CHAPTER 23 Rick Fraser1, Rosemary Purcell2 and Danny Sullivan2,3 1 Early Intervention in Psychosis Service, Sussex Partnership NHS Foundation Trust, Sussex, UK 2 Centre for Forensic Behavioural Science, Swinburne University of Technology, Melbourne, Victoria, Australia 3 Victorian Institute of Forensic Mental Health, Clifton Hill, Victoria, Australia There is a well-established relationship between experiencing severe mental illness, particularly psychotic disorders such as schizophrenia, and increased rates of violence and criminal offending. Despite this association, it is important to recognise that only a minority of those with mental illness will offend as a result of their mental illness and that the vast majority of individuals with mental illness do not engage in criminal offending. Furthermore, the excess incidence of violent crime associated with mental illness is not observed across all diagnoses, but occurs within particular categories and/or circumstances that increasingly can be specified. Because of the potential to prevent or ameliorate this offending in those with mental illness, this is a vital issue to address; however, it is equally important to be mindful of the risk of stigmatising those with mental illness as ‘dangerous’ (or other pejorative terms) when the overwhelming majority will never engage in violence or offending. Violence in those with mental illness represents a significant public health issue, given not only the substantial personal costs to the victim and patient/perpetrator, but the economic costs conferred via health services for treating physical injuries, mental health services for treating emotional harm suffered, lost productivity if the victim is unable to work due to injury or death and the involvement of the criminal justice and forensic hospital systems in cases of serious violence. In the United Kingdom, the estimated economic cost per homicide by a mentally ill offender has been calculated to be £1.72 million, equivalent to AUD $2.76M [1, 2]. A preventative or early intervention framework for reducing violence among the mentally ill is warranted and would likely be highly cost-effective. However, clinical service models for putative interventions are limited. While specialist forensic mental health services are available in some mental health catchments, these services typically become involved only after a patient has offended. This chapter will canvass the rationale for early intervention to reduce violence and criminal offending among the mentally ill, with an emphasis on younger patients, since the opportunities for prevention and early intervention are likely to have the greatest impact with this population. We will focus on interventions that have been used to reduce these ‘forensic’ outcomes and their effectiveness, including within special clinical populations prone to other problem behaviours or offending. Finally, service provision issues, including the challenge of capacity building in this field, will be reviewed, along with potential research and service reform initiatives that can further advance the critically (and chronically) overlooked agenda of early intervention in forensic mental health. The overwhelming majority of people who experience mental disorders are never violent. Nonetheless, there is a robust, well-established association between experiencing a major mental illness, particularly a psychotic or severe mood disorder and increased rates of violence and criminal offending [3–5]. Evidence of this relationship has been derived from three distinct research paradigms: (i) studies of violence and/or offending among individuals with a mental illness [e.g. 6, 7]; (ii) studies of mental illness among known offenders, such as prisoners [e.g. 8, 9] and (iii) epidemiological studies of offending and mental illness in community samples [5, 10]. In a landmark, methodologically rigorous Australian study, Wallace et al. [5] used a register which recorded all contacts with public mental health services in the state of Victoria to establish the prior psychiatric histories of over 4000 individuals convicted in the higher courts of serious offences. The results demonstrated that, compared to a general population sample matched for age, gender and area of residence (a socio-economic proxy measure), individuals with schizophrenia were 3 times more likely to be convicted of a sexual offence, 4 times more likely to be convicted of a violent personal assault, and 10 times more likely to have been convicted of homicide. These elevated rates of violence were similarly observed in patients with severe affective disorders (predominantly bipolar disorder and major depression). Analysis indicated that the rates of offending increased substantially when co-morbid substance use was included, although highly significant associations remained for all forms of offending in the absence of substance use. Using data from 20 discrete studies (n = 18, 423), Fazel et al. [7] conducted a meta-analysis which demonstrated that the level of association for general violence was 4–5 times greater in patients with psychosis compared to the general population and between 14–25 times higher for homicide. Consistent with earlier studies, the relationship between violence and psychosis was mediated in part by co-morbid substance abuse. There were no differences in the rates of violence between patients with schizophrenia versus other forms of psychotic illness (e.g. schizoaffective disorder, schizophreniform disorder, delusional disorder, psychotic disorder not otherwise specified) or between the study period or study location (including Scandinavia, the United States, the United Kingdom and Australia). This finding attests to both the consistency over time and across communities of the strong association1 between psychosis and offending. The evidence of a link between psychosis, violence and offending is compelling, and argued by one expert in the field to be similar in magnitude to the association between smoking and lung cancer [4]. What is less clear is how these constructs are related, despite considerable research in this regard. Conceptually, there are at least three possible mechanisms by which these constructs might be related [11]. The first (Pathway 1) asserts that violence emerges as a function of the symptoms of psychosis experienced by individuals with an otherwise unremarkable and ‘unblemished’ background. Within this pathway, there are three putative categories of psychotic symptoms which may play a role. Psychotic symptoms typically cluster into three domains: (i) positive symptoms, which are characterised by the presence of delusions and hallucinations; (ii) negative symptoms, which are characterised by blunted affect, poverty of speech, amotivation, anhedonia and social withdrawal and (iii) disorganized symptoms, which combine both positive and negative symptoms, but are characterised by thought disorder, disorientation, confusion and cognitive deficits. The three-domain model of psychosis has been theoretically proposed to account for illness-related violence [12], although the bulk of empirical research has focused only on the relationship with positive symptoms. Evidence from clinical, prison and community-based studies support the contention that violence is significantly related to the patient’s positive psychopathology [see 13–17]. This is particularly related to the experience of command auditory hallucinations, whereby ‘voices’ instruct the patient to harm another person; and persecutory delusions, in which the patient falsely believes that another person intends them harm. One major community study however failed to find a significant association between the patient’s symptomatology and violence [18]. Negative and disorganised symptoms have also been argued to play a role in the emergence of violence via their interference with goal-directed behaviour, logical thinking and cognition. For example, it has been suggested that these symptoms may frustrate patients and increase the likelihood that they act violently in response to managing interpersonal interactions and conflicts [11, 12], however there is a lack of empirical data to support this contention. The second pathway (Pathway 2) posits that behavioural and conduct difficulties, as well as offending behaviours, are evident before the frank emergence of psychosis or are apparent at first contact with mental health services [e.g. 19–22], and that personality disturbance and substance abuse among affected patients are influential in the expression of violence [23–25]. For example, in a study of 205 inpatients with severe mental illness, a greater number of conduct disorder symptoms prior to 15 years was found to be significantly associated with increased risk of serious assaults over the lifespan, aggressive behaviour in the past 6 months and violent crime after controlling for alcohol and illicit drug use [22]. Finally, the third pathway (Pathway 3) posits that violence may be a simple correlate of psychosis [26]. That is, violence and psychosis share a statistical relationship through their links with other mediating variables, such as younger age, poverty, low socio-economic status, or co-morbid substance abuse. If this were the case, no clear temporal relationship would exist between violence and psychosis, after controlling for confounding factors. The lack of consensus regarding the temporal relationship between psychosis and offending mainly reflects disparities, and in some cases major limitations, in study methodologies. This is not to say that each route to violence may not operate, but that it cannot be discerned on the basis of the extant research which pathway is particularly influential for which patients. Research limitations include sampling (e.g. Pathway 1 studies that include a high proportion of non-psychotic patients who are violent – such as prisoners or general community samples – are unlikely to find associations with the positive symptoms of psychosis); the assessment of violence (relying on self-reported violence being the most obvious weakness) and the methods of assessing pre-morbid behavioural or conduct difficulties (which in most ‘Pathway 2’ studies has been retrospective – often 2 to 3 decades after the events – and subject to biases in recall or reporting). Furthermore, studies reporting the existence of prior offending at ‘first contact with mental health services’ have typically taken ‘first contact’ to mean the onset of psychosis. However, psychotic disorders are often episodic and may have an insidious onset, such that the person may experience months or years of deteriorating functioning before the onset of acute positive symptoms that usually precipitate help-seeking and psychiatric treatment. It may be that the observed offending in these studies in fact occurred during the prolonged period of illness onset and does reflect an illness-related outcome, rather than reflecting a general disposition to antisocial conduct. The latter point is salient, particularly as a more thorough analysis of the literature on violence among the mentally ill indicates that a significant proportion of offending occurs during the first episode of psychosis. Several studies and reports focusing on homicide as the outcome have found that between 38 [27] and 61% [28] of individuals were experiencing their first episode of psychosis at the time they committed the offence. These findings were subsequently confirmed by a systematic review and meta-analysis, which estimated the rate of homicide during the first episode of psychosis to be approximately 15 times higher compared to the rate of homicide after the initiation of treatment [29]. Another systematic review by these authors also demonstrated a significant association between the duration of untreated psychosis (DUP) and homicide, such that patients who experienced a longer period of untreated illness were more likely to have killed [30]. There is a growing literature to suggest that a significant minority of patients experiencing their first episode of psychosis demonstrate aggression and violent behaviour prior to first psychiatric admission [20%; 31] or at first presentation to mental health services [40%; 32]. In light of the research demonstrating the relationship between severe mental disorder and violence and offending, particularly in those with emerging and first-episode psychosis, it has been suggested that early intervention may be critical to preventing or reducing these outcomes among the mentally ill, and thereby ultimately saving lives [33, 34]. Research indicates that there is an association between traumatic experiences and later offending [35]. Victims of violence are at increased risk of mental health problems including depression, dissociation and PTSD, as well as substance use and offending behaviours. A study by Widom [36] demonstrated a link between early childhood trauma/abuse and antisocial behaviour. These youngsters were at greater risk of later involvement with the criminal justice system. Retrospective studies looking at criminal populations also show this association between early trauma and later offending behaviour [37]. A German study of 54 prison inmates found a lifetime prevalence of 36% for PTSD and a point prevalence of 17% [38]. There is a clear association between traumatic experiences and offending behaviours, with mental illness and/or substance abuse as links in this pathway. Early identification and treatment of these disorders in those who have been traumatised may decrease risk of future offending and violence. While there has in the past been some ambivalence about diagnosing personality disorder (PD) in young people, in recent years it has become increasingly clear that these constructs exhibit moderate–high stability over time, and that personality difficulties and vulnerabilities may be apparent well before the firm diagnosis of PD is made [39]. In particular, childhood conduct disorder is strongly correlated with adult antisocial or dissocial PD [40]. When there are concerning personality traits which may be associated with offending, early intervention could putatively reduce the rate of transition to adult PD, or might attenuate its consequences [41]. A separate debate about the pejorative effects of labelling young people with PD diagnoses may in fact reflect clinicians’ (and researchers’) own negative attitudes to PD. Regardless, a reluctance to make an appropriate diagnosis may reduce access to clinical services and thus be detrimental to young people who exhibit personality vulnerabilities or difficulties. Young people displaying marked impulsivity and aggression are also at increased risk of delinquent behaviours and a progression to adult offending. Not only are there genetic [42] and familial [43] components to these traits and to cluster B PDs, but both the traits and related PDs associated with their early life manifestations are strongly correlated to adult offending [44].
Early Intervention to Reduce Violence and Offending Outcomes in Young People with Mental Disorders
Introduction
The rationale for early intervention to reduce violence and offending among people with mental disorders
Populations at risk for offending
Psychosis
Post-traumatic stress disorder (PTSD)
Personality disorder
Related posts:
Perinatal Preventive Interventions in Psychiatry: A Clinical Perspective
How Early Intervention Can Turn Things Upside Down and Turn a Patient Into a Psychiatrist
Transiting Out of Child and Adolescent Mental Health Services – Influences on Continuities and Discontinuities in Mental Health Care
Early Intervention for Delirium
Early Intervention for Self-Harm and Suicidality
Early Intervention in Older Adults – A Focus on Alzheimer’s Dementia
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
