Abstract
Objective:
Although a minority of persons with schizophrenia (SCZ) commits violent acts, SCZ remains a risk factor for violence. Here, we present a broad overview of evidence-based treatments for violence in SCZ, including biological and psychosocial interventions.
Method:
We conducted MEDLINE and PsychINFO literature searches to retrieve articles relating to treatments for violent, hostile, or aggressive behaviours in SCZ.
Results:
Clozapine shows the strongest evidence for treating the acute violence of SCZ. Other atypical antipsychotics also possess antiaggressive effects, although the evidence is not as robust as that for clozapine. Psychosocial treatments can be useful adjuncts to pharmacotherapy once patients’ positive symptoms have stabilized. Cognitive behavioural therapy for psychosis and cognitive remediation are 2 psychosocial interventions that have demonstrated positive outcomes for violence in SCZ. Most psychosocial studies that examined violence as an outcome were conducted in forensic psychiatric settings.
Conclusions:
Effective treatments exist for persons with SCZ who pose a risk for violent and aggressive behaviour, although the overall evidence base remains relatively weak. More randomized controlled trials of programs showing evidence for reduction of violence in SCZ are required. Further research should delineate which patients could benefit from multimodal treatment and where and when such treatments are optimally delivered.
Although most people with schizophrenia (SCZ) are not violent, SCZ is still a risk factor for violence. 1 Epidemiological evidence confirms that men and women with SCZ are at elevated risk of conviction for violent offenses 2 –5 even after controlling for substance misuse and personality disorders. 6,7 The risk of committing homicide is even greater among persons with SCZ compared with the general population. 8 However, violence in SCZ frequently occurs in clinical settings, and much of this violence takes the form of physically assaultive behaviour directed at mental health professionals. 9,10
Effective treatment of violence and aggression in SCZ has the potential to lead to several positive outcomes. 11 First, it would decrease victimization of the perpetrators that is closely linked to their own violent behaviour. 1,12 –14 Second, it would result in reduced victim injury. Third, stigma against patients with SCZ and other psychiatric illness would diminish. 15,16 Fourth, the financial cost of incarcerating and hospitalizing offenders would decrease. Fifth, successful treatment could reduce crime rates. 17,18
At the outset, it should be noted that persons with SCZ who engage in aggressive behaviour have multiple problems, each of which needs to be addressed by different treatment components. 19 Although antipsychotic medication is necessary to treat acute violence driven by psychosis, it is insufficient on its own to prevent future violence. Nonetheless, medication compliance is imperative for patients to engage in other aspects of treatment to reduce the risk of violence over time and improve their quality of life.
The purpose of this article is to provide a broad overview of evidence-based treatments for violence in SCZ, with a focus on pharmacological and psychosocial interventions. These treatments are discussed in the context of phase of illness as some interventions can only be applied once acute psychotic symptoms have resolved.
Methods
We performed a MEDLINE and PsychINFO search. Combining the search terms schizophrenia or schizoaffective AND treatment AND violen* or aggress* yielded 1805 unique results. Articles were reviewed, and references were also scanned for other pertinent literature. Recent conference papers were also scrutinized. Relevant studies were selected based on predetermined inclusion criteria: publications with quantitative data, relatively recent publications (articles published from 1990 onward were considered with prioritization of more recent publications), violence as an outcome measure (e.g., rating scale items measuring violence or aggression, documented violent incidents, and charges or convictions for violent offenses), and SCZ and/or schizoaffective (SCZA) disorder comprising a significant proportion of the population under investigation. Exceptions were made to some of these inclusion criteria when the authors believed that readers would benefit from information that could inform the future of the field.
Publications were rated for the quality of evidence based on the Oxford Centre for Evidence-Based Medicine rating scheme. 20 Generally, level 1 evidence refers to systematic reviews of randomized trials; level 2 evidence implies randomized trials; level 3 evidence involves nonrandomized controlled cohort or follow-up studies; level 4 evidence refers to case-series, case-control, or historically controlled studies; and level 5 evidence includes mechanism-based reasoning. Levels were upgraded or downgraded based on specific strengths or weaknesses of methodologies; these adjustments are noted in the text.
Results
Biological Treatments
Pharmacotherapy
Pharmacotherapy, primarily with antipsychotic medications, is a first-line treatment for SCZ and the most appropriate choice for acute aggression. A complete discussion of the evidence linking treatment with pharmacologic agents to reduced aggression and hostility in SCZ is beyond the scope of this review. We focus, instead, on clozapine, for which there is substantial literature to support its use targeting aggression in SCZ.
Clozapine
Clozapine is a second-generation antipsychotic typically reserved for treatment-resistant SCZ that is distinguished from other agents in its class by its high affinity for dopamine 4 versus dopamine 2 receptors. This mechanism has been proposed as one reason for the demonstrated superior antiaggressive properties of clozapine over other antipsychotic agents. 21 Evidence further suggests that clozapine exerts its antiaggressive effects independent of its sedative or general antipsychotic properties. 22 –24 Level 4 retrospective studies of inpatients with SCZ report that treatment with clozapine is associated with reduced need for seclusion and restraint 25,26 and decreased aggressive behaviour. 27,28 At least 2 level 2 publications report analogous findings. Clozapine was shown in 1 randomized controlled trial (RCT) to reduce hostility among hospitalized patients with SCZ, and it also outperformed haloperidol on measures of clinician-rated aggressive behaviour. 24 In another RCT, clozapine treatment was associated with reduced frequency and severity of physical aggressiveness. 29 The findings from these robust studies are notable in that they demonstrated the superiority of clozapine over other second-generation and first-generation antipsychotics in reducing acute aggression. However, despite these benefits, clozapine may not be suitable for acutely aggressive patients who are unlikely to comply with the necessary bloodwork and metabolic monitoring.
Other Antipsychotics
Although evidence suggests that other second-generation antipsychotics can attenuate acute violence in SCZ, 30,31 it is unclear how much benefit this class offers over first-generation antipsychotics. One cross-sectional study of 331 patients with SCZ or SCZA, rated as level 3 evidence because of its larger sample size, reported an association between treatment with second-generation antipsychotics, but not first-generation agents, and reduced physical and verbal aggression. 32 The study’s main finding is consistent with an earlier level 3 prospective study that demonstrated a relative benefit of second- over first-generation antipsychotics in mitigating aggressive behaviour. 31 On the other hand, level 2 data from the Clinical Antipsychotic Trials of Intervention Effectiveness found no benefit of atypical antipsychotics (risperidone, olanzapine, ziprasidone, or quetiapine) over perphenazine in reducing aggression. 33 Notably, clozapine was not included in this study. On the other hand, other level 2 evidence has also shown that olanzapine has effects against hostility 34,35 and overt aggression, 29 which are superior to most other antipsychotics, save clozapine.
Medication Adherence and Depot Antipsychotics
Medication nonadherence is a well-documented risk factor for violence in SCZ. 36 A recent meta-analysis of 110 studies with 45,533 patients (87% with SCZ) found medication nonadherence to be one of several modifiable risk factors for violent behaviour. 37 This study was rated as level 2, as it included several methodologically weaker studies. Medication nonadherence may also increase violence risk independent of patients’ insight about their illness. 38 Level 2 evidence indicates that treatment with depot antipsychotic medications improves patient adherence, 39 although violence was not tested as an outcome measure in this study. However, a cohort study of 92,647 individuals (11% with SCZ) reported that patients prescribed depot antipsychotics had significantly fewer violent offenses. 40 This study was rated as level 4 evidence as it was an uncontrolled investigation. One single-blinded RCT of level 2 evidence found that SCZ patients with a history of violence who were treated with intramuscular zuclopenthixol had fewer violent episodes than patients taking oral zuclopenthixol. 41
In summary, clozapine shows the strongest evidence for amelioration of acute violence in schizophrenia. Less robust evidence indicates that second-generation antipsychotics are also efficacious treatments for episodic aggression. Although these treatments cannot prevent all future violence, medication compliance, perhaps reinforced through use of depot agents, is necessary to allow patients with SCZ to participate in psychosocial programs once acute psychotic symptoms have stabilized.
Brain Stimulation Approaches
Electroconvulsive Therapy
One landmark RCT demonstrated that electroconvulsive therapy (ECT) offers additional benefit to treatment-refractory patients with SCZ already taking clozapine. 42 The main outcome measures were global psychotic symptoms, which did not include violent behaviour. Level 1 meta-analytic data also indicate that ECT combined with antipsychotics can provide rapid global symptom improvement. 43 Few ECT studies in SCZ have examined violence as primary outcome measures, although some weaker studies, rated as level 4 evidence, report reduced aggressive incidents 44 and hostility scores. 45 Some experts recommend augmenting pharmacotherapy with ECT when rapid reduction of aggression is warranted. 46 Combined treatment with ECT and pharmacotherapy may be particularly indicated during the early stages of admission when patients present more acutely ill and are demonstrating high levels of aggression or violence as a result of their psychosis.
Noninvasive Brain Stimulation Techniques
Repetitive transcranial magnetic stimulation (rTMS) is a noninvasive method of electrically stimulating cortical neurons that has been shown in level 2 studies to ameliorate positive and negative symptoms of SCZ. 47 –49 However, to our knowledge, rTMS has not been tested as a treatment for aggressive behaviour in SCZ. Although there is no direct evidence connecting rTMS to violence reduction, it is possible that risk may be reduced in patients whose acute psychotic symptoms respond well to rTMS, given the relationship between violent behaviour and positive symptoms of SCZ. 50,51
To summarize, weak evidence links ECT treatment to improvement of violence in SCZ. As far as we are aware, no studies have directly investigated potential antiaggressive properties of rTMS in SCZ.
Psychosocial Treatments
Individuals with SCZ who have acted violently may receive treatment in general psychiatric settings, forensic hospitals, and correctional facilities. Overall, studies of forensic psychiatric services show the best outcomes in terms of violence reduction. 52 Below, we describe psychosocial treatment modalities that have been investigated for their potential to reduce violence in SCZ. These interventions assume more importance once acute symptoms of psychosis have been adequately treated and may be more effective for violence due to impulsivity or antisociality. 53
Group Psychoeducation
Psychoeducation involves teaching patients about the nature of their illness and available treatments, with the goal of improving insight and preventing relapse. Psychoeducation is commonly offered in a group format in forensic settings, and all studies reviewed below were embedded within forensic psychiatric programs. One observational study, rated as level 4 due to lack of a control group, reported that group psychoeducation was associated with improved knowledge, insight, and attitudes toward medication among patients with SCZ in a high-secure forensic hospital. 54 A level 2 RCT documented increased insight, knowledge, and self-esteem in a group of forensic patients with SCZ randomized to an 8-session psychoeducation group. 55 Notably, neither of these studies tested violence as an outcome measure, although many of the skills imparted (e.g., medication adherence) could theoretically protect against future violence.
Cognitive-Behavioural Therapy Interventions
Cognitive-behavioural therapy (CBT) focuses on addressing cognitive distortions that promote continuation of maladaptive attitudes and behaviours. Level 2 evidence indicates that CBT for psychosis (CBTp) can reduce the positive symptoms of psychosis 56 and has been applied for this purpose in forensic settings. 57 Weaker, level 4 studies of programs incorporating elements of CBTp for forensic SCZ inpatients reported improvement on clinician-rated measures of hostility 58 and aggression. 59
Another long-term CBT-informed inpatient program (the Service for Treatment and Abatement of Interpersonal Risk [STAIR] program) was evaluated on 145 patients with arrest histories (86% with a psychotic disorder). STAIR is a modification of a cognitive-skills program initially designed for nonpsychotic offenders. Participants were followed for 6 to 60 months following discharge from STAIR. Significantly fewer arrests, hospitalizations, and days institutionalized were noted posttreatment. 60 Although 13% of the participants had rearrests for violent or potentially violent offenses following program completion, baseline data for violent arrests were not available. This study, rated as level 4 evidence, did not include a control group.
On the other hand, 2 RCTs specifically examined the use of CBT-based psychotherapies in patients with SCZ and a history of violence. In the first study, 61 77 inpatients and outpatients were randomized to a modified CBTp program or social activation therapy (SAT). Follow-up at 6 and 12 months demonstrated the superiority of CBTp over SAT in decreasing aggressive incidents. The authors suggested that improvement in delusional thinking led to the reduced aggression in the CBTp group. This study, rated as level 2 evidence because of its robust design, is one of the few CBT studies that directly measured violence as an outcome. In a second study, 62 20 male inpatients were randomly assigned to receive metacognitive therapy, an offshoot of CBT aimed at increasing awareness of individual cognitive biases, or treatment as usual (TAU). The authors found a significant reduction in suspiciousness among those who had received the cognitive intervention, although violence was not evaluated.
Reasoning & Rehabilitation Programs
The observation that some offenders lack social, attitudinal, and cognitive-reasoning skills to engage in prosocial behaviours led to development of Reasoning & Rehabilitation (R&R) programs that aim to impart these skills and mitigate risk of recidivism. R&R programs are widely used in the criminal justice system and, more recently, have been studied for forensic patients.
Few studies have directly examined the effect of R&R on violent attitudes and offending. One investigation found that an R&R-like cognitive skills program resulted in fewer arrests for violent and nonviolent offenses among community-dwelling offenders with severe mental illness (70% with SCZ) involved in a court diversion program. 63 In addition, Young et al. 64 evaluated an R&R program adapted for violent offenders with serious mental illness and reported reduced violent attitudes and disruptive behaviours among program completers versus wait-list controls. Both of these studies featured high attrition rates and were rated as level 4 evidence.
Indirect measures of violence risk have been evaluated in studies with stronger designs. For example, one study gauged as level 3 evidence because of its controlled design compared R&R with TAU in violent offenders (72% with SCZ) and found that R&R conferred improvement in social problem-solving ability and coping responses. 65 This same research group conducted an RCT that randomized patients with a psychotic disorder and history of violence to R&R or TAU. 66 R&R program completers demonstrated improvement in social problem solving and evidenced decreased criminal attitudes, even at 12 months posttreatment. Since half of the 44 original participants did not complete the program, the study was assessed as level 3 evidence, despite its randomized design. Dropouts were associated with diagnoses of psychopathy or antisocial personality disorder and recent violence. As noted above, these 2 R&R studies did not assess violence or aggression as outcome measures.
Other Cognitive Remediation Programs
Neurocognitive ability, social cognition, alexithymia, emotion regulation capacity, and the therapeutic milieu of inpatient units may all influence aggression in SCZ. 67 Hence, interventions that target these factors could have the potential to reduce violent outcomes in SCZ. One recent level 2 RCT examined the effect of a cognitive remediation program on cognition in patients with SCZ or SCZA. 68 A high proportion (43 of 78 participants) had a history of violence. The remediation program was associated with significantly greater improvement on a number of neurocognitive measures in addition to reduced physical and verbal aggression.
Animal Assisted Therapy
Animal-assisted therapy (AAT) involves the use of specific animals in a goal-directed therapeutic process. 69 AAT has had a wide range of applications, including management of agitation and aggression in dementia. 70 A recent RCT evaluated the effect of AAT on reducing violent and “severely regressed” behaviours (defined as persistent social isolation and difficulty engaging in discharge-related programming) among 90 psychiatric inpatients, 75% of whom had SCZ or SCZA and 63% with a history of violence. 71 Participants were randomized to a canine-assisted intervention, equine-assisted intervention, social skills group, or TAU. Equine-assisted therapy alone was found to significantly reduce violent incidents for several months following treatment initiation. The authors noted that the low incidence of preintervention violence in the canine-assisted group may have masked positive effects. This study was scored as level 3 despite its randomized design because of this important group difference.
Addressing Comorbid Substance Use Disorders
Given that comorbid substance use disorders are associated with elevated risk of violence in SCZ, 5,72 –76 substance use treatment remains a priority for attenuating violent outcomes in patients with concurrent disorders. Substance misuse can increase violence risk via multiple pathways. 77 First, the direct pharmacological effects of substances, such as alcohol and stimulants, may facilitate emergence of aggression. 78 Second, substance use is linked to treatment nonadherence, which, as previously discussed, is independently associated with violent behaviour in SCZ. 79 Third, in a subpopulation of patients, maladaptive personality traits that onset in childhood may amplify substance use that contributes to violence risk. 78,80 There are several psychosocial treatment approaches for substance use disorders, including motivational interviewing, 12-step programs, CBT, family therapy, contingency management, individual counselling, and residential treatment programs. 81 One might expect that treatments that curb problematic substance use may also lead to violence risk reduction. However, very few studies have addressed this question. One level 2 RCT found that individuals with severe mental illness (20% with SCZ) and comorbid substance use disorders who received combined substance use and psychiatric treatment had fewer hospitalizations and arrests, although the issue of violence was not specifically addressed. 82
In summary, very few high-quality studies link delivery of psychosocial interventions to reduced violence or aggression in SCZ. On the other hand, most of the reviewed studies describe improvement in other aspects of functioning (e.g., neurocognition) or attitudes that could positively affect risk of violence. Most studies that examined violence as an outcome and reported positive results were conducted in forensic psychiatric settings. Among the modalities reviewed, CBTp and cognitive remediation showed the strongest evidence for reduction of violence in SCZ.
Discussion
This review provided an overview of the pharmacologic and psychosocial interventions for violence in SCZ. In Canada, as in many Western jurisdictions, persons with SCZ who commit violent offenses often attract significant media and political attention. 83 The relative dearth of studies, particularly RCTs, that have investigated potential treatments for violence in SCZ is thus surprising.
One challenge to studying treatments for violence in SCZ relates to heterogeneity in the definition of violence. For example, the studies reviewed measured violence with rating scales or defined it based on the emergence of physically or verbally aggressive incidents or manifestation of new charges. Lack of RCTs of forensic programs or specific program components targeting violent behaviour as a direct outcome is particularly problematic. Finally, meta-analyses of psychosocial interventions are necessary to parse out the active antiaggressive components of treatments that have an established evidence base.
Once effective treatments are identified, future studies should attempt to identify when is the optimal time to initiate treatment (e.g., before or following the first or multiple instances of violence), which patients require it, and the ideal place for treatment to occur (e.g., general inpatient wards, forensic psychiatric hospitals, outpatient clinics, residential facilities). 11 Determining which patients are most likely to benefit from specific treatments and/or multimodal strategies should be an additional priority. Aggressive patients who refuse treatment pose a particular challenge: harnessing the skills of a diverse and integrated treatment team, engaging patients’ families, and emphasizing individual strengths could facilitate the therapeutic alliance and discovery of violence reduction as a mutual goal. Given evidence that aggressive behaviour may be equally prevalent among general and forensic inpatients with severe mental illness, 1 the treatments described in this review are likely applicable to patients in both civil and forensic settings. Ultimately, the effective delivery of evidence-based treatment for violence in SCZ should improve the health outcomes of patients, their families, health care providers, and the public.
Footnotes
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.
