Are principles of recovery-oriented practice evident in staff and service user perspectives on seclusion?

Abstract

Purpose

This paper aims to explore the extent to which principles of recovery-oriented practice are evident in the published perspectives and experiences of health professionals and service users on seclusion in adult mental health services.

Design/methodology/approach

A systematic review informed by PRISMA guidelines was conducted, drawing from four databases, which were searched in August 2018 and August 2022. Only original empirical studies rated as having “major” relevance were included. Data were extracted from 31 studies and qualitatively synthesised through deductive analysis using recovery principles as themes.

Findings

There was limited evidence of perceptions of seclusion being being consistent with recovery principles, with greater evidence of perceptions that directly opposed them. Studies of service user perspectives highlighted this more often than staff perspectives. The findings highlight paradoxical relationships between care and control and conflicting rights and emphasise the need to openly acknowledge the complexity of seclusion and its interface with recovery.

Research limitations/implications

This review was developed in line with international best practice and the protocol was registered. Using a search string with only three components maximised sensitivity during searches and minimised the risk of relevant literature being missed. Limitations include the focus on studies where the full text was published in English.

Originality/value

This review makes a unique contribution, highlighting that, to the best of the authors’ knowledge, no studies to date have explicitly explored the perspectives and experiences of staff and service users on the use of seclusion in the context of recovery-oriented practice. The findings are relevant to clinical practice, policy and future research, including amending procedures and practices to partially reconcile seclusion and recovery where the seclusion is deemed necessary.

Citation

Stíobhairt, A., Staunton, D. and Guerin, S. (2023), "Are principles of recovery-oriented practice evident in staff and service user perspectives on seclusion?", Mental Health Review Journal, Vol. 28 No. 2, pp. 144-166. https://doi.org/10.1108/MHRJ-02-2022-0009

Plain language summary

Current approaches to mental health focus on the importance of recovery. However, in many mental health settings, individuals may still experience practices such as seclusion, where they are placed in a space they cannot leave for a period of time. We reviewed published research to see if studies on seclusion considered aspects of recovery. We found little evidence of recovery principles in how individuals experienced seclusion and highlight the need for mental health services to consider the implications of this relationship.

Introduction

Seclusion has been defined by the World Health Organisation (WHO, 2012, p. 82) as the “placement of an individual alone in a locked room or secured area from which he or she is physically prevented from leaving”. While formal definitions vary within the literature, they generally describe a relatively homogenous practice with a set of core characteristics that has largely remained unchanged over time (National Institute for Health and Care Excellence, 2015). To the best of the authors’ knowledge, four systematic reviews of staff and service user perspectives on seclusion in adult mental health (AMH) services have been published to date (Askew et al., 2019; Happell and Harrow, 2010; Mellow et al., 2017; Van der Merwe et al., 2013). Taken together, these reviews indicate that many service users report negative experiences of seclusion, with only a minority reporting positive experiences. In contrast, the majority of staff believed that seclusion was necessary, therapeutic and should continue to be used, despite acknowledging negative service user experiences and a perceived risk of abuse.

These findings should be considered in the context of increasing calls for reductions in restrictive practices in mental health settings (Muir-Cochrane et al., 2018) and the evidence for negative outcomes associated with these practices (Butterworth et al., 2022). In addition, models such as Safewards (Mullen et al., 2022) and trauma-informed approaches (Bloom and Farragher, 2013; Ross, 2020) offer alternatives to harmful restrictive practices. Despite this, seclusion remains a part of AMH services, with evidence of on-going support among clinical staff (Happell and Harrow, 2010; Van der Merwe et al., 2013), indicating that efforts to minimise or eliminate it may face significant barriers.

One of the main counterpoints to the use of seclusion highlighted in the literature is the importance of recovery-oriented approaches to AMH (Mullen et al., 2022). A review of qualitative studies of seclusion and restraint by Hawsawi et al. (2020) highlights the need for more recovery-oriented care. Recovery-oriented practice has increasingly occupied a central role in AMH policy and practice internationally in recent years. While there is no international consensus on the principles of recovery-oriented practice, the framework published by the Australian Government (2010) has considerable international support (World Health Organisation, 2012). It contains six principles and includes “uniqueness of the individual”, “real choices”, “attitudes and rights”, “dignity and respect”, “partnership and communication” and “evaluating recovery”.

It has been argued that seclusion and recovery-oriented practice are incompatible (Ashcraft and Anthony, 2008), with research by Brophy et al. (2016) highlighting seclusion and restraint as “anti-recovery”. However, research also suggests that a recovery orientation can promote the safety in AMH settings that seclusion might be perceived to ensure (Marshall et al., 2019). It is interesting that while recovery-informed practices are regularly discussed in the context of reducing restrictive practices in general, the literature on experiences of seclusion has not been examined though the lenses of recovery. With limited exceptions, its only since the 2010s that attention has been given to the view that these constructs are opposing approaches to AMH. Therefore, this review aimed to address the research question:

RQ1.

Are principles of recovery-oriented practice evident in the perspectives of staff and service users on seclusion?

Methods

Design

This systematic review was conducted in line with the Preferred Reporting Items for Systematic Reviews and Meta-analyses (PRISMA) statement (Moher et al., 2009) and was registered on the online prospective register PROSPERO (Record ID = CRD42019111869; www.crd.york.ac.uk/PROSPERO/).

Search strategy

Searches were conducted on PsycINFO (ProQuest interface), CINAHL Plus (EBSCOhost interface), MEDLINE (ProQuest interface) and Embase (Elsevier interface) based on their relevance to disciplines associated with inpatient mental health services and their frequent inclusion in related reviews (Hui et al., 2013; Mellow et al., 2017; Van Der Merwe et al., 2013). The components of the search string included the individuals of interest (i.e. staff and service users), the practice of interest (i.e. seclusion) and the subject of interest (i.e. perspectives). See the following database search terms for details, with minor adaptations made to match the requirements of each database. Search terms in bold were included in both the primary search on 28/08/2018 and the updated search on 26/08/2022. Those not in bold were excluded from the updated search.

service user* OR patient OR inpatient OR client OR consumer OR staff* OR nurse* OR psychiatrist* OR psychologist* OR therapist* OR advoca*

AND

seclu* OR restricti* OR restrai* OR coerc* OR confin* OR segregat* OR locked OR padded OR involuntar*

AND

Perspective* OR perception OR perceive* OR "views" OR "thoughts" OR opinion* OR attitude* OR experience OR feel*

Inclusion and exclusion criteria

The SPIDER framework (Cooke et al., 2012) was used to structure inclusion and exclusion criteria at the title/abstract and full-text screening stages to increase specificity. For each criterion below, studies containing non-relevant data were included if relevant data could be selectively extracted.

Sample: Studies were included where they involved human participants aged 18 years or above. No minimum sample size was required. Data from service users that had not personally been secluded were excluded to ensure expressed perspectives were based on lived experience. Such service users have also been excluded from some related reviews (Mellow et al., 2017; Tingleff et al., 2017).

Phenomenon of interest: Studies were included if they reported on the use of seclusion with service users in AMH services.

Study design/research type: Only articles with original empirical data were included. All methodologies and designs were included. Only pre-intervention data were included from studies involving interventions that affected staff perspectives.

Evaluation: Studies that reported staff and/or service user perspectives on seclusion were included. For the purpose of this review, the term “perspective” was considered to encompass views, attitudes and experiences, with studies included where data relating to perspectives could be selectively extracted. A system was developed in which relevance was rated based on the degree of focus on perspectives on seclusion within a study, the quantity of relevant data and the degree to which the data would contribute to the aims of the review (see Supplementary Table 1 for details). Tingleff et al. (2017) excluded some studies on the basis that relevant data were “too scarce” (p. 684), however details of how this was operationalised were not reported. Thirty-one studies were rated as having “major” relevance; therefore, studies with lower ratings were excluded. A subset of studies (n = 39) was double rated to allow for an estimation of the reliability of this system. The reviewers were blind to each other’s judgements until this stage was complete.

Publication status: The limits placed on the age of included studies vary among previous reviews. As recovery first emerged as a distinct concept in the 1990s (Anthony and Ashcraft, 2010), searches were limited to peer-reviewed studies from 1990 to 26/08/2022, when final searches were conducted. Advanced online publications were included where the final manuscripts were available and reasonable efforts were made to obtain inaccessible manuscripts (e.g. by contacting corresponding authors directly or seeking inter-library loans). Studies for which only abstracts have been published were excluded (e.g. conference presentations). Duplicates and republication of the same data with a markedly similar analysis and findings were excluded.

Language: No limits were placed on language when conducting searches, however studies were only included if the full-text manuscript was published in English. Titles and abstracts published in languages other than English were translated using Google Translate and then screened. Those that appeared relevant are listed in Supplementary Table 2.

Study identification and selection

Search results were uploaded to www.covidence.org. All study titles/abstracts and full texts were screened by two reviewers. Decisions at both stages were made independently, other than a small number of early disagreements that were reviewed at the outset of each stage to promote reliability. Disagreements at both stages were resolved by a third researcher, except for studies identified during updated searches, which were resolved collaboratively by the two raters. The reference sections of all studies included in the review were manually searched to identify additional potentially relevant studies. Identified studies underwent the screening procedure outlined above.

Data extraction and management

Data were extracted from all sections except abstracts, with summary data and direct quotations compiled based upon the Joanna Briggs Institute Qualitative Data Extraction Tool (Lockwood et al., 2015). The extraction sheet included publication details, study characteristics and findings. Data were extracted by a second reviewer from six randomly selected studies, representing 19% of included studies. Both reviewers were blind to each other’s extraction until this stage was complete. Inter-rater reliability was calculated with Gwet’s (2008) AC1 used where information on individual rater decisions was available.

Quality appraisal

The Mixed Methods Appraisal Tool 2018 (MMAT; Hong, 2019) was used to evaluate the quality of included studies, as it was anticipated that a range of methodologies and designs would be included in the review. Double rating was conducted for six studies, and inter-rater reliability was calculated. Studies of poor quality were not excluded; however, ratings were used to facilitate interpretation of the findings. The quality appraisal ratings assigned to each study are presented in Supplementary Table 3.

Data synthesis

A thematic synthesis was conducted based on Thomas and Harden (2008), with additional guidance from Braun and Clarke (2006). A deductive analysis was conducted, whereby themes were defined a priori, based on the principles of recovery-oriented practice published by the Australian Government (2010). Coding was primarily conducted at a semantic level, with a small proportion of data coded at a latent level (Terry et al., 2017). Data from three randomly selected studies were initially coded by two researchers to promote validity and reliability in this process. In reporting the findings, it is noted that studies varied in how they reported frequencies. Some studies used descriptive terms (e.g. “some”, “most”) to refer to the frequency of particular perceptions among their sample and some made no reference to frequency at all (e.g. “patients said […]”). Where possible, statistics such as percentages have been reported, while descriptive terms have been used where these were used in the relevant papers.

Results

The 31 included studies were published between 1992 and 2022 (see Figure 1 for PRISMA flowchart). The publication details and methodology of individual studies are presented in Table 1.

The studies were conducted across eight countries (Australia = 8; the USA = 6; The Netherlands = 5; England = 6; Canada = 2; Finland = 2; Lesotho = 1; and New Zealand = 1). Fifteen studies used a qualitative design (n = 2–26), eight were mixed-methods (n = 24–193), five were quantitative non-randomised (n = 50–106) and three were quantitative descriptive (n = 78–540). There were no randomised controlled trials. There was 96.66% (15077/15476) agreement between raters for title/abstract screening. There was substantial agreement for full-text screening (y = 0.74) and moderate agreement for the study relevance rating system using Gwet’s AC1 (y = 0.56).

Among the 31 included studies, recovery was explicitly referenced once among staff perspectives on seclusion (23). As such, there was almost no evidence that recovery was explicitly considered in relation to seclusion by staff or service users. Authors of only four studies made explicit reference to recovery elsewhere in their article and this largely amounted to only brief consideration (1;9;11;28). Findings relating to each of the six Australian recovery principles are presented in turn in the following sections, where quoted extracts from the principles are used as section headings (Australian Government, 2010). As the recovery principles are closely related, a significant proportion of data were coded under two or more principles.

Uniqueness of the individual

There was no evidence that the unique needs of service users were recognised in relation to seclusion at a service level. There was only one example of individualised practices, whereby a service user was allowed their guitar during seclusion. Explicit acknowledgement of this was reported once in relation to service users actively seeking seclusion (13) and once in relation to their perception and therapeutic benefit (5).

Recognises that recovery is not necessarily about cure but is about having opportunities for choices and living a meaningful, satisfying and purposeful life, and being a valued member of the community:

There was no evidence of this. In contrast, comments from four studies indicated that some service users experienced further marginalisation from an already marginalised population. Some reported feeling stigmatised, abandoned, deprived of freedom and embarrassed and less accepted by peers afterwards (3;6;15;21). Holmes et al. (2004) described seclusion as a “catalyst in reawakening and intensifying already existing feelings of being alone and abandoned by significant others” (p. 570).

Accepts that recovery outcomes are personal and unique for each individual and go beyond an exclusive health focus to include an emphasis on social inclusion and quality of life:

Service users across multiple studies reported that seclusion enabled them to have valued experiences that indicated quality of life. They noted that time alone provided them with peace (11) and privacy (10;15) and opportunities for quiet reflection (10;14), meditation, prayer (6;22), sleep/rest (13;17;18) and to calm (28). When asked, 82% of service users across multiple Australian services responded they would like to use the seclusion room when they chose to be alone (19), however Stolker et al. (2006) found that service users residing in multiple-occupancy bedrooms held more favourable views prior to seclusion. The authors speculated that the lack of privacy and rest service users experienced may have made seclusion somewhat more attractive. Some service users also cited opportunities to “learn a lesson” (10, p. 580), sober up (15), de-escalate and allow medication to take effect (13), learn to control their behaviour and emotions (15;18) and have a positive effect on their mental health (15).

Notably, most findings were inconsistent with this aspect of the principle, with service users from four studies perceiving that even health needs were not adequately met during seclusion. Female service users across several English units described having no access to underwear or menstrual products, either at all or those they feel comfortable using (25). Many US service users complained that they were “not listened to when relating their medical histories, symptoms and methods of coping” (7, p. 36). One service user worried that her insulin would be forgotten during seclusion (21) and another reported that his diabetes-related need for frequent meals and use of the toilet were unmet (22). An Indigenous Australian noted that his head injury went untreated for three days during seclusion (24). In another study, glasses were listed among the objects most missed during seclusion (21).

There was abundant evidence of the adverse impact of the physical seclusion environment. Participants described an impaired sense of reality and control, feelings of going “mad” (1; 18, p. 374), loneliness (1;4;11) and rumination precipitated by the lack of social contact (3). Some described perceptual disturbances (28), including hyper-sensitivity to sensory stimulus (18) and interactions with psychotic hallucinations and delusions, contributing to fear and anxiety (11;14;22). Many service users across three studies (1;13;18;25) and two staff members (16) referred to significant boredom. Some service users noted that they had significantly limited social contact or opportunity for physical activity, which they relied on for emotion regulation (7;18). Notably, up to 95% of service users (4) and 62%–85% of mixed health professionals (4;8;9) believed that changes were needed to seclusion rooms. Varying levels of support was expressed by staff for modifications that could increase social contact and quality of life (4;9;8;13;15;21). Notably, van der Venne et al. (2021) found no significant statistical difference between the experiences of service users in a regular seclusion room versus one with designed to be more soothing; however, qualitative data suggested that it made the experience more bearable.

Service users wished for staff to understand their “personal needs and idiosyncrasies” (7, p. 35). Multiple service users in Lesotho were unhappy eating without first washing their dirty hands, with one reporting that he opted to go without food as a result (22). Service users in two studies described feeling uncomfortable in such small spaces (21;10). Findings from eight studies illustrated the adverse impact that unique personal histories had on service user lived experiences of seclusion. There was no evidence of individualised practices based on unique histories. Service user comments indicate that having normal clothing removed – often forcefully – and being made to wear hospital gowns is common practice across services and countries (7;13;15;16;21;25;28). One service user noted that this is a particularly difficult experience for those who had been sexually abused (7). Service users across several studies reported that seclusion caused them to revisit negative experiences with law enforcement and prison (6;22;24). Individual service users described felt they had done “something wrong against God” (21, p. 13), may be killed (3) or feared being raped (22).

Empowers individuals so they recognise that they are at the centre of the care they receive:

There was no evidence of this. Many findings demonstrated disempowerment and are discussed under other principles.

Real choices

Supports and empowers individuals to make their own choices about how they want to lead their lives and acknowledges choices need to be meaningful and creatively explored:

There was no evidence of this. In contrast, there was abundant evidence that the power to make choices in relation to seclusion was perceived to sit almost entirely with staff, while service users were greatly disempowered. Across Australian open, closed and forensic units between 46 and 62% of nurses said they would “never” seclude a service user that asked to be secluded (8;31). Some service users in The Netherlands perceived that nurses had “absolute power” (11, p. 280). Service users in across studies referred to the “unspoken rules” of the seclusion “game” (both 18, p. 374), “playing along” (12, p. 572) and behaving passively (3) to satisfy staff judgements that seclusion was no longer needed and “walking on eggshells” (25 p. 382) to avoided threats of seclusion. A US service user commented “I cried out for help and they didn’t do it till they were good and ready” (21, p. 10). In Australia, 89%–98% of mixed health professionals believed that seclusion “sometimes” or “often” disempowers service users (8;9). Across studies, 76%–89% of nurses (8;9;27; 31) and 91% of service users (27) believed that service users felt “helpless” during seclusion. Feelings of powerlessness and hopelessness were described by service users elsewhere (3;6;11;18;21;22;28). One English service user stated that some of his challenging behaviour during seclusion was a deliberate attempt to reassert some control (3).

Service users in multiple studies referred to social and physical force used in relation to seclusion, which precluded meaningful choices (e.g. 1, 24). Some English forensic service users perceived that staff were “exerting power over them” (10, p. 577), and an Australian nurse noted that a “show of staff is usually enough for [service users] to comply” (20, p. 560). Service users in the USA stated that seclusion was “forced on them” (7, p. 36) and one in England reported “being dragged […] and locked in” (10, p. 571). Another service user in Lesotho described being assaulted when they refused seclusion (22). Some felt “unable to get out” (6, p. 307), “trapped” (23, p. 28), imprisoned (6;10;17;18;25), in a “gas chamber” (21), subject to “torture” (1;17) or that they were being punished (3;7;10;12;15;28). Service users across several studies complained of a lack of privacy (e.g. 3;25;28), with some adding that they abused by being observed using the shower and toilet (3). Two nurses that voluntarily spent time in seclusion also viewed it as a punishment (16).

The adverse impact of being unable to make choices is particularly evident in several comments. A service user in the USA stated, “I cannot get things that I want to do or get out […] nobody was listening […] the doors are shut behind you […] and you are there alone” (6, p. 307). Another in Australia reported “they even take away your option to change the circumstances to try and lift your mood” (18, p, 373). An Australian service user stated, “when they first put me in there and I was just screaming and kicking and yelling because I didn’t want a needle and then I remembered just bursting into tears and I think I cried myself to sleep” (18). Findings discussed under other principles demonstrate that service users had virtually no power to alleviate their boredom, loneliness and discomfort.

Supports individuals to build on their strengths and take as much responsibility for their lives as they can at any given time:

There was no evidence that seclusion was perceived as supporting service users to build on their strengths. In reference to abilities, some English psychiatrists reportedly believed that seclusion “did not encourage the development of internal boundary setting” (5, p.428). Service users in four studies noted that the restrictive procedures surrounding seclusion prevented them from taking responsibility for some of their own basic needs. Some reported urinating (12) or defecating on the floor (24) or incontinence (7;24), as they were denied access to a toilet. A service user in the Netherlands described how staff chose his clothing and laid out the things he needed for a shower, as he was not allowed to do this for himself (11).

Findings illustrate a perception among some nurses that service users are given control within a broader context ultimately controlled by staff. Some nurses in Australia reportedly aimed to “facilitate the patients remaining in control” (20, p. 559) and avoid the use of seclusion, but viewed it as a legitimate form of social control that would “provide an opportunity for the individual to regain control” (20, p. 558). These comments indicate that some nurses perceived themselves as essentially supporting service users to take as much responsibility for their lives as they could at any given time.

Although many service users experienced seclusion as a transgression, comments from two studies indicated that some later agreed with staff judgements. Haw et al. (2011) found that 56% of service users in a forensic service thought they should have been secluded during the last episode of coercive treatment. Some service users recruited by El-Badri and Mellsop (2008) also reportedly felt “grateful at being kept under control and out of an overwhelming high-stimulus where they might do something impulsively they will later regret” (p. 251).

Ensures that there is a balance between duty of care and support for individuals to take positive risks and make the most of new opportunities:

There was no evidence of positive risk-taking to make the most of new opportunities, but instead some evidence of stringent or conservative criteria surrounding seclusion. Muir-Cochrane (1996) noted that “acceptable patient behaviour” (p. 561) was guided by clinicians’ individual assessment, and that nurses perceived control to be regained if service users “could reason with clinicians, talk about what had happened and accept limits by staff” (p. 560). Some service users complained that opportunities to discuss or defend their actions were not available (7;18), their emotion regulation strategy was misunderstood (7) or that seclusion continued for staff convenience (18).

Attitudes and rights

Involves listening to, learning from and acting upon communications from the individual and their carers about what is important to each individual:

There was no evidence that the views of carers were sought or listened to in relation to seclusion. Findings from several studies supplement those discussed elsewhere that are inconsistent with this aspect of the principle. Service users across five studies felt that seclusion could have been avoided through open, respectful communication or an opportunity for them to calm down (6;7;10;18;21), however some described feeling mocked or ignored (7). Some Australian service users perceived a lack of support following seclusion and desired an opportunity to “off-load emotions” afterwards (18, p. 375).

Promotes and protects individual’s legal, citizenship and human rights:

The studies illustrate a paradoxical relationship in which seclusion was frequently perceived to infringe upon rights in some respects, but to support them in others. This includes conflicts between the various rights of the secluded individual, the rights of others, changes to perceptions over time and the perceived therapeutic benefit of seclusion.

Findings from several studies related to the inappropriate use of seclusion in ways that infringe upon service user rights. Nurses in English services perceived that seclusion was or could be abused in several ways, including unnecessary duration; initiating or discontinuing based on staff levels; using medication as an additional restraint; not-adhering to policies; failing to use more appropriate alternatives; using seclusion to manage “difficult” service users; and using it as a punishment (2). Service users across other studies felt that their seclusion was inappropriate (18), initiated without sufficient efforts to help them (13, p. 205) and accompanied by over-sedation (24).

In describing their findings, Meehan et al. (2000) reported a sense that Australian service users “lose claim to any individuality or rights” (p. 375). Previously discussed experiences indicating infringed rights include perceptions of disempowerment, imprisonment and punishment; the use of social and physical force; forced wearing of hospital gowns; and minimalistic seclusion rooms with restricted facilities for toileting and hygiene. Service users (4;21;24) and staff (4;30) across four studies viewed seclusion as dehumanising. Both groups across three studies felt it was “degrading” (12, p. 570;30, p. 410) and that service users were “abused” throughout the process (4, p. 251). Three service users likened seclusion to being a caged animal (1;4;13). Conversely, nurses across three studies believed that seclusion maintained a person’s dignity and privacy when they were particularly unwell (2;5) or engaged in embarrassing or humiliating behaviour (20). Several viewed seclusion as often safer and more dignified than chemical restraint or prolonged physical restraint (5;2), highlighting the relativity of perceptions.

Most New Zealand health professionals (4) and all nurses in a Canadian forensic study (13) reported primarily using seclusion for the safety of the secluded individual, other service users and staff, thus being perceived as promoting rights in this regard. Health professionals across six studies (2;4;8;13;23;27) and service users in two forensic studies (13;15) viewed seclusion as helpful or necessary for this reason. Sixty-seven per cent of service users in the USS (17) and some in the Netherlands (11) reported feeling safe during seclusion. Findings from several studies highlighted the perception that seclusion supported the rights of other service users (e.g. 2;20). Seventy-one per cent of Australian nurses (19) and some across various English settings (2) perceived that seclusion benefitted other service users. Eighty-six per cent of a large sample of psychiatrists working in secure services supported the use of seclusion to manage threats of physical violence to others (5).

Three studies explicitly referenced the paradoxical relationship between control and therapeutic care whereby an initial transgression may be considered to ultimately support rights in a broader sense (20;8;13). Some staff felt they had betrayed the service user by secluding them (4) while some nurses felt it was “probably a necessary evil” (2, p. 792). Meehan et al. (2004) reported that 100% of nurses and 67% of service users felt that seclusion “sometimes” or “often” helps service users to calm down, while 100% of nurses and 96% of service users felt that it “sometimes” or “often” also made them feel frustrated. Service users across four studies were retrospectively “glad” or “relieved” that they were secluded, noting that this ensured their safety and that of others (4;10;13;17;18). Further support for a perceived therapeutic benefit to seclusion was expressed by health professionals elsewhere (23;27).

Supports individuals to maintain and develop social, recreational, occupational and vocational activities which are meaningful to the individual:

There was no evidence of this, except for the opportunity for quite time and the perceived paradoxical longer-term benefits to recovery previously discussed. In contrast, abundant evidence discussed elsewhere indicates that seclusion precluded most meaningful activities. One service user in a Canadian forensic service stated that “we all have programmes in this building […] we miss all the programs, we miss our job, we miss outings” (13, p. 204), while another in the Netherlands stated, “when you’re alone, you no longer exist” (28, p. 8):

Instils hope in an individual’s future and ability to live a meaningful life:

Evidence of this was limited to the few service user comments about opportunities to sober up and learn to control one’s emotions and behaviour, as previously discussed. In contrast, findings from numerous studies indicated that seclusion adversely affected some service users’ ability to live meaningful lives, both in the immediate aftermath and after discharge. Frequent feelings of disempowerment, hopelessness and the reawakening of previous traumas are discussed elsewhere. In addition, El-Badri and Mellsop (2008) reported that 30% of service users and 36% of staff described people feeling worthless as opposed to valuable when secluded. Approximately 44% of service users in the US (17) and some in Finland described a fear of re-seclusion (15). Several Dutch previous inpatients said they had not come to terms with their seclusion experience, with one describing difficulties with enclosed spaces, locked doors and the sounds of keys and another having difficulties relating to loved ones (11). An Australian service user noted that the experience “broke my spirit” (24, p. 527) while an English service user experienced ongoing nightmares about seclusion (25). An Australian health-care professional notably acknowledged “feelings of responsibility for possibly traumatising patients” (23).

Dignity and respect

Consists of being courteous, respectful and honest in all interactions:

A small number of service user comments provided evidence of this. Some in the USA felt respected when staff explained the reason for their seclusion and gave them a timeframe for release (7). Up to a quarter US service users also reported being called by their name, kindness, helpfulness, reassurance, a comforting voice and caring smile (21).

The vast majority of findings conversely indicated the absence of this aspect of the principle. Service user perceptions of limited communication, their needs being unmet and being abused, abandoned, treated inhumanely, sedated with medication and punished are discussed elsewhere. Across studies, some service users additionally described staff as “not caring” (1;6;13), showing a “lack of concern” (21, p. 12), having a “negative attitude” (22, p. 9), “patronising” (1, p. 728;28) and being “very rude” (24, p. 527), mean, disrespectful, humiliating (6), hostile (22) and using “threats” (1, p. 727). Some US service users perceived that staff talked about them, mocked them or shouted at them during the seclusion process (6;7). All 20 service users in a single US service felt that staff should be educated or retrained in communication skills (6), with some reporting that staff lacked compassion and empathy. Notably, a member of staff in the Netherlands stated that “I will not hurt him on purpose, but I will jam your wrist in such a way, that as soon as you try as much as kick someone, you will howl with pain” (30, p. 410). Van der Nagel et al. described this as “power perversion” (p. 490), which is in clear opposition to respectful and courteous interactions.

Service users across multiple studies (1;6;7;28) and some staff (16) described seclusion as humiliating. Some service users described eating meals from the floor (21), eating in a dirty and foul-smelling room and being “ordered” to clean the menstrual blood of another service users from the floor (22, p. 10). Most of a US sample of 20 service users were bitter about being accompanied to the bathroom and felt hurried (21). Service users across multiple studies, particularly females, described humiliation, disempowerment and a loss of dignity at being forcefully stripped in front of staff and made to wear hospital gowns (10;12;13;18;21;25;28) and being unable to maintain their hygiene without menstrual products (25). Notably, some staff in New Zealand recognised that service users sometimes felt mistreated, disrespected, abused or inhumanely treated during the process of seclusion (4). Nearly a third of staff in this study perceived that service users felt neglected. All seven service users in one study across English forensic units also felt neglected (3).

Involves sensitivity and respect for each individual, particularly for their values, beliefs and culture:

There was no evidence of this beyond the few examples discussed in the previous section. Abundant evidence in opposition to this is discussed elsewhere and includes little recognition of the uniqueness of individuals or accommodation of their wishes or needs, limited sharing of important information or responsiveness to service user communication and the use of social and physical force.

Challenges discrimination and stigma wherever it exists within our own services or the broader community:

There was no evidence of this. As discussed elsewhere however, some service users described feeling embarrassed, ostracised and stigmatised following seclusion.

Partnership and communication

Acknowledges each individual is an expert on their own life and that recovery involves working in partnership with individuals and their carers to provide support in a way that makes sense to them:

There was no evidence of individuals being acknowledged as experts on their own lives or working in partnership with carers in relation to seclusion. Explicit evidence of partnership with service users was found in only two studies. Australian Nurses described “watching out” for service users, providing counselling and medication and discussing their emotions to support them to maintain control and avoid seclusion (20). A Dutch service user stated: “I have made an arrangement, it’s in black and white. That I will get my medication the moment I am to be secluded” (11, p. 280). Notably however, they added that “it didn’t happen” (p. 280), suggesting that this partnership was unreliable.

Abundant findings illustrated an absence of partnership in relation to seclusion. Findings discussed elsewhere include perceptions of service user disempowerment, abandonment and punishment, the use of social and physical force and a lack of open and respectful communication. Nurses in a Canadian forensic service commented that good therapeutic rapport can reduce the need for seclusion (13); however, Hoekstra et al. (2004) noted that seclusion has a strong negative impact on this relationship and service user trust in staff. Service users in an English service were divided on the helpfulness of seclusion reviews, with some noting a fear of violence during reviews and relief when staff leave again (1).

Values the importance of sharing relevant information and the need to communicate clearly to enable effective engagement:

There was some evidence that information-sharing and clear communication were valued and occurred. A member of staff and a service user in a Canadian forensic service explicitly acknowledged the importance of the service users’ understanding of their seclusion, with one service user adding that it influences their reactions (13). Australian nurses felt that explaining what was happening was “integral” to seclusion (20, p. 560). They reported discussing unacceptable behaviour, explaining what was going to happen, giving reassurance, setting limits, negotiating and parenting (20, p. 560). Some US service users also described having such conversations (14). A Canadian nurse noted they always debriefed after seclusion and they, along with another nurse, reportedly reviewed events and discussed future potential episodes with service users (13). All 106 service users across various Finnish services (15) and some in the USA (7) could explain why they were secluded. Some of the US service users noted that this contributed to better engagement.

Most findings conversely indicated limited communication and information-sharing, and loss of social connection which Allikmets et al. (2020) concluded was central to negative service user experiences. Some Australian service users perceived that they were “generally inadequately informed of ward rules” (p. 372) and were therefore at risk of being secluded (18). Norris and Kennedy (1992) reported that up to 30% did not know why they were secluded, while six other studies presented evidence of limited understanding but did not specify the prevalence (1;4;6;10;13;22;23). Roughly a quarter of service users did not know the conditions under which they would be released (4) and some did not know how long seclusion would last (10;18). The latter issue was also acknowledged by some staff in New Zealand (4). Notably, all ten English service users (3) and “most” service users in various Finnish services reportedly had no debriefing, while 78% perceived a need for this (15). An Australian service user commented: “they could have sat me down and explained why I’d been through all that hell” (18, p. 375).

Service users in three studies reported being dissatisfied with the level of interaction with staff during and following seclusion (18;21;22). Only one-third of Finnish service users from a large sample were satisfied with their level of social contact with staff during seclusion (15). Service users in other studies reported feeling “ignored” (1, p. 727; 7, p. 36; 13, p. 205), cut-off and “estranged” from staff (12, p. 573) during seclusion. One service user noted: “I could not speak to anyone for all that time, when I really needed to talk to someone. It was horrific” (18, p. 375). The lack of information also contributed to suspicion (22) and mistrust (23) of staff.

Involves working in positive and realistic ways with individuals and their carers to help them realise their own hopes, goals and aspirations:

There was no evidence of this.

Evaluating recovery

Ensures and enables continuous evaluation of recovery based practice at several levels:

There was no evidence of this.

Individuals and their carers can track their own progress:

There was no evidence of this. In contrast, staff and service users in two studies perceived decisions as ambiguous (18;22). Muir-Cochrane (1996) indicated that decisions often related at least partially to factors other than the presentation of the service user themselves. Some service users noted that family members were not informed that they were in seclusion (22) or were not permitted to visit them during this time (4), significantly limiting their ability to track progress:

The mental health system reports on key outcomes that indicate recovery including (but not limited to) housing, employment, education and social and family relationships as well as health and well being measures:

There was no evidence of this.

Services demonstrate that they use the individual’s experiences of care to inform quality improvement activities:

There was no evidence of this.

Discussion

Overall, the findings of this review demonstrated that there was limited evidence of the principles of recovery-oriented practice in the published perspectives of health professionals and service users on seclusion. There were very few occasions in which the word “recovery” appeared among the 31 studies, indicating that there was almost no explicit consideration of recovery among staff and service user perceptions and limited consideration by authors. At most, comments by staff and service users provided some evidence in support of particular aspects of individual recovery principles. Notably however, this was consistently accompanied by a larger body of findings indicating that seclusion was perceived as being in direct opposition to these principles. Service user perspectives consistently placed seclusion and recovery at greater odds than staff perspectives.

With respect to the principles of recovery, several points are noteworthy. For the principle “uniqueness of the individual”, there was almost no evidence that the unique needs of service users in relation to seclusion were recognised or that individualised practices were used. However, the breadth of service user experiences discussed demonstrated that they indeed have unique needs. Although some service users reported that seclusion gave them opportunities for valued experiences, none of these were unique to seclusion. In relation to the principle “real choices”, there was no evidence that service users were empowered to make choices about seclusion or that choices were explored. In contrast, most staff and service users perceived that the power to make choices in relation to seclusion lay almost entirely with staff, while service users were greatly disempowered. According to Muir-Cochrane (1996), this raises ethical concerns about the way in which “in control” and “unwanted behaviour” is defined and by whom. While the reality of the risks that sometimes present in inpatient units must be acknowledged and idealism should be avoided, the findings suggest that the values of services and staff may be privileged over those of service users. Ironically, seclusion often precipitated emotional experiences and behaviour that some staff may perceive as evidence of an on-going lack of control, thus extending seclusion and acting as a self-fulfilling prophecy. Such a pattern has previously been described by Holmes et al. (2004).

The majority of the findings indicated that seclusion was perceived to be in opposition to the principle of “attitudes and rights”. This included perceived failures to act upon service user communication, the potential misuse and abuse of seclusion and negative short- and long-term effects on the ability of service users to live meaningful lives. At the same time, both staff and service users perceived that it supported the safety of the individual and other service users. Some staff also considered seclusion therapeutic. Reflecting on such findings, Muir-Cochrane (1996) noted that tighter control measures might fulfil the bureaucratic need of the service to provide a safe environment for everybody, but that this may be at the expense of the individual. The findings again suggest that seclusion was valued relative to the general ward environment and suggest an implicit belief that the right to safety could not be ensured otherwise.

The vast majority of the findings relating to the principle of “dignity and respect” indicated that seclusion was primarily perceived as being in opposition to this, with most of the relevant findings coming from service users. There was no evidence of sensitivity or respect for the values, beliefs and culture of service users or that practices were individualised. Conversely, some staff perceived that seclusion actually maintained a person’s dignity by providing privacy where this may otherwise be compromised due to unusual behaviours associated with poor mental health. Limited evidence was found in support of the principle “partnership and communication” in relation to seclusion. In contrast, there was plentiful evidence for its absence, particularly among service user perceptions. Although information sharing and clear communication was valued by staff in some studies, there was only limited evidence of this transpiring in reality. The vast majority of findings from service users highlight a perception of limited information sharing, poor communication and reduced contact with staff, however.

Very few of the findings contained in the 31 included studies related to the principle of “evaluating recovery”. No aspects of this principle were evident in staff or service user perspectives on seclusion. Instead, the limited relevant findings suggested the absence of any systems to evaluate recovery. There was no evidence that service user experiences informed quality improvement activities, despite apparently frequent adverse experiences. Importantly, this review did not aim to explore systems for tracking recovery or quality improvement activities in relation to seclusion. The studies included predominantly focused on personal perspectives and experiences rather than organisational practices, therefore these findings should be interpreted with caution.

They highlight several points that are relevant to clinical practice, governance and future research. Given that reviews of the qualitative literature provide strong evidence of various detrimental effects of seclusion (Happell and Harrow, 2010; Mellow et al., 2017; Van der Merwe et al., 2013) there is a need to more explicitly consider the impact of seclusion on an individual’s recovery journey. In judging the value, need for and impact of seclusion, there is a need to look beyond behaviour, as recovery is not considered to be synonymous with “symptom” reduction (Clarke et al., 2016). The function of behaviour and the role of meaning-making also shape the lived-experience.

The findings underscore the need to openly acknowledge the reality of conflicting priorities in relation to seclusion and recovery in AMH services. Many of these needs are paradoxical and cannot be easily reconciled, such as care, control and conflicting rights. They can occur both within and across various levels of social systems and include the various needs of secluded individuals, other service users, staff, the mental health system and society more broadly. They are dynamic rather than static over time and need to be satisfied under imperfect and challenging circumstances. This results in ethical dilemmas in clinical practice. Further, Muir-Cochrane et al. (2018) and Jansen et al. (2022) highlighted the moral distress experienced by frontline staff who face these dilemmas while being aware of a growing societal expectation to avoid restrictive practices and engage in recovery-oriented practices. Findings from this study and other research (Mullen et al., 2022) emphasise the active role that staff and systems play in the co-production of social situations. They suggest that revised practices could support the partial reconciliation of seclusion and recovery, where the seclusion is deemed necessary. For example, some recent research suggests that seclusion rooms could be redesigned to make them actively soothing and therapeutic, rather than principally containment spaces (Güsewell et al., 2019; van der Venne et al., 2021). Overall, the findings suggest that in the absence of system-wide changes in AMH services, calls for the complete elimination of seclusion and strict adherence to a recovery model could be somewhat idealistic for some settings.

Reflections on the evidence

The inclusion of a large body of findings from 31 studies has promoted the reliability of this review. The impact of the methodological limitations of individual studies was minimised as each one made only a small contribution to the overall findings and interpretation. Of the 15 qualitative studies included, 12 met all five MMAT quality criteria, indicating good quality overall. No accommodations were made during synthesis and analysis however, as findings from the three remaining qualitative studies did not differ noticeably from the other studies. None of the quantitative non-randomised studies included met all five MMAT criteria. There was no evidence that any of these studies accounted for confounders in the design and analysis, though this had no implications for interpretation as the perspectives were the focus of this research, rather than group differences or influencing factors. The quantitative descriptive studies included met four of the five MMAT criteria, suggesting good quality overall. Varying proportions of staff and service users declined to participate in the non-randomised and qualitative studies, contributing to a risk of bias. Finally, there was considerable variation in quality among the eight mixed-methods studies included in the review. Typically, studies provided insufficient information on the qualitative method used for data collection and analysis in various respects. Care was therefore taken when extracting interpretative comments, however direct findings were taken at face value. This represents a limitation in a considerable proportion of studies including in this review.

The included studies focused either primarily or exclusively on perceptions of seclusion. The absence of collateral reports or records of events to corroborate accounts may raise concerns about the validity of the data. However, data were only included from service users who had personally experienced seclusion and staff that worked in AMH services that had some familiarity with seclusion, therefore perceptions were likely influenced by tangible events related to seclusion. It is important to recognise that the perspectives expressed are contextualised and many factors contributed to the range of perspectives found within and across studies. Generalisations and comparisons should only be made with an understanding of their contextualised nature. Previous research has indicated that contextual factors include: service user history of seclusion (Van der Merwe et al., 2013); the number of people with which a bedroom was shared before seclusion (Stolker et al., 2006); staff’s professional experience and exposure to seclusion (Van der Merwe et al., 2013; Happell and Harrow, 2010); the stage of the seclusion process that staff are referring to as well as their role in any given episode (VanDerNagel et al., 2009); and the passing of time (Keski-Valkama et al., 2010; Van Doeselaar et al., 2008). Mellow et al. (2017) noted that seclusion practices are likely to be influenced by country, culture and time.

Strengths and limitations of the review

The methodology applied in this research possesses a number of strengths and limitations. As noted, using a search string with only three components maximised sensitivity during searches and minimised the risk of relevant literature being missed. The methods were developed in line with international best practice and the review protocol was registered publicly. Two reviewers were used at multiple stages including screening, data extraction and coding to promote validity and quality and enable calculations of inter-rater reliability.

Despite these positives, there are some limitations in the search. For example, one limitation is evident in the decision to only include studies where the full-text manuscript was published in English. However, titles and abstracts published in languages other than English were translated using Google Translate and those that appeared relevant studies are listed in the supplementary material. Notably, all but one of the studies included in this review (Ntsaba and Havenga, 2007) were completed in westernised developed countries. As such, perspectives from less westernised and developed countries remain largely unknown. In addition, variation in reporting the frequency of particular perceptions contributed to difficulties interpreting and synthesising findings from studies and the terms used by authors were incorporated into this review insofar as possible.

An important limitation comes from the fact that, while many of the findings presented in the included studies related explicitly to the Australian recovery principles, none of the studies actively aimed to explore the interface of seclusion and recovery. There was no evidence of explicit and meaningful consideration of recovery by staff and service users in any of the studies. Importantly, studies in which this interface is intentionally explored may produce somewhat different findings. The approach taken was viewed as an effective alternative in the absence of studies that have actively explored this interface. As previously noted, this review did not aim to specifically explore systems for tracking recovery or quality improvement activities in relation to seclusion, therefore findings in relation to the principles of “evaluating recovery” should be interpreted with caution.

Conclusion

This review makes a unique contribution to the academic literature, as no studies to date have explicitly explored the perspectives and experiences of staff and service users on the use of seclusion in the context of recovery-oriented practice. Overall, there was limited evidence of the principles of recovery-oriented practice in the published perspectives of staff and service users on seclusion. Where there was some evidence supporting aspects of individual principles, this was consistently accompanied by a larger body of findings indicating that seclusion was perceived as being in direct opposition to these principles. Notably, service user perspectives consistently placed seclusion and recovery at greater odds than staff perspectives. This is despite the developments mentioned in the Introduction including practices such as Safewards.

The findings have relevance to service user advocacy groups, policymakers and a range of health professions, including clinical psychology. They highlight the importance of nuanced understanding and efforts to conscientiously balance conflicting priorities as best as possible in any given moment to ensure ethical practice, rather than consistently prioritising some needs to the detriment of others. Care must be taken to avoid polarised stances based on overly simplified narratives that may ultimately be detrimental to service user outcomes. In such cases, future clinical practice could suffer from different but equally problematic limitations, leading to calls for yet another ideological revision.

Figures

Figure 1

PRISMA flowchart for inclusion/exclusion

Table 1

Details of included studies

Author(s)	Country	Service type/setting	Methodology	Data collection	Analysis	Participants
1 Allikmets et al. (2020)	England	Psychiatric intensive care unit	Qualitative (phenomenology [inferred])	Structured interview	Phenomenology (inferred) (Miles and Huberman (1994) method)	10 inpatients service users
2 Alty (1997)	England	Psychiatric hospital; various settings (wards [unspecified] and day care services)	Mixed-methods (convergent)	Structured interview with questionnaire	Qualitative: thematic analysis (inferred); Quantitative: descriptive statistics	64 nurses
3 Askew et al. (2020)	England	Rehabilitation and acute assessment wards in a forensic medium secure hospital	Qualitative (phenomenology)	Semi-structured interviews	Interpretative phenomenological analysis	7 inpatient service users
4 El-Badri and Mellsop (2008)	New Zealand	“General adult mental health service”	Mixed-methods (convergent)	Questionnaire	Quantitative: descriptive statistics; Qualitative: content analysis	55 outpatient service users 138 mixed professionals
5 Exworthy et al. (2001)	England	Medium and maximum security hospitals/units	Mixed-methods (convergent)	Questionnaire	Quantitative: descriptive and inferential statistics; qualitative: method unclear)	117 medical professionals (various grades)
6 Ezeobele et al. (2014)	USA	Acute care psychiatric hospital	Qualitative (phenomenology)	Semi-structured interview	Phenomenology (Husserl’s philosophy)	20 inpatient service users
7 Faschingbauer et al. (2013)	USA	Inpatient psychiatric unit(s)	Qualitative (phenomenology)	Unstructured interview	Phenomenology (Van Manen’s approach)	12 inpatient service users
8 Happell and Koehn (2010)	Australia	Acute inpatient units across multiple services	Quantitative description	Questionnaire	Descriptive and inferential statistics	123 nurses (various grades)
9 Happell et al. (2012)	Australia	Inpatient/community mental health service	Quantitative non-randomised	Questionnaire	Descriptive statistics	54 mixed professionals
10 Haw et al. (2011)	England	Forensic psychiatric hospital	Mixed-methods (convergent)	Semi-structured interview; case note review	Quantitative: descriptive and inferential statistics; qualitative: thematic analysis	57 inpatient service users
11 Hoekstra et al. (2004)	The Netherlands	Outpatient mental health service (long-term care)	Qualitative (grounded theory)	Semi-structured interview	Grounded theory	7 previously inpatient service users (inferred)
12 Holmes et al. (2004)	Canada	Psychiatric unit; psychiatric hospital	Qualitative (phenomenology)	Semi-structured interview	Phenomenology (Heideggerian framework/Colaizzi’s approach)/content analysis	6 inpatient service users
13 Holmes et al. (2015)	Canada	Forensic psychiatric service	Qualitative (phenomenology)	Semi-structured interview	Interpretative phenomenological analysis/content analysis	26 staff/inpatient service users (ratio not reported)
14 Kennedy et al. (1994)	USA	Acute adult psychiatric admission unit at a mental hospital	Mixed-methods (convergent [inferred])	Semi-structured [inferred]; file review	Qualitative: method unclear; quantitative: descriptive and inferential statistics	25 inpatient service users
15 Keski-Valkama et al. (2010)	Finland	Forensic psychiatric hospitals and general psychiatric units	Quantitative non-randomised (cross sectional-analytic)	Structured interview	Descriptive and inferential statistics	106 inpatient service users
16 Kuosmanen et al. (2015)	Finland	Psychiatric unit	Qualitative description [inferred]	Semi-structured self-report [inferred]	Inductive content analysis	2 nurses
17 Mann et al. (1993)	USA	Psychiatric unit; community teaching hospital	Quantitative non-randomised (cross-sectional analytic) [inferred]	Questionnaire	Descriptive and inferential statistics	50 inpatient service users
18 Meehan et al. (2000)	Australia	Open acute care units, tertiary mental health facility	Qualitative description (inferred)	Semi-structured interview	Content analysis	12 inpatient service users
19 Meehan et al. (2004)	Australia	Acute unit, general hospital; acute unit and medium secure unit in a tertiary mental health facility	Quantitative non-randomised (cross-sectional analytic)	Questionnaire	Descriptive and inferential statistics	29 inpatient service users 60 nurses (various grades)
20 Muir-Cochrane (1996)	Australia	Closed psychiatric wards in a teaching hospital	Qualitative (grounded theory)	Reflective written report and semi-structured interview	Grounded theory	7 nurses (various grades)
21 Norris and Kennedy (1992)	USA	Multiple psychiatric units in a university hospital for “short term, intensive treatment”	Qualitative description	Questionnaire and interview	Thematic summaries	20 inpatient service users
22 Ntsaba and Havenga (2007)	Lesotho	Psychiatric hospital	Qualitative (phenomenology)	Semi-structured interview, field notes	“Tesch’s method of open coding”	11 inpatient service users
23 Roberts et al. (2009)	Australia	Acute care mental health unit	Mixed-methods (explanatory)	Questionnaire, chart review, focus groups	Quantitative: descriptive statistics; qualitative: thematic analysis	≤12 service users (inferred) Survey: 71 mixed professionals Focus groups: >12 (inferred) mixed professionals
24 Sambrano and Cox (2013)	Australia	Acute mental health services	Qualitative (phenomenology)	Unstructured interview	Phenomenology (Gadamer’s approach)	3 outpatients
25 Scholes et al. (2022)	England	Rehabilitation, low/medium secure and acute units	Qualitative description	Semi-structured interview	Thematic analysis	20 previous/current inpatient service users
26 Stolker et al. (2006)	The Netherlands	Psychiatric hospital; locked ward	Quantitative non-randomised (cross-sectional analytic [inferred])	Questionnaire	Descriptive and Inferential statistics	54 inpatient service users
27 Tooke and Brown (1992)	USA	Short-term psychiatric inpatient unit	Mixed-methods (convergent [inferred])	Questionnaire	Qualitative: method unclear; quantitative: inferential statistics	11 inpatient service users 13 nurses
28 van der Venne et al. (2021)	The Netherlands	Closed forensic units	Mixed-methods (convergent)	Questionnaire and semi-structured interview	Qualitative: descriptive and inferential statistics; qualitative: thematic analysis	52 inpatient service users
29 van Doeselaar et al. (2008)	The Netherlands	Various unspecified mental health services	Quantitative description [inferred]	Questionnaire	Descriptive and inferential statistics, exploratory methods	540 mixed professionals
30 VanDerNagel et al. (2009)	The Netherlands	Psychiatric hospital; closed unit	Qualitative (grounded theory)	Semi-structured interview	Grounded theory	8 nurses/social workers
31 Wynaden et al. (2001)	Australia	Open, locked and forensic units across multiple acute mental health services	Quantitative description	Questionnaire	Descriptive statistics	78 nurses (various grades)

Notes:

Some studies contained data/components that did not meet inclusion criteria. Information in this table relates to data/components that fully met the inclusion criteria

Supplementary material

Supplementary material for this material can be found online.