Talking to Forensic Patients About Risk

Why Is SDM Problematic in Forensic Psychiatry?

Despite the optimism of the NHS England policy documents, within the bioethics literature it has long been recognized that respect for patient autonomy and choice is more complicated in mental health than in many physical health contexts. SDM in mental health is sometimes challenging because of the potential for mental disorders to undermine the capacity needed to make complex decisions; these include decisions about the nature of reality, other peoples' identities, the assessment of danger, and how to weigh options and choices. Good-quality research has shown that both mental and physical disorders can impair the capacity to make complex decisions, such as the decision to accept or refuse treatment or the decision to participate in altruistic activities such as research.

However, the best evidence to date suggests that mental conditions do not always impair decision-making capacity, and when they do, the impairment may be temporary and may be decision-specific. For example, a person with psychosis may lack capacity to refuse treatment for mental disorder but may still have capacity to consent to surgical treatment or make a good-quality decision about where to live. The recovery movement in mental health has done much to support people living with mental health problems in making the decisions that they can make by challenging both professional and family assumptions about mental disorder and incapacity.

Much mental health law exists to stand in for the autonomy of psychiatric patients who are incapacitated by mental disorder. The law can protect people who lack capacity from interference or exploitation by others; at the same time, it may ensure that their lack of capacity to exercise autonomy does not cause harm to the patient or other people. This is a complex balance ethically, and the legal solutions are imperfect and continue to be in development, both in court and in jurisprudence. In general, psychiatrists have a duty to obtain informed consent from their patients before treating them, just as in general medicine and surgery. Only in very particular circumstances will the law allow psychiatrists to impose treatment on people without their consent or in the face of a flat refusal.

Within forensic mental health, there are additional reasons that SDM might be problematic between patients and their doctors. First, forensic patients are both people who break the law and people who live with mental disorders, whose values and choices have often caused great suffering, fear, and harm to the communities from which they come.⁷ Their identities and values as both psychiatric patients and offenders bring them into conflict with the values of the wider society represented by the criminal and civil law; these values are represented by the forensic services that both detain and care for them. There is a sense in which forensic services have to stand for and impose the pro-social values that offender patients have challenged by their behavior.

Second, it cannot be assumed that patients' antisocial values are a symptom of mental illness that will go away when treated. Decades of research on the relationship between violence and mental disorder suggest that symptoms of mental disorders are only one of many risk factors for violence and are arguably made worse by substance abuse and antisocial attitudes.⁸ Forensic psychiatrists who offer treatment to patients may find that the patients still have their antisocial values and beliefs at the end of treatment and do not want to give them up, even if they do increase risk. SDM requires practitioners to have a dialogue with patients that respects the patients' values, but this seems hard to justify in settings that are committed to reducing risk.

Third, it is a challenge to have genuinely shared collaboration and conversation about treatment decisions in coercive settings where autonomy is restricted in the name of security. Forensic patients know that they stand to benefit in terms of access to leave and ultimately freedom if they comply with the treatment and rehabilitation regime advised for them and agree to the formulation of risk that the team asserts explains their violence risk. It is not clear that forensic patients are allowed to hold formulations different from their clinical teams; even if they are behaviorally settled on a unit, continuing dissent from the risk formulation is likely to be perceived as risky in itself. It is hard not to think that patients are tacitly encouraged, if not actively pressured, to express views about risk and rehabilitation that fit with the professional view.

This brings us to the fourth problem, the formulation of risk itself, which is by no means a simple or static process. There are real ethics questions about how a risk formulation is made in a way that is just in terms of balance, objectivity, and acceptance of uncertainty. For criminally sentenced patients, the formulation process usually begins in the trial stage and is first framed at conviction, when the mental disorder is determined to be either functionally linked to the offense or an explanation for it. Many forensic patients will do what they are told by their defense team at this point, and their view of the offense and their part in it is not usually given much weight unless it goes to a positive verdict in terms of outcome. I am thinking here of many patients I have met who strongly disagreed with the idea that they were mentally ill when they committed their act of violence, despite apparently complying with their lawyers' advice.

Once in mental health settings, the responsible clinical team now develops their own formulation of how the offense took place, based not only on what the court heard and found but on what they see and hear during the treatment process. Relationships with fellow patients and staff, as well as behavior on the ward, are likely to be used in reviews of formulation; and there may be additional information in terms of childhood adversity and family history that were not in evidence at court but are seen as clinically relevant to formulation (in terms of structured risk assessments like the Historical Clinical Risk Management-20, for example). Again, in my experience, the patient's own formulation of risk may be quite different; I am thinking here of an individual who was undoubtedly psychotic at the time of his offense, a psychosis exacerbated by substance misuse. He did not accept that his violence was driven by psychosis; rather, he believed that he was acting in self-defense against dangerous drug dealers. He accepted that he had acted illegally and riskily, but he did not accept that mental illness played a significant part. The clinical team took the view that his risk to others would be reduced by his taking appropriate medication. However, he thought his risk would be reduced by living somewhere different and finding more reliable people from whom to obtain drugs.

Finally, there may be revisions to the formulation that draw on material from therapy sessions or information from family members. Both therapists and family members are assumed to be reliable informants, but they may have a very particular perspective on risk, depending on their experience, knowledge, and personal reactions to the patient. I am thinking here of one patient who spent many decades in psychiatric hospital for a sexual offense that would usually warrant a sentence of less than 10 years (rightly or wrongly). His risk of sexual offense to others was constantly reformulated based on the new therapists who saw him for sexual risk-assessment interviews, and there was a mass of conflicting views about his risk of further violence. His family were strongly opposed to his leaving secure care and asserted repeatedly that he would offend again if he were not detained and constrained; their anxiety was partly driven by concern that he would not be as well looked after in a shared home or rehabilitation hostel. The patient's view was that he wanted a chance to live in the community; he completely accepted all the clinical formulations and was keen to take medication to abolish all and any sexual desire if he could live in the community and have a little more autonomy. However, some members of the clinical team thought this acceptance itself was a kind of false compliance that might indicate a manipulative nature and over-dependence on staff. In this kind of scenario, the patient's view is seen as unreliable and therefore not worthy to be part of a risk assessment.

I suggest that risk-related SDM is complex because forensic professionals find it hard to trust patients' accounts of themselves, often seeing them as generally untrustworthy and also because being suspicious of patients' motives and values is seen as being tough, realistic, or apparently objective. Forensic professionals must struggle with the compound organizational task of helping patients recover good mental health and well-being while simultaneously taking steps to ensure that patients reduce the risk they pose to others. The professional values that forensic professionals have to pursue can clash in many ways, some obvious, some subtle. There is little guidance about which values should dominate in a conflict of values in which the value of reduced security is weighed against the value of harm prevention and public safety. Collaboration with patients about their risk is likely to end when patients and professionals fundamentally disagree; unlike in general medicine, it is not possible to let forensic patients make decisions that could have bad outcomes for themselves and other people.

How Can We Improve SDM in Forensic Settings?

I think it is not too strong to say that forensic services have traditionally been tasked with making patients change their minds and values, as well as changing their behavior. However, changing minds is harder than controlling behavior. Forensic professionals have always been uncomfortably aware that one can use environments and rules to enforce behavioral compliance, but this may be no guarantee of a persistent change of values and moral perspectives.

The application of the language of recovery to mental health services has been helpful to forensic services in several ways. First, the idea of living with a condition that affects your identity neatly mirrors the idea of having to live with an identity as an offender; this idea also grounds an understanding of stigma and its reality in communities that socially exclude mentally abnormal offenders. Second, the recovery literature encourages attention to personal narratives and the way that people tell stories about their experiences, especially those that are painful, challenging, and chronic.

Third, and most important in the context of this commentary, a recovery perspective encourages patients to pay attention to the reality of their difficulties and ownership of responsibility for the process of change. This aspect is elegantly demonstrated in work by Maruna⁹ on desistance in delinquency, who used a narrative approach to explore what makes people give up offending. Maruna found that those who desisted from offending were more likely to use the language of personal agency and responsibility, accepting (it seems) that behavioral change starts with a transformation of values and ownership of personal choices, past and future. Maruna's findings have been replicated in other groups and are mirrored in Ward's Good Lives Model of offender rehabilitation,¹⁰ where again there is emphasis on the values and choices that are important to living a life without violence. There is also an intriguing resonance with the literature on the linguistic expression of attachment security, whereby securely attached people use language that demonstrates a fresh and autonomous sense of self, as well as a comfort with vulnerability and need.¹¹

A forensic service that only made patients feel better in themselves (as a traditional mental health service does) would arguably be of little value in terms of public interest and social capital; forensic services need to be seen to help patients behave better, not just feel better.¹² The task of helping forensic patients behave better, however, entails getting them to agree that our prosocial values are better than their antisocial ones, and this task involves building relationships where values, choices, and attitudes can be explored in depth. This is a building block of what we call relational security,¹³ and if this is in place, SDM will be a natural part and process. Without therapeutic relationships that take forensic patients' values seriously and allow exploration of how they differ from the norm, there is a danger that SDM will be no more than a cover for patient compliance with the values of forensic professionals, a cover that makes forensic professionals feel good about themselves and reassures them that they are not coercing the patients into a prosocial world view.

There is an irony here, however, in that some degree of coercion is inevitable in forensic services. Forensic patients cannot consent to be detained by courts, and they, at least implicitly, do not agree to be treated. There is a paradoxical complexity to the life of the forensic patient; offenders who say that they want to have treatment for their mental disorder are often viewed with skepticism and would rarely be offered a place, even if voluntary treatment were possible. Further, there is a profound philosophical assumption that those who commit offenses, whether mentally ill or not, are people who cannot be trusted, who do not share the values of honesty, integrity, and respect for others. The situation is even more complicated for forensic patients who are detained because they repeatedly assault fellow patients and staff; they may not have antisocial values and attitudes, but their mental illness is so treatment-resistant that it makes them insightless and paranoid. In such circumstances, their condition may make it difficult to establish the kind of therapeutic relationship needed for SDM about risk. To reiterate the problem in a different way, you need some shared vision of recovery and desistance from the antisocial life before you can have SDM about risk.

The Language of Risk and Recovery

I argue that shared decision-making must involve a rich and nuanced dialogue between two people. In forensic psychiatry, this will be a dialogue about prosocial and antisocial values and choices, and what it might cost a patient to give those up. It might be counter-argued that, in relation to risk assessment, it is hard for clinicians to give weight and attention to the patient's values and views. For example, this review cites a study using the DUNDRUM tools and quotes the authors' view that “concordance between clinician and patient ratings could be a useful index of the degree to which understanding of risk has become shared” (Ref. 1, p 6). In terms of dialogue, however, we might also wonder whether concordance implies compliance with a narrative that is imposed by the clinical team, i.e., a narrative that may or may not be accurate and may or may not fit with the patient's narrative.

There remain profound questions about what a shared understanding of risk means and what scope there is for patients to dissent or have a different narrative. To reach a shared understanding, we might first need a shared language of offending and what it means. I am thinking here of ideas from psychodynamic psychiatry that actions are communications, especially for those who lack language. Forensic professionals who want to develop a shared understanding of risk with patients will need to pay attention to the language of risk; they will have to check whether their language is comprehensible to the patient, and whether they can understand the patient's language of offending. It may be helpful here to consider the literature on offender narratives and how offenders may use language about their offense that tends to reduce or neutralize a sense of guilt or shame.¹⁴ Lamb¹⁵ has written thoughtfully about how the use of language in professional reports about offenders subtly affects perceptions of agency; for example, agency can be communicated by the use of the active voice (He broke her jaw.) as opposed to the passive voice (Her jaw was broken.). I take the view, drawing on Lamb's arguments, that forensic professionals have a duty to use the language of agency with their patients, and to emphasize the offender's agency and choices in relation to the offending. Not only is this stance respectful of patient autonomy, but it conveys the possibility that, if bad choices were made in the past, better choices could be made in the future. In contrast, if we use passive language (such as “the offense took place,” or “when the index offense occurred,” or “when the victim died”), it promotes a linguistic sense of passive helplessness in the face of offending that keeps people stuck in both their offender identity and their mental illness identity.

I suspect that if SDM around risk is going to have utility and be respectful of the values of both patients and professionals, then it needs to start with developing and refining formulations of risk through an examination of narrative. This is a process that Griffith and Baranoski¹⁶ have described as valuable in different ways within the criminal court process. The difference in the clinical context is that narratives about past crimes and cruelties that inform risk assessment may be understood by victims, perpetrators, and justice agents in different ways at different times.¹⁶ These differences of perspective will be crucial to developing a shared narrative of risk, which includes areas of “dis-sensus” as well as consensus.¹⁷

It might be said that such conversations could be painful and distressing, and professionals may be rightly concerned about their patients experiencing feelings of guilt and shame because of these discussions. However, the cognitive therapy literature tells us that avoidance of negative feelings only entrenches them; it is the job of the forensic clinical team to help patients manage such feelings, not avoid them at all cost. On that point, of course, it is worth remembering that many violent acts represent an attempt to escape from horrible feelings by acting them out in someone else's physical space, so helping offenders mentalize uncomfortable feelings and understand them as part of human experience may be crucial to their recovery.