Hostname: page-component-6bf8c574d5-h6jzd Total loading time: 0.001 Render date: 2025-02-23T06:53:07.578Z Has data issue: false hasContentIssue false

A systematic review and meta-analysis of the effect of community treatment orders on aggression or criminal behaviour in people with a mental illness

Published online by Cambridge University Press:  20 February 2025

S. Kisely*
Affiliation:
The University of Queensland School of Medicine, Brisbane, QLD, Australia Metro South Addiction and Mental Health Services, Metro South Health Service, Woolloongabba, QLD, Australia Griffith Criminology Institute (GCI), Griffith University, Brisbane, QLD, Australia Departments of Psychiatry, Community Health and Epidemiology, Dalhousie University, Halifax, NS, Canada
C. Bull
Affiliation:
The University of Queensland School of Medicine, Brisbane, QLD, Australia Queensland Centre for Mental Health Research, The University of Queensland, Brisbane, QLD, Australia
N. Gill
Affiliation:
School of Medicine and Dentistry, Griffith University, Brisbane, QLD, Australia Mental Health Policy Unit, Health Research Institute, University of Canberra, Canberra, ACT, Australia Mental Health and Specialist Services, Gold Coast Health, Southport, QLD, Australia
*
Corresponding author: S. Kisely; Email: s.kisely@uq.edu.au
Rights & Permissions [Opens in a new window]

Abstract

Aims

There has been concern about violent acts and other criminal behaviour by people with a possible history of mental health problems. We therefore assessed the effects of community treatment orders (CTOs) on self-, third-party-, and agency-reported criminal behaviour when compared to voluntary treatment.

Methods

A systematic search of PubMed/Medline, Embase, PsycINFO and criminal justice bibliographic databases for observational or randomised controlled trials (RCTs) comparing CTO cases with controls receiving voluntary psychiatric treatment. Relevant outcomes were reports of violence and aggression or contacts with the criminal justice system such as arrests and court appearances.

Results

Thirteen papers from 11 studies met inclusion criteria. Nine papers came from the United States and four from Australia. Two papers were of RCTs. Results for all outcomes were non-significant, the effect size declining as study design improved from non-randomised data on self-reported criminal behaviour, through third party criminal justice records and finally to RCTs. Similarly, there was no significant finding in the subgroup analysis of serious criminal behaviour.

Conclusions

On the limited available evidence, CTOs may not address aggression or criminal behaviour in people with mental illness. This is possibly because the risk of violence is increased by comorbid or nonclinical variables, which are beyond the scope of CTOs. These include substance use, a history of victimisation or maltreatment, and the wider environment. The management of risk should therefore focus on the whole person and their community through social and public health interventions, not solely legislative control.

Type
Original Article
Creative Commons
Creative Common License - CCCreative Common License - BYCreative Common License - NCCreative Common License - ND
This is an Open Access article, distributed under the terms of the Creative Commons Attribution-NonCommercial-NoDerivatives licence (http://creativecommons.org/licenses/by-nc-nd/4.0), which permits non-commercial re-use, distribution, and reproduction in any medium, provided that no alterations are made and the original article is properly cited. The written permission of Cambridge University Press must be obtained prior to any commercial use and/or adaptation of the article.
Copyright
© The Author(s), 2025. Published by Cambridge University Press.

Introduction

Community treatment orders (CTOs) allow compulsory treatment in the community for people with mental illness. They are an example of civil commitment, as opposed to forensic orders following criminal proceedings. CTO use in Australia and New Zealand is high by international standards although rates vary considerably within both countries (Light et al., Reference Light, Robertson, Boyce, Carney, Rosen, Cleary, Hunt, O’Connor, Ryan and Kerridge2017; O’Brien, Reference O’Brien2014). Concerns have been raised regarding ethical and human rights implications of the CTOs (Brophy et al., Reference Brophy, Edan, Kisely, Lawn, Light, Maylea, Newton-Howes, Ryan, Weller and Zirnsak2021).

One of the other main criteria for CTO placement is the imminent danger of harm to self or others. Despite this, most of the relevant systematic reviews and meta-analyses on CTOs concentrate on health outcomes and do not consider violent acts or other criminal behaviour. There have been just three systematic reviews that have considered criminal behaviour or aggression in comparisons of CTO cases with voluntary controls, generally as a secondary outcome. The first was a Cochrane Systematic Review that was restricted to just three randomised controlled trials (RCTs) (Kisely et al., Reference Kisely, Campbell and O’Reilly2017). Because of the challenges of conducting RCTs in this area, it is unlikely that the participants were representative of people on CTOs, especially as dangerousness is generally an exclusion for entry into an RCT. The two other systematic reviews considered observational trials, the results of which may be more generalisable. However, one was published 17 years ago and the literature search in a second completed in 2019 (Churchill et al., Reference Churchill, Owen, Singh and Hotopf2007; Segal, Reference Segal2020). As criminal behaviour was not the primary outcome, neither review attempted to meta-analyse the data nor extend their search to bibliographic databases specific to criminology.The two reviews came to very different conclusions. The earlier one concluded that CTOs were not associated with significant changes in criminal behaviour or aggression and therefore questioned their utility. By contrast, the more recent review reported significant reductions in both violence and crime in mirror image and controlled studies, concluding that these results were evidence of the benefits of CTOs. However, this review was limited by methodological concerns such as not following Preferred Reporting Items for Systematic Reviews and Meta-Analyses guidelines (PRISMA) (Kisely et al., Reference Kisely, Zirnsak, Corderoy, Ryan and Brophy2024; Moher et al., Reference Moher, Liberati, Tetzlaff and Altman2009). There may also have been further developments in the area. Given the continuing uncertainty, we therefore assessed the possible effects of CTOs on self-, third-party-, and agency-reported criminal behaviour when compared to voluntary treatment.

Method

Search strategy

The protocol for this systematic review was registered with PROSPERO (CRD42020155996). We followed PRISMA guidelines (Moher et al., Reference Moher, Liberati, Tetzlaff and Altman2009). Given the limitations of the most recent systematic review, we updated the search by Churchill et al. (Reference Churchill, Owen, Singh and Hotopf2007) of PubMed/Medline, Embase and PsycINFO (2005 onwards). Following consultation with a professional librarian, we also searched Criminal Justice Abstracts, the Australian Criminology Database and ProQuest from inception onwards. All the searches were completed in June 2024 and restricted to peer-reviewed papers published in English. Table 1 shows our search terms that were finalised with the help of the same professional librarian. Ethical approval was not required for this systematic review, as all included primary data had previously been published.

Table 1. Search terms

Two of the authors (SK and CB) independently screened records and abstracts. A third reviewer (NG) was consulted in the case of any disagreement and consensus was achieved in all cases. The reference lists of selected retrieved papers were screened to identify additional studies that met inclusion criteria.

Inclusion criteria

We included any of the following study designs that compared people on CTOs for severe mental illness with contemporaneous controls receiving voluntary psychiatric treatment: RCTs, cohort, case control and cross-sectional studies.

Exclusion criteria

We excluded studies of inpatient treatment (including extended leave), forensic orders and compulsory treatment in the community for drug or alcohol related disorders, as well as those that did not have controls receiving voluntary psychiatric treatment.

Outcomes

We extracted data for the following outcomes: 1) violence or aggression measured, where possible, with standardised instruments; 2) contacts with the criminal justice system such as arrests or court appearances following CTO placement. Outcomes could be self-, third-party-, or agency-reported criminal behaviour. We focused on outcomes at one year as this is the most common endpoint in the literature and the impact of an intervention on health service beyond one year is difficult to ascertain (Kisely et al., Reference Kisely, Campbell and O’Reilly2017). However, we also assessed for the presence of these events up to 10 years from entry into the study.

Study quality

All studies identified for inclusion were cohort studies and RCTs. Two of the authors (SK and CB) independently assessed quality using the appropriate Joanna Briggs Institute tool for each. Both cover the following areas: selection of the study groups in terms of case definition, representativeness and similarity of controls; comparability of the groups such as the use of matching or multivariate techniques, and measurement of exposure and outcomes in a valid and reliable way. This included consideration of allocation concealment and blinding in the case of RCTs. The version for cohort studies has 11 items and the one for RCTs, thirteen. In the case of RCTs and quasi-experimental studies, we only appraised the primary outcome as it related to the aim of our review.

Similarly, we used the Grading of Recommendations Assessment, Development and Evaluation (GRADE) framework to assess the overall certainty of evidence from very low to high. Outcomes from RCTs were initially graded as high and those from observational studies as low. These were then graded up or down depending on study quality, inconsistency, indirectness, imprecision and publication bias.

Analysis

Where data were available for two or more studies, they were combined in a meta-analysis giving preference to data that had been adjusted for potential confounders. We used log odds ratios (ORs) to pool dichotomous and continuous data using inverse variance. We used an I 2 statistic value of greater than 50% as an indicator of significant heterogeneity. We explored any heterogeneity further through sensitivity analyses of the effect of omitting each study in turn. The random effects model was used for all the analyses because we could not definitively exclude between-study variation even in the absence of statistical heterogeneity. We ran separate analyses for criminal justice contacts and reports of violence or aggression. We did not combine analyses from non-randomised designs and RCTs. If recorded, we undertook subgroup analyses of criminal behaviour coded as being severe. Finally, we undertook sensitivity analyses of the effect of excluding studies of lower quality such as those that only reported unadjusted or selectively chosen results.

Results

We found 3,929 citations of interest in the updated search. Of these, 23 full-text papers were potentially relevant and assessed for eligibility (Fig. 1). Eight papers met inclusion criteria (Gilbert et al., Reference Gilbert, Moser, Van Dorn, Swanson, Wilder, Robbins, Keator, Steadman and Swartz2010; Link et al., Reference Link, Epperson, Perron, Castille and Yang2011; Ogilvie and Kisely, Reference Ogilvie and Kisely2022; Phelan et al., Reference Phelan, Sinkewicz, Castille, Huz, Muenzenmaier and Link2010; Pollack et al., Reference Pollack, McFarland, Mahler and Kovas2005; Segal, Reference Segal2019; Segal et al., Reference Segal, Rimes and Badran2023; Swanson et al., Reference Swanson, Borum, Swartz, Hiday, Wagner and Burns2001). Reasons for exclusion were that records were unpublished theses or were not of a relevant design, setting and outcome (Fig. 1). Adding these studies to the previous search meant that there were 13 papers from 11 studies in total as different aspects of one study were reported in two papers (Fig. 1 and Table 2).

Figure 1. PRISMA diagram.

Table 2. Included studies

CTO = Community Treatment Order, OR = Odds Ratio, SEIFA = Socio-Economic Indexes for Areas, CBA = Controlled before and after, RCT = Randomised Controlled Trial

Of the 13 papers, nine were from the United States (Gilbert et al., Reference Gilbert, Moser, Van Dorn, Swanson, Wilder, Robbins, Keator, Steadman and Swartz2010; Hiday and Scheid-Cook, Reference Hiday and Scheid-Cook1987, Reference Hiday and Scheid-Cook1989; Link et al., Reference Link, Epperson, Perron, Castille and Yang2011; Phelan et al., Reference Phelan, Sinkewicz, Castille, Huz, Muenzenmaier and Link2010; Pollack et al., Reference Pollack, McFarland, Mahler and Kovas2005; Steadman et al., Reference Steadman, Gounis, Dennis, Hopper, Roche, Swartz and Robbins2001; Swanson et al., Reference Swanson, Borum, Swartz, Hiday, Wagner and Burns2000, Reference Swanson, Swartz, Wagner, Burns, Borum and Hiday2001) and four were from Australia (Ogilvie and Kisely, Reference Ogilvie and Kisely2022; Power, Reference Power, Churchill, Owen, Hotopf and Singh1992; Segal, Reference Segal2019; Segal et al., Reference Segal, Rimes and Badran2023). Five studies had follow-up periods of up to 12 months, otherwise follow-up ranged between 36 months and 12.4 years or was unspecified (Table 2). Details for one study (Power, Reference Power, Churchill, Owen, Hotopf and Singh1992) were only available from a subsequent systematic review (Churchill et al., Reference Churchill, Owen, Singh and Hotopf2007).

Three papers were from two RCTs (Steadman et al., Reference Steadman, Gounis, Dennis, Hopper, Roche, Swartz and Robbins2001; Swanson et al., Reference Swanson, Borum, Swartz, Hiday, Wagner and Burns2000, Reference Swanson, Swartz, Wagner, Burns, Borum and Hiday2001). In the case of the RCT from North Carolina (Swanson et al., Reference Swanson, Borum, Swartz, Hiday, Wagner and Burns2000, Reference Swanson, Swartz, Wagner, Burns, Borum and Hiday2001), randomised results were supplemented by several post-hoc analyses. In the first, they analysed a non-random sample who underwent more than 180 days of CTO placement (Swanson et al., Reference Swanson, Swartz, Wagner, Burns, Borum and Hiday2000). However, analysis of a group that has not been randomly assigned to treatment of less or more than 180 days may reflect a bias when the order was selectively extended when it appeared to be of benefit. In the second, a non-random group of individuals who were judged too violent to be included in the RCT were incorporated into a series of multiple staged, stepwise logistic regression models (Swanson et al., Reference Swanson, Borum, Swartz, Hiday, Wagner and Burns2001). Notably, usable data on arrests were only available for a subgroup of participants with a prior history of multiple admissions combined with prior arrests and/or violent behaviour, not the whole sample (Swanson et al., Reference Swanson, Borum, Swartz, Hiday, Wagner and Burns2001).

Overall, study quality was no more than moderate (Table 2, Supplementary Material 1). All but two studies adjusted for potential confounders through matching, adjustment or randomisation (Hiday and Scheid-Cook, Reference Hiday and Scheid-Cook1987; Segal et al., Reference Segal, Rimes and Badran2023).

Outcomes

All but two studies used externally recorded criminal justice contacts such as arrests or court appearances as measured by administrative databases (Table 2). The two remaining studies used standardised measures (Phelan et al., Reference Phelan, Sinkewicz, Castille, Huz, Muenzenmaier and Link2010; Power, Reference Power, Churchill, Owen, Hotopf and Singh1992). In a further two studies, externally recorded criminal justice data were supplemented by information from participants, families, staff or hospital records (Hiday and Scheid-Cook, Reference Hiday and Scheid-Cook1987; Swanson et al., Reference Swanson, Swartz, Wagner, Burns, Borum and Hiday2000). Three studies considered serious criminal behaviour (Link et al., Reference Link, Epperson, Perron, Castille and Yang2011; Segal, Reference Segal2019; Segal et al., Reference Segal, Rimes and Badran2023).

Table 3 displays the results indicating either improvement, decline, or no change, for people on CTOs compared to controls in terms of violence reported by participants or others, as well as criminal justice contacts, with subgroups for serious and minor criminal behaviour. Unadjusted results showed increases in criminal justice contacts while randomised data uniformly showed no significant difference. The results for controlled trials with adjusted analyses either showed a significant decrease or no difference.

Table 3. Studies by outcome

Unadj = Unadjusted; Adj = Adjusted; RCT = Randomised Controlled Trial

Figure 2 shows the results of the meta-analysis. There were no significant differences for any of the outcomes between CTO cases and voluntary controls, the effect size declining as study design improved from non-randomised data on self-reported criminal behaviour, through third-party criminal justice records and finally to RCTs. Similarly, there was no significant findings in the subgroup analysis of serious criminal behaviour (OR = 1.01; 95% CI = 0.52–1.97; p = 0.97; I 2 = 98%; k = 3).

Figure 2. Aggression or criminal justice contacts.

In terms of sensitivity analyses, there was no change in the results when we excluded the study where details on aggression were derived from secondary sources (OR = 0.54; 95% CI = 0.17–1.75; p = 0.30; I 2 = 80%; k = 3), or when analyses were restricted to studies that presented adjusted results for informant reported outcomes (OR = 0.49; 95% CI = 0.22–1.05; p = 0.07; I 2 = 65%; k = 3) and criminal justice contacts (OR = 0.78; 95% CI = 0.53–1.15; p = 0.21; I 2 = 58%; k = 6). There was also no change in the results when we excluded the study that only had usable data from selected subgroups of participants rather than the whole sample (OR = 0.94; 95% CI = 0.62–1.42; p = 0.76; I 2 = 95%; k = 7).

Heterogeneity, publication bias and certainty of evidence

All the above results showed significant heterogeneity apart from those of criminal justice contacts as reported in RCTs and the subgroup analysis of adjusted informant reported outcomes. We further explored this by excluding each study in turn in every analysis but this did not result in any I 2 value of less than 50%. However, removal of one study from the controlled trials of criminal justice contacts did reduce the I 2 value from 94% to 56% (Segal et al., Reference Segal, Rimes and Badran2023). We were unable to analyse for the effects of publication bias as none of the analyses had 10 or more studies. As most of the included studies were observational, both outcomes (violence/aggression and criminal justice contacts) were initially rated as being of low certainty. This was downgraded to being very low given high heterogeneity (inconsistency), and because some of the studies had low numbers (imprecision) or other outcomes as the primary focus (indirectness).

Discussion

This is the first such meta-analysis of the effect of the CTOs on aggression and criminal behaviour. We only included studies with contemporaneous controls as mirror image designs are subject to regression to the mean and may overestimate any effects (Eccles et al., Reference Eccles, Grimshaw, Campbell and Ramsay2003). For instance, in the most comprehensive meta-analysis of the effect of CTOs on health outcomes, before-and-after studies showed changes in health service use, that were no longer apparent in RCTs or observational studies with contemporaneous comparisons (Barnett et al., Reference Barnett, Matthews, Lloyd-Evans, Mackay, Pilling and Johnson2018). Mirror image studies may also be affected by secular trends or sudden changes in practice or policy that are unrelated to CTOs. We identified 13 papers from 11 studies, of which two were RCTs. There were no significant differences for any of the outcomes between CTO cases and voluntary controls, the effect size declining as study design improved from non-randomised data on self-reported criminal behaviour, through third party criminal justice records and finally to RCTs. Similarly, there was no significant finding in the subgroup analysis of serious criminal behaviour.

Both CTOs and forensic orders have similarities in balancing patient autonomy with reducing symptoms and the potential harm to self and others. However, the use of CTOs is much higher than that of forensic orders and, as a result, it is important to assess any potential benefits in terms of criminal behaviour (Gill et al., Reference Gill, Amos, Muhsen, Hatton, Ekanayake and Kisely2020). This is also important because of the human rights concerns regarding the use of CTOs. The United Nations Convention on the Rights of Persons with Disabilities (CRPD), which Australia and New Zealand have ratified, requires the States Parties to ensure that people with disabilities enjoy legal capacity on an equal basis with others (article 12) (United Nations, 2006). It encourages supported, rather than substitute decision-making and requires that any measures related to the exercise of legal capacity must be proportional and tailored to the person’s circumstances, apply for the shortest time possible and subject to appropriate and effective safeguards. Article 14 of the CRPD states that the existence of disability shall not justify a deprivation of liberty, and any deprivation of liberty must be on an equal basis with others. While some have interpreted the CRPD as prohibiting involuntary psychiatric treatment under all circumstances, a balanced and realistic interpretation requires that involuntary psychiatric treatment must be used only as the last resort, with strong safeguards and for the shortest possible time to promote safety and the right to health, where less restrictive interventions cannot achieve that outcome (Gill and Sartorius, Reference Gill and Sartorius2024). This raises ethical concerns on long term restriction of autonomy and liberty of individuals in the community through legislative mechanisms like the CTOs and requires empirical analysis of the evidence of their effectiveness.

Qualitative research illuminates the implications that CTOs have on human rights. In their meta-synthesis of stakeholder perspectives – including those of service users, relatives, mental health professionals and psychiatrists – Goulet et al. (Reference Goulet, Pariseau-Legault, Côté, Klein and Crocker2020) highlight that many recipients of CTOs perceive them to be a mechanism of medication compliance, and a means of ensuring community mental healthcare contact. While many see the value of CTOs in improving treatment adherence, and protecting recipients and others from harm, others (including service providers) raise concerns of outright legal abuse and coercion (Goulet et al., Reference Goulet, Pariseau-Legault, Côté, Klein and Crocker2020). That is, some believe that the legal criteria for enforcing CTOs are too rigid, and greater flexibility would enhance the human rights of recipients. Significant variations in CTO exist across the jurisdictions in which they are used, reinforcing concerns about the ethical justification of CTOs, particularly if recipients did not break the law in the first instance (Rugkåsa et al., Reference Rugkåsa2016).

Although some people with severe mental illness may be at greater risk of aggression, it is often comorbid substance or alcohol use and conduct or personality disorder that are the major drivers of either violence or contacts with the criminal justice system (Swanson et al., Reference Swanson, Van Dorn, Swartz, Smith, Elbogen and Monahan2008; Witt et al., Reference Witt, Van Dorn and Fazel2013). Using CTOs to improve adherence to the treatment of positive symptoms of psychosis does have a role but other contributors merit consideration (Swanson et al., Reference Swanson, Swartz, Van Dorn, Elbogen, Wagner, Rosenheck, Stroup, McEvoy and Lieberman2006). These include homelessness and a history of criminal behaviour, violence, victimisation and sexual or physical child maltreatment, as well as a parental history of criminal involvement or alcohol use (Witt et al., Reference Witt, Van Dorn and Fazel2013). Wider social factors may also be relevant such as exposure to violence in the surrounding environment (Swanson et al., Reference Swanson, Swartz, Essock, Osher, Wagner, Goodman, Rosenberg and Meador2002). There are similar findings when analyses are restricted to severe violence (Witt et al., Reference Witt, Van Dorn and Fazel2013). This therefore points to the need to also address these factors even though they are less easily modifiable through direct clinical intervention.

Further research should therefore consider the role of interventions that address previous victimisation and maltreatment, comorbid alcohol or substance use and the wider social context such as homelessness. The human rights framework adopted by the CRPD requires the ratifying countries to promote the rights to habitation and rehabilitation, education, employment, health, adequate standard of living and social inclusion of persons with disabilities (Gill, Reference Gill2019). Most mental health laws, however, continue to focus on civil commitment through legislative mechanisms such as CTOs rather than these economic, social and cultural rights (McSherry, Reference McSherry2014).

Simply using legislation to enforce treatment is unlikely to achieve the twin goals of promotion of community safety and protecting the human rights of the individuals with mental illness. Promotion of the economic, social and cultural rights of people with mental illness, through provision of housing, access to health services and other social goods is more likely to pave the way for protection of civil and political rights and lead to a safer and just society. This is especially relevant when there is no evidence that legislative mechanisms like CTOs lead to a reduction in aggression or criminal behaviour, as found by this meta-analysis. Alternative approaches therefore should give greater emphasis to addressing social determinants rather than solely focussing on individuals (Kirkbride et al., Reference Kirkbride, Anglin, Colman, Dykxhoorn, Jones, Patalay, Pitman, Soneson, Steare, Wright and Griffiths2024). One example from the United States was a community building initiative for young people that reduced substance use, violence, delinquency, and behavioural issues (Kuklinski et al., Reference Kuklinski, Fagan, Hawkins, Briney and Catalano2015).

Limitations

All but two of the included studies were observational, and many were of administrative health data. These may be subject to recording bias and lack information on social aspects of disability. Cases and controls may also have differed in ways for which it was not possible to match or adjust. It is therefore possible that the CTO cases were more seriously ill, or at greater risk of aggression and other criminal behaviour than the controls. The reported levels of aggressive or criminal behaviour may therefore still represent a reduction from what might have been the case if these individuals had not been on a CTO. On a related issue, there was no information on relevant comorbidities such as antisocial personality disorder. The focus of this review was on 12-month outcomes as this was the timeframe considered by most of the included studies. This may have under-estimated any benefits that might have arisen from actual CTO placement if participants had been on voluntary treatment for part of the relevant follow-up period.

Many of the current studies relied on data originally collected for other purposes, not specifically for studying the relationship between mental illness and violence. For instance, findings of higher levels of criminal justice contact in people with mental illness may reflect that they are at an increased risk of being monitored and arrested, rather than of committing offences. The lack of evidence of any effect may also have been due to the limited quality of the available data as highlighted by the very low rating for the certainty of evidence using the GRADE framework. More studies are therefore indicated with violence as the primary outcome and sufficient power to identify a range of relevant risk factors. Where possible, outcomes should be from external sources such as criminal justice records.

The majority of our meta-analyses showed a high degree of heterogeneity. Although we tried to accommodate this with random effects models, our results should still be viewed with caution. Finally, we were unable to test for publication bias as none of the outcomes had 10 or more studies.

Conclusion

On the limited available evidence, there remains uncertainty regarding the role of CTOs in addressing aggression or criminal behaviour in people with mental illness. Given that the risk of violence is increased by comorbid or nonclinical variables, including victimisation, maltreatment and the wider environment, management should also focus on the whole person and their community through social and public health interventions, not just legislative control.

Supplementary material

The supplementary material for this article can be found at https://doi.org/10.1017/S2045796025000058.

Availability of data and materials

All relevant data for this study are provided in-text or supplementary materials. Request for further information is available from the corresponding author on reasonable request.

Acknowledgements

PROSPERO registration: CRD42020155996 (prospective).

Financial support

This research received no specific grant from any funding agency, commercial or not-for-profit sectors.

Competing interests

None.

Ethical standards

Not applicable.

References

Barnett, P, Matthews, H, Lloyd-Evans, B, Mackay, E, Pilling, S and Johnson, S (2018) Compulsory community treatment to reduce readmission to hospital and increase engagement with community care in people with mental illness: A systematic review and meta-analysis. The Lancet Psychiatry 5, 10131022.CrossRefGoogle ScholarPubMed
Brophy, L, Edan, V, Kisely, S, Lawn, S, Light, E, Maylea, C, Newton-Howes, G, Ryan, CJ, Weller, PJ and Zirnsak, T-M (2021) The urgent need to review the use of CTOs and compliance with the UNCRPD across Australian jurisdictions. International Journal of Mental Health and Capacity Law 28, 39.Google Scholar
Churchill, R, Owen, G, Singh, S and Hotopf, M (2007) International Experiences of Using Community Treatment Orders. London. Department of Health, Institute of Psychiatry.Google Scholar
Eccles, M, Grimshaw, J, Campbell, M and Ramsay, C (2003) Research designs for studies evaluating the effectiveness of change and improvement strategies. Quality and Safety in Health Care 12, 4752.CrossRefGoogle ScholarPubMed
Gilbert, AR, Moser, LL, Van Dorn, RA, Swanson, JW, Wilder, CM, Robbins, PC, Keator, KJ, Steadman, HJ and Swartz, MS (2010) Reductions in arrest under assisted outpatient treatment in New York. Psychiatric Services 61, 996999.CrossRefGoogle ScholarPubMed
Gill, NS (2019) In Human Rights Framework: An Ethical Imperative for Psychiatry. London, England: Sage Publications Sage UK, 810.Google ScholarPubMed
Gill, NS, Amos, A, Muhsen, H, Hatton, J, Ekanayake, C and Kisely, S (2020) Measuring the impact of revised mental health legislation on human rights in Queensland, Australia. International Journal of Law and Psychiatry 73, .CrossRefGoogle ScholarPubMed
Gill, N and Sartorius, N (2024) The application of the CRPD and other international human rights instruments in mental health care. In Gill N, Sartorius N (eds) Mental Health and Human Rights. Sustainable Development Goals Series. Springer, Cham. 516.CrossRefGoogle Scholar
Goulet, M-H, Pariseau-Legault, P, Côté, C, Klein, A and Crocker, AG (2020) Multiple Stakeholders’ perspectives of involuntary treatment orders: A meta-synthesis of the qualitative evidence toward an exploratory model. International Journal of Forensic Mental Health 19, 1832.CrossRefGoogle Scholar
Hiday, VA and Scheid-Cook, TL (1987) The North Carolina experience with outpatient commitment: A critical appraisal. International Journal of Law Psychiatry 10, 215232.CrossRefGoogle ScholarPubMed
Hiday, VA and Scheid-Cook, TL (1989) A follow-up of chronic patients committed to outpatient treatment. Hospital and Community Psychiatry 40, 5259.Google ScholarPubMed
Kirkbride, JB, Anglin, DM, Colman, I, Dykxhoorn, J, Jones, PB, Patalay, P, Pitman, A, Soneson, E, Steare, T, Wright, T and Griffiths, SL (2024) The social determinants of mental health and disorder: Evidence, prevention and recommendations. World Psychiatry: Official Journal of the World Psychiatric Association (WPA) 23, 5890.CrossRefGoogle ScholarPubMed
Kisely, SR, Campbell, LA and O’Reilly, R (2017) Compulsory community and involuntary outpatient treatment for people with severe mental disorders. Cochrane Database Systematic Reviews 3, Google ScholarPubMed
Kisely, S, Zirnsak, T, Corderoy, A, Ryan, CJ and Brophy, L (2024) The benefits and harms of community treatment orders for people diagnosed with psychiatric illnesses: A rapid umbrella review of systematic reviews and meta-analyses. Australian & New Zealand Journal of Psychiatry 58, 555570.CrossRefGoogle ScholarPubMed
Kuklinski, MR, Fagan, AA, Hawkins, JD, Briney, JS and Catalano, RF (2015) Benefit-cost analysis of a randomized evaluation of communities that care: Monetizing intervention effects on the initiation of delinquency and substance use through grade 12. Journal of Experimental Criminology 11, 165192.CrossRefGoogle ScholarPubMed
Light, EM, Robertson, MD, Boyce, P, Carney, T, Rosen, A, Cleary, M, Hunt, GE, O’Connor, N, Ryan, CJ and Kerridge, IH (2017) How shortcomings in the mental health system affect the use of involuntary community treatment orders. Australian Health Review 41, 351356.CrossRefGoogle ScholarPubMed
Link, BG, Epperson, MW, Perron, BE, Castille, DM and Yang, LH (2011) Arrest outcomes associated with outpatient commitment in New York State. Psychiatric Services 62, 504508.CrossRefGoogle ScholarPubMed
McSherry, B (2014) Mental health laws: Where to from here. Monash University Law Review 40, 175197.Google Scholar
Moher, D, Liberati, A, Tetzlaff, J and Altman, DG (2009) Preferred reporting items for systematic reviews and meta-analyses: The PRISMA statement. Annals of Internal Medicine 151, 264269.CrossRefGoogle ScholarPubMed
O’Brien, AJ (2014) Community treatment orders in New Zealand: Regional variability and international comparisons. Australasian Psychiatry 22, 352356.CrossRefGoogle ScholarPubMed
Ogilvie, JM and Kisely, S (2022) Examining the health and criminal justice characteristics for young people on compulsory community treatment orders: An Australian birth cohort and data linkage study. International Journal of Law and Psychiatry 83, . Epub 2022 Jun 24 . doi:10.1016/j.ijlp.2022.101813.CrossRefGoogle ScholarPubMed
Phelan, JC, Sinkewicz, M, Castille, DM, Huz, S, Muenzenmaier, K and Link, BG (2010) Effectiveness and outcomes of assisted outpatient treatment in New York State. Psychiatric Services 61, 137143.CrossRefGoogle ScholarPubMed
Pollack, DA, McFarland, BH, Mahler, JM and Kovas, AE (2005) Outcomes of patients in a low-intensity, short-duration involuntary outpatient commitment program. Psychiatric Services 56, 863866.CrossRefGoogle Scholar
Power, P (1992) A controlled Study of the clinical effectiveness of community treatment orders in Australia. In Churchill, R, Owen, G, Hotopf, M, and Singh, S (eds), International Experiences of Using Community Treatment Orders. London: Institute of Psychiatry, Google Scholar
Rugkåsa, J (2016) Effectiveness of community treatment orders: The international evidence. The Canadian Journal of Psychiatry 61, 1524.CrossRefGoogle ScholarPubMed
Segal, SP (2019) The utility of outpatient commitment: Reduced risks of victimization and crime perpetration. European Psychiatry 56, 97104.CrossRefGoogle ScholarPubMed
Segal, SP (2020) The utility of outpatient civil commitment: Investigating the evidence. International Journal of Law Psychiatry 70, .CrossRefGoogle ScholarPubMed
Segal, SP, Rimes, L and Badran, L (2023) Need for treatment, a less restrictive alternative to hospitalization, and treatment provision: the utility of community treatment orders. Schizophrenia Bulletin Open 4, .CrossRefGoogle ScholarPubMed
Steadman, HJ, Gounis, K, Dennis, D, Hopper, K, Roche, B, Swartz, M and Robbins, PC (2001) Assessing the New York City involuntary outpatient commitment pilot program. Psychiatric Services 52, 330336.CrossRefGoogle ScholarPubMed
Swanson, JW, Borum, R, Swartz, MS, Hiday, VA, Wagner, HR and Burns, BJ (2001) Can involuntary outpatient commitment reduce arrests among persons with severe mental illness? Criminal Justice and Behavior 28, 156189.CrossRefGoogle Scholar
Swanson, JW, Swartz, MS, Essock, SM, Osher, FC, Wagner, HR, Goodman, LA, Rosenberg, SD and Meador, KG (2002) The social-environmental context of violent behavior in persons treated for severe mental illness. American Journal of Public Health 92, 15231531.CrossRefGoogle ScholarPubMed
Swanson, JW, Swartz, MS, Van Dorn, RA, Elbogen, EB, Wagner, HR, Rosenheck, RA, Stroup, TS, McEvoy, JP and Lieberman, JA (2006) A national study of violent behavior in persons with schizophrenia. Archives of General Psychiatry 63, 490499.CrossRefGoogle ScholarPubMed
Swanson, JW, Swartz, MS, Wagner, HR, Burns, BJ, Borum, R and Hiday, VA (2000) Involuntary out-patient commitment and reduction of violent behaviour in persons with severe mental illness. The British Journal of Psychiatry 176, 324331.CrossRefGoogle ScholarPubMed
Swanson, JW, Van Dorn, RA, Swartz, MS, Smith, A, Elbogen, EB and Monahan, J (2008) Alternative pathways to violence in persons with schizophrenia: The role of childhood antisocial behavior problems. Law and Human Behaviour 32, 228240.CrossRefGoogle ScholarPubMed
United Nations (2006) Convention on the Rights of Persons with Disabilities, adopted on 13 December 2006, GA Res 61/106, UN Doc A/Res/61/106 (entered into force 3 May 2008). Available at: https://www.un.org/development/desa/disabilities/convention-on-the-rights-of-persons-with-disabilities.html.Google Scholar
Witt, K, Van Dorn, R and Fazel, S (2013) Risk factors for violence in psychosis: Systematic review and meta-regression analysis of 110 studies. PLoS One 8, .CrossRefGoogle ScholarPubMed
Figure 0

Table 1. Search terms

Figure 1

Figure 1. PRISMA diagram.

Figure 2

Table 2. Included studies

Figure 3

Table 3. Studies by outcome

Figure 4

Figure 2. Aggression or criminal justice contacts.

Supplementary material: File

Kisely et al. supplementary material

Kisely et al. supplementary material
Download Kisely et al. supplementary material(File)
File 20.3 KB