospitals were simply advertised as 'consultant psychiatrist'. The addition of the F-word came about in the hope that standards would be raised and in order to help develop regional services. But times have changed, the needs of individuals with severe mental illness have become prioritised, and it is now time for reintegration. As long as we continue to debate whether patients are that strange hobbledehoy 'forensic', or not, we put ourselves at the mercy of the government's agenda, as in the case of patients with dangerous and severe personality disorder (DSPD). Emphasis on similarities rather than differences will especially help us respond to proposals for new mental health laws. Do we wish to serve the Home Office or the Department of Health? Do we wish to be Home Office apparatchiks or Department of Health therapists? Ultimately, we need, as a profession, to get back together again in the interests of patients and the future of psychiatry. We could only feel better with less gatekeeping and more care and treatment.

Declaration of interest
None. It is indeed time for general and forensic psychiatrists to work together to improve services for mentally disordered offenders and others with similar problems. To achieve this, we must understand one another. Turner & Salter (2008, this issue) are unhappy with the definition of forensic psychiatry as 'patients and problems at the interface of law and psychiatry'. I consider this to be an accurate, pithy and practical definition. It establishes forensic psychiatry as the branch of psychiatry that deals specifically with mentally disordered offenders (patients at the interface of law and psychiatry) and that works alongside criminal justice agencies, including courts and prisons, to meet their needs. The authors are quite wrong in equating the development of forensic psychiatry with society's preoccupation with risk. Forensic psychiatric services were developed in the context of a liberal public policy tradition that seeks to divert mentally disordered offenders from criminal justice to health and social care. In 1990 the Home Office and Department of Health produced the widely quoted circular 66/90, which stated that public policy in regard to mentally disordered offenders is that they should receive their care and treatment within the National Health Service (NHS) rather than the penal system. This long-standing liberal tradition in English Law is reflected in Section 37 of the Mental Health Act 1983 (originating in the 1959 Act), which allows for offenders who have been found guilty of even the most serious violent offences to be dealt with by means of a disposal to healthcare rather than punishment in prison. This sets forensic psychiatric provision in the UK apart from other jurisdictions without such an enlightened and liberal attitude towards offender patients. The authors may, as citizens, object to the allocation of significant resources to the management of offenders, but as psychiatrists they should celebrate the commitment by society of resources to provide treatment for offenders with mental disorder in health settings rather than in a penal institution.
The authors seem to fall into the trap of minimising the correlation between violence, offending, substance