Mental health practitioner experiences of engaging with service users in community mental health settings: a systematic review of qualitative evidence

Introduction: Effective mental health care is dependent on engaging service users, but some individuals do not actively attend appointments, and may stop engaging with mental health services. Quantitative studies reveal some salient factors that seem to predict engagement but these studies miss the nuances of good clinical practice in this area. A number of qualitative studies of health professionals’ experiences and understanding of effective engagement have been published. Aim: This review aimed to systematically identify, evaluate and synthesise results from these studies with a view to informing effective practice in this area. Methods: Electronic databases Medline, EMBASE, CINAHL, PsychINFO and AMED were searched (PROSPERO systematic review protocol registry (www.crd.york.ac.uk/prospero/; ID CRD42017083976). Of 799 records, ten papers met the inclusion criteria. All papers were subjected to quality appraisal based on the CASP checklist and data systematically extracted. A thematic synthesis of included studies examining mental health practitioners’ experiences of engagement in community mental health settings was conducted. Results: Mental health practitioners see engaging service users as depending upon complex, multi-dimensional phenomena which should include individualised person-centred approaches as well as practical, social and clinical support. Mental health practitioners demonstrate qualities such as determination and adaptability to establish and maintain engagement with service users. A cc ep te d A rt ic le


Introduction and background
Mental health services are largely community oriented and a variety of specialist community services for those with long-term mental health problems have developed over recent years.
These include early intervention services (Bertolote & McGorry, 2005), focusing of psychosis and its prevention; community mental health teams (Sayce, Craig & Boardman, 1991), focusing on those who need enduring support and assertive outreach teams, which specifically target individuals who are otherwise difficult to engage (Sainsbury Centre for Mental Health 1998).
Further developments have seen a shift towards generic referral to specialist community services via a single point of access to ensure timely, appropriate, recovery-based care (Department of Health, 2019). Moreover, some National Health Service (NHS) Trusts have developed Recovery Colleges to compliment community mental health services by actively engaging people with lived experiences of mental health problems in the co-production of recovery-based interventions (Gilburt, 2015;Ebrahim, Glascott, Mayer & Gair, 2018).
For interventions to be effective, they need to be delivered to relevant service users. However, a significant number of those who experience psychosis or other serious mental illnesses and use services (service users) are often challenging to engage (Tait, Birchwood & Trower, 2003;O'Brien, Fahmy & Singh, 2009;Doyle et al., 2014;Tindall, Francey & Hamilton, 2015). Disengagement rates in mental health services are higher than other health services (Mitchell & Selmes, 2007).
Up to 50% of individuals who use mental health services disengage, with adolescents and young people being at particularly high risk (Lal & Malla, 2015). Young people may find it difficult coming to terms with a psychiatric diagnosis due to the associated stigma that surrounds mental ill health and may doubt the usefulness of professional help (O'Brien et al., 2009;Gulliver,

Accepted Article
This article is protected by copyright. All rights reserved Griffiths & Christensen, 2010). Moreover, research has shown that medication side-effects impact upon an individual's willingness to engage (Stanhope, Henwood & Padgett, 2009). Some individuals may disengage from services due to perceiving that medication is being enforced and sustaining mental ill health (Priebe, Watts, Chase & Matanov, 2005). Similarly, individuals may avoid contact with mental health services for fear of being compulsorily detained and treated (Sweeney, Gillard, Wykes & Rose, 2015). However, engagement with specialist services such as Early Intervention Services has also been found to have positive impacts upon a service user's social sense of identity in terms of re-integrating back into society and viewing themselves as a member of a wider community (Loughlin et al., 2020).
However, engagement as a concept lacks a clear definition (O'Brien et al., 2009) and the term is used inconsistently (Bright, Kayes, Worrall & McPherson, 2015;Doyle et al, 2014). Thus the term engagement has been used in a number of ways, including accessing services, retention within services, enthusiasm and self-management, service provision and the interaction between the patient and healthcare provider. Burns and Firn (2002) have suggested that engagement involves a contact between mental health service providers and service users with both parties agreeing that this contact is beneficial. Tait, Birchwood and Trower (2002) developed a Service Engagement Scale (SES) purporting to measure engagement. Items were based on clinical experience and a review of the literature. Although these authors do not define engagement in explicit terms, the SES focusses on availability (client available for arranged appointments), collaboration (client actively participating in the management of mental health problem), help seeking (client seeking help) and treatment adherence (attitude toward taking medication). The SES is a useful, valid and reliable tool for practitioners to identify key areas of concern in terms of service user engagement (Roeg, van de Goor & Garretsen, 2015). It is within the broad terms outlined within the SES that the present study should be considered.

Accepted Article
This article is protected by copyright. All rights reserved Stewart, 2012). This is consistent with the idea that the relationship between the service user and practitioner being a central tenet for a service user's engagement in treatment, both as a stand-alone intervention and as a platform for delivering other interventions (McCabe & Priebe, 2004). A trusting therapeutic relationship depends upon appropriate communication, developing rapport, demonstrating empathy and instilling hope (Adam, Tilley & Pollock, 2003;Shattell, Starr & Thomas, 2007;Stanhope, Henwood & Padgett, 2009). The importance of this two-way relationship as a foundation for engagement is further corroborated by Bright et al., (2015, p.651), in which engagement was 'co-constructed through interpersonal connection' and thus challenging the notion that engagement was only dependent on the service user. Engagement can thus be viewed as a complex, multi-dimensional process that emphasises the relationship between service users and practitioners to work in collaboration towards goals as opposed to just physical attendance at appointments (Tait et al., 2002;Kreyenbuhl, Nossel & Dixon, 2009;O'Brien et al., 2009;Tait et al., 2010;Tindall et al., 2015). Such aspects of the therapeutic alliance in influencing engagement within mental health care must be considered.  (2015) specifically recommend that the wider perspectives of relevant stakeholders such as service providers should be accounted for to further inform strategies to enhance service engagement.
Early scoping searches indicated that some qualitative studies have been undertaken to explore mental health practitioners' perspectives of engagement. However, no known qualitative systematic reviews were identified that synthesised the findings from individual studies around this topic. Systematic reviews of qualitative literature are increasingly regarded as important in evaluating the efficacy of therapeutic approaches in mental health settings (Wood, Burke & Morrison, 2015). Moreover, Thomas and Harden (2008) state that qualitative studies provide

Accepted Article
This article is protected by copyright. All rights reserved important perspectives and should be subject to the same rigour as quantitative studies to examine a specific evidence base. Therefore, this paper presents a systematic review of qualitative studies to further understand mental health practitioner's experiences of engagement with service users across a range of community mental health settings.

Aims
This review aimed to synthesise qualitative evidence regarding mental health practitioners' experiences of engagement. The research question was: 'What are the experiences of mental health practitioners in relation to engaging with service users in community mental health settings?' The SPICE components (Setting, Perspective, Intervention/interest, Comparison, Evaluation) were used to develop the review question (Booth, Papaioannou & Sutton, 2012 showed no previous or present reviews on this topic.

Inclusion and exclusion criteria
Studies were included if they were reported in English and published in a peer-reviewed journal.
Studies had to examine mental health practitioners' experiences of engaging with service users in community mental health settings by using interviews and/or focus groups and utilise qualitative methods of analysis. Studies which used quantitative methods or mixed-method studies where qualitative data could not be separated were excluded. Conference abstracts, reviews, editorials, opinion pieces and grey literature were also excluded.

Accepted Article
This article is protected by copyright. All rights reserved

Search strategy
A comprehensive search of Medline, EMBASE, PsychINFO, CINAHL and AMED was performed in January 2018. The search strategy was developed in consultation with an academic librarian.
Given the paucity of qualitative studies focusing on practitioner engagement in community mental health services, the search terms were broad to ensure that no potential studies were missed. Key search terms were based on SPICE components and used symbols such as truncations and wildcards to capture spelling variations across international studies. Boolean operators such as AND, and OR were used to combine key search words. The limits of peerreviewed journal and English language were applied. The following keywords were used: Additional manual searches of reference lists of included papers were conducted through a process of pearl diving to identify further studies (Booth et al., 2012). Specific date ranges were not imposed to maximise the identification of potentially relevant literature.

Selection of studies
The search led to the identification of 765 articles. An additional 39 records were identified by searching through NHS databases and reference lists of key papers. After removing duplicates, this resulted in 775 papers. Titles and abstracts were screened, and 31 full-text studies were assessed against the review question and inclusion criteria. Ten studies were included in the review and this process is summarised in the PRISMA (preferred reporting items for systematic reviews and meta-analyses -Jakimowicz, Stirling & Duddle, 2015) flow chart in Figure 1. P.H. had

Accepted Article
This article is protected by copyright. All rights reserved responsibility for selecting studies for inclusion. J.B. independently reviewed five randomly included papers and five excluded papers to confirm eligibility. There was 100% agreement between these raters regarding papers for inclusion.

Accepted Article
This article is protected by copyright. All rights reserved studies presented findings in a clear, explicit way and were further illuminated by using participant quotes. Moreover, all studies received a positive final rating in terms of the value of the research to further enhance knowledge and understanding of engagement across community mental health settings. As a result, weaker studies were not excluded following quality assessment to avoid eliminating potentially valuable insights in the synthesis.

Study characteristics
Characteristics of included studies are provided in Table 2

Data synthesis
Thematic synthesis (Thomas and Harden, 2008) was used to analyse and synthesise content across studies. This method has been used in systematic reviews that address questions about lived experiences (Booth et al., 2012). Due to the multi-disciplinary nature of studies included within this review, thematic synthesis was deemed as suitable given its ability to translate the findings across disparate literature into common themes for comparison and analysis (Thomas, Harden & Newman, 2012). In line with guidance set out by Thomas and Harden (2008), this method followed three steps:

Accepted Article
This article is protected by copyright. All rights reserved The first stage of the thematic synthesis conducted by P.H. involved free line-by-line coding of the findings section of every included study. Each sentence and paragraph were carefully read with a view to identify underpinning themes and concepts. Text was highlighted if it was considered to represent mental health practitioner experiences of engagement with service users, with a code being created to summarise its content. A code was represented as a single word (such as 'empathy') or a short phrase ('being person-centred') to summarise and describe a sentence or paragraph of text. In total, forty-five initial codes were developed .
In the second stage, conducted by P.H., codes were juxtaposed and cross-compared across studies with similarities being grouped together to construct descriptive themes. Practitioner quotes were taken from included studies to further support the descriptive themes. The third stage involved the development of analytical themes by collapsing of the descriptive themes.
Descriptive themes with clear commonalities were merged to develop analytical themes.

Reflexivity
This review was primarily conducted by P.H. who has previously worked as a mental health nurse in several community mental health teams. By virtue of this extensive background, a wealth of experience in engagement work has been gained which is a notable strength for the focus of this review. However, to address the potential influence of prior experiences on the data extracted and synthesis, reflexive notes were kept identifying biases and assumptions. Regular discussions with the research team allowed for assumptions to be scrutinised and facilitated ongoing reflection.

Results
Findings from this review indicate that engagement is seen by practitioners as a multi-faceted and complex phenomenon. Nine descriptive themes were identified: 1) "Building rapport so that they can feel safe" 2) "And I go with their choices, because they've got right too" 3) "Showing a more human side to myself" 4) "You actually have to show that person that you are interested in helping them and in what they've got to say" 5) "You can't force someone to like you" "6) "I have got a time restriction I have got to go and see someone else" 7) Anxiety and fear about the unknowingness of engagement work 8) "You mustn't give up on them" 9) "It's about the things

Accepted Article
This article is protected by copyright. All rights reserved you do alongside them". Further development generated three analytical themes which are discussed below: 1) Being authentic based on real dialogue and collaboration 2) Pushing against barriers-engaging against all odds.
3) The chameleon effect-the skill of being adaptable. The findings are structured below according to the analytical themes, along with the descriptive themes. Although the three analytical themes are identified as distinct categories, some overlap was evident. Figure 2 indicates the level of overlap between the three themes.

Being authentic based on real dialogue and collaboration
The first theme emerged as the largest analytical theme as this was informed from four of the nine descriptive themes "Building rapport so that they can feel safe", "And I go with their choices, because they've got rights too", "Showing a more human side to myself" and "You actually have to show that person that you are interested in helping them and in what they've got to say".
It was consistently identified that practitioners attached a high value to engaging as a person rather than a practitioner to facilitate this process. An important finding within this theme is that engaging service users is experienced as more successful when relationships are open, honest and respectful, where collaborative approaches are the norm and where there is mutual trust.
Interestingly, concepts of trust and rapport were strong, recurring descriptions throughout all the reviewed studies and perceived as fundamental for increased engagement.
"Building rapport so that they can feel safe" Building rapport was identified as an integral process for engagement. One practitioner stated:

Accepted Article
This article is protected by copyright. All rights reserved Building rapport so that they can feel safe, so that you're a safe person for them to be with, so that they can start telling you more. Because if they don't feel safe they're not going to tell you hardly anything. (Procter et al. 2015a, p.431).
Three studies demonstrated that when time was invested to build trust and develop rapport, there was evidence of greater help seeking and engagement from service users (Hitch, 2009;Procter et al., 2015a;George et al., 2016).
"And I go with their choices, because they've got rights too" Five studies described how person-centred qualities such as trust and rapport aided collaborative approaches to engagement (Addis & Gamble, 2004;Hitch, 2009;Procter et al., 2015a;Wright et al., 2011;George et al., 2016). Practitioners identified that working at a service user's pace and respecting their choices were essential for successful engagement, exemplified by the following quote: "This person that you are working for, you are working for them you are not deciding on what they should be doing" (Wright et al., 2011, p.828). Moreover, the latter description 'working for' gave emphasis to engagement being a service user-led process rather than necessarily directed by the practitioner.
However, one study found that respecting choices and pacing work also meant that active engagement could be experienced as a lengthy, time-consuming process depending on the service user's perception of time and immediate priorities (Addis & Gamble, 2004).

"Showing a more human side to myself"
Three studies emphasised that being human as well as professional would further encourage greater rapport and trust and influence future engagement (

Pushing against barriers-engaging against all odds
The second theme was the next largest analytical theme being informed from three of the nine descriptive themes "You can't force someone to like you", "I have got a time restriction I have got to go and see someone else", "Anxiety and fear and the unknowingness of engagement work". This theme illustrated that engagement was influenced by several external pressures. It was noted throughout the reviewed literature that practitioners faced organisational pressures to engage with service users which could be perceived as artificial or forced. Similarly, organisational requirements to engage service users, would suggest that this was to the expense of engendering person-centred approaches such as service user choice and autonomy.

"You can't force someone to like you"
Practitioners experienced hostility, ambivalence or rejection in engaging some service users (Addis & Gamble, 2004;Killaspy et al., 2009). Furthermore, they described the reluctance of some service users to engage for reasons including fear, anger, stigma or shame (Addis & Gamble, 2004;Hitch, 2009;Killaspy et al., 2009;Clutterbuck et al., 2009;Procter et al., 2015b;Gairns et al., 2015). Moreover, two studies emphasised that a service user's engagement was hindered by past negative experiences of mental health services rather than being attributable to the practitioners working alongside them (Hitch, 2009;Gairns et al., 2015). However, one study

Accepted Article
This article is protected by copyright. All rights reserved identified how persistent efforts to engage service users placed pressure on staff to form artificial relationships as described by the following quote: "Just because I'm working in assertive outreach…. you can't force someone to like you, and I think sometimes you've got to wrestle with that" (Hitch, 2009, p.487). Equally, another study reported how organisational pressures to engage with service users would lead to practitioners experiencing self-criticism or doubt if engagement was unsuccessful (Addis & Gamble, 2004).

"I have got a time restriction I have got to go and see someone else"
Five studies described how time pressures were regarded to be a major hurdle in terms of engagement (Addis & Gamble, 2004;Coombes & Wratten, 2007;Killaspy et al., 2009;Wright et al., 2011;Gairns et al., 2015). Various workload demands were perceived to impact upon the frequency of engagement with service users as described by the following statement: "I think that you're seeing people fortnightly, but there are people like xx who need more that, and you're just not able to offer it due to time constraints placed on you by large caseloads, chronic caseloads and all the rest" (Killaspy et al., 2009, p.537).
Engaging hard-to-reach service users was also experienced as emotionally demanding. Two studies identified that staff experienced feelings of frustration, despondence or hopelessness when engagement was unsuccessful (Addis & Gamble, 2004;Killaspy et al., 2009). One practitioner stated: "It makes me feel pretty useless at times" (Addis & Gamble, 2004, p.455) Similarly, engagement with service users with dual diagnoses could be experienced as an almost impossible challenge due to the length of time taken to establish trust (Coombes & Wratten, 2007). Furthermore, the issue of practitioner fatigue was also experienced after working intensively to engage service users (Addis & Gamble, 2004).

Anxiety and fear about the unknowingness of engagement work
One study described how possible risk in community settings could trigger anxiety which could create barriers to further engagement with service users (Clutterbuck et al., 2009). Similarly, another study described the management of risk and the impact upon engagement as highlighted by the following quote:

Accepted Article
This article is protected by copyright. All rights reserved

The chameleon effect-the skill of being adaptable
The third theme was the least weighted analytical theme being informed from two of the nine descriptive themes "You mustn't give up on them", "It's about the things you do alongside them". There was a recurring thread throughout the reviewed literature to indicate that practitioners were creative in engaging service users as the process could be experienced as challenging and unpredictable. This theme indicated that engagement with service users was not experienced as a single discrete phenomenon.

"You mustn't give up on them"
Four studies described how using ordinary human qualities and getting to know the service user as a person were viewed as invaluable to engage hard-to-reach service users (Addis & Gamble, 2004;Killaspy et al. 2009;Wright et al., 2011;George et al., 2016). Some practitioners described that there was an element of being human but persistent to allow for active connections. For example: "With a lot of our clients, initially, they don't want any kind of contact with us whatsoever, and we come out regardless of how many times they slam the door in our face. We do it consistently" (George et al., 2016, p.884).
Conversely, one study reported how persistent efforts to engage service users was experienced as "forced" at times due to organisational pressures to maintain contact (Hitch, 2009, p.486).

Accepted Article
This article is protected by copyright. All rights reserved Furthermore, this was perceived to pressure staff to engage with service users who may not want services (Hitch, 2009).

"It's about the things you do alongside them"
Five studies highlighted how practitioners used creative, flexible and sometimes unconventional methods to engage (Addis & Gamble, 2004;Hitch, 2009;Killaspy et al., 2009;Procter et al., 2015a;George et al., 2016). Working alongside service users to provide practical assistance for their other relevant needs beyond their mental health was emphasised as an important engagement strategy. For example: "Really we've used a lot of non-nursing and non-medical ways of engaging her. Going to the cinema…. getting a pair of trainers…one of the really good ways of getting to see her is going to cafes" (Killaspy et al., 2009, p.535).
Highlighted was how practitioners actively worked alongside family members to engage service users as described by the following quote: 'He agreed to go to hospital with his sister, she'd pop inside and talk to us. And it was a rather unusual way of engaging…It was much better than getting the ambulance and police of course.

Discussion
From the outset, we aimed to consolidate practitioners' experiences of engagement in community mental health settings. We identified ten papers for this review and inductively developed three analytical themes. Mental health practitioners identified several barriers and facilitators associated with engagement. However, the review finds that humanistic, personcentred relationships are key to successful engagement. This is in line with the 'Tidal Model' wherein mental health recovery is built upon a genuine human alliance. To engage with an individual rather than manage, treat or otherwise fix their problem is integral to good outcomes (Barker & Buchanan-Barker, 2005). Within the first overarching theme, a trusting, collaborative relationship was identified as the single most important factor in determining whether engagement between practitioners and service users was positive or negative as illustrated by the descriptive theme" And I go with their choices, because they've got rights too". However, although the first theme suggests the emphasis given to the therapeutic alliance, practitioners

Accepted Article
This article is protected by copyright. All rights reserved were also aware that their efforts to maintain contact particularly within AOT contexts could be perceived as coercive or pressuring. This increases the need to further improve understanding of such pressures and how they affect the relationship between service users and services.
A further concept proposed by the Tidal Model is that genuineness is an important feature within engagement (Barker & Buchanan-Barker, 2005). This was paralleled within theme one in which authenticity was central to successful engagement. However, there was a lack of clarification in the reviewed literature about how practitioners maintained professional boundaries. For example, three of the ten studies highlighted how practitioners adopted role duality by sharing aspects of their personal self (Addis & Gamble, 2004;Procter et al., 2015b;George et al., 2016). It could be argued that some practitioners based on previous experiences may place more emphasis to the human aspect of their interactions with service users to establish a more common ground and encourage successful engagement. This is consistent with Egan (2014) who argues that practitioners are more effective when they adopt a position of being themselves.
However, whilst maintaining professional boundaries are necessary in mental health settings, purposeful human interactions are regarded as key qualities to effectively engage service users in meaningful partnerships (Clarke & Walsh, 2009). Likewise, the emphasis on ordinary conversation taking place between the professional and the service user may re-shape the relationship, enabling engagement to be transformative and generate new trains of thought rather than just exchanging facts (Zeldin, 2000).
The Tidal Model acknowledges the emotional challenges faced by some practitioners in that he/she may risk subconsciously addressing his/her own feelings rather than those of the service user which may lead to distant forms of care and engagement (Barker & Buchanan-Barker, 2005).
The Tidal Model further re-iterates the importance of practitioners accessing debriefing to reevaluate their human qualities rather than their skills to effectively engage with individuals (Barker & Buchanan-Barker, 2005). This is an ongoing consideration for community mental health practices as within the second overarching theme, practitioners often felt pressured as illustrated by its descriptive theme "I have a got a time restriction, I have got to go and see someone else".
Although practitioners had the skills to engage, levels of anxiety and worry due to time pressures, larger caseloads, and risk management considerations could hinder engagement. This

Accepted Article
This article is protected by copyright. All rights reserved is consistent with Kielhofner et al. (2002) in that emotions are important to understanding engagement experiences. However, although 'pushing against barriers' identified varying pressures that practitioners faced to engage service users, it was also perceived to result in engagement that was more outcome-orientated than collaborative. It is suggested that this approach may place further pressure on service users to engage with services with subsequent loss of autonomy and feelings of powerlessness (Priebe et al., 2005). Consequently, further research has suggested that individuals with mental health problems may disengage if they cannot see benefit from the service or if there is a sense of loss of control over their own lives (Kreyenbuhl et al., 2009). Moreover, although practitioners in the reviewed studies did not explicitly describe mandated community treatment such as Community Treatment Orders, it is claimed that such approaches can create further barriers to contact with services (Sweeney et al., 2015). In contrast, mental health practices that embrace service user autonomy and decision making are suggested to facilitate greater engagement (Priebe et al., 2005).
A further consideration within the Tidal Model is its notion of the 'helpful helper' (Barker & Buchanan-Barker, 2005, p.134). In this instance, it is argued that the practitioner as 'helpful helper' is aware that there is not a one-size-fits-all approach and to only do what needs to be done to meet the service user's immediate needs. Furthermore, this practical approach recognises human experience as a fluid, dynamic process that acknowledges the importance of critical, short-term needs as well as longer-term developmental care to support recovery (Barker & Buchanan-Barker, 2010). However, there was no explicit discussion of how a practitioner's role was defined within the reviewed literature. Notwithstanding, a role within mental health teams is regarded to be a multi-dimensional one with practitioners taking on several different roles that overlap with those of other professional groups (Newbigging, 2004). Indeed, it was consistent within the third theme that engagement was experienced as successful when practitioners were flexible whether it be the social, practical, economic or clinical needs of the service user. The latter point is supported by Repper and Perkins (2003) in that practitioners are pivotal to linking service users to other services or providing practical help to reduce social exclusion. Practitioners working within AOT contexts would take on multiple positions/roles that appeared to be similar to advocacy, family work and social support depending on individual needs of service users (Addis & Gamble, 2004;Hitch, 2009;Wright et al., 2011;George et al., 2016).

Accepted Article
This article is protected by copyright. All rights reserved AOTs have been found to be more successful than CMHTs in engaging service users due to smaller caseloads (Killaspy, 2007;Killaspy et al., 2009). However, some care must be taken in considering the wider organisational impacts of such services. Although AOTs have been mostly welcomed by service users and their families, it could be argued by their 'assertive' nature that this leads to different engagement strategies being used including persuasion to compulsion (Molodynski, Rugkasa & Burns, 2010). Within the descriptive theme "You mustn't give up on them", practitioners were critical of the assertive outreach model on engagement as this was generally perceived as pressuring staff to make connections with service users who may not want services (Hitch, 2009). Hence, there is a potential tension between building a therapeutic alliance based on mutual trust with the need to manage risk and potentially implement mandated community treatment orders. The latter point needs further explicit acknowledgement to allow realistic expectations for mental health staff and service users.

Limitations
Upon application of the CASP quality appraisal tool, lower quality studies were included in this review to further allow for richer insights of engagement to be understood as experienced by practitioners across community mental health settings. However, although this review specifically sought to highlight practitioner perspectives, it acknowledged that engagement may be experienced and understood differently by service users. It is possible that practitioners may experience genuineness, flexibility and collaboration as effective qualities when service users are agreeable to engagement. However, many widely held assertions about engagement are not based upon consistent evidence (O' Brien et al., 2009;Doyle et al., 2015). Given that a universal definition of engagement is not fully agreed upon, more detailed investigation is needed to further understand how engagement works for individuals who do not fully accept services.
The potential influence of the lead author's experiences as a mental health nurse in community mental health services are acknowledged. The analysis and synthesis of the qualitative evidence being conducted by the first author only is acknowledged as a further limitation of the review process and thus may introduce reviewer's bias. However, regular and frank discussions with the research team, who are from clinical and non-clinical backgrounds, throughout the review process were used to minimise the potential for individual assumptions and biases to influence

Accepted Article
This article is protected by copyright. All rights reserved the review's overall findings. Another limitation is the low number of studies included in the review for thematic synthesis. Although Thomas and Harden (2008)  A final limitation is that the engagement experiences of carers, families and other service users within mental health services were not considered. Given the importance of the two-way relationship between staff, service users and families/carers, future research could consider integrating findings regarding engagement from these group's perspectives.

Implications for research and practice
The findings indicate that mental health professionals see engagement with service users as complex, multi-dimensional, person-centred and includes practical and social approaches as well as clinical interventions. Due to an increasing emphasis within contemporary nurse education around collaborative, person-centred and relationship building approaches (Nursing and Midwifery Council, 2019), registered mental health nurses and other professionals would benefit from further systematic guidance around engagement strategies.
Notwithstanding, it is important to be mindful of the applicability of the review's findings to wider service contexts as most included studies focused on engagement in AOT or CMHT settings. Furthermore, there has been a reduction in the specialist remit of AOTs due to the dismantling of such services or integration of some functions into CMHTs (Firn et al., 2013;Gilburt, 2015). Although EISs have retained a distinct function, some of these services have also been integrated into CMHTs and now provide care for people aged between 14 and 65 years (NHS England, 2016). This has important implications for practitioners working within

Accepted Article
This article is protected by copyright. All rights reserved traditionally youth-focused services such as EIS in utilising engagement skills that best fit with people of varying ages. Additionally, less is known about the engagement experiences of practitioners working within specialist settings such as EIS. Given the paucity of qualitative evidence that focuses on the engagement practices of EIS practitioners (Tindall, Simmons, Allott & Hamilton, 2018), there is a need to gain deeper insights of such experiences within this setting.
Future research should aim to qualitatively explore EIS practitioner's experiences of engagement for which it is hoped will add to the knowledge-base about what contributes to successful engagement.

Conclusion
This review has aimed to understand mental health practitioners' experiences of engagement with service users with the intention of providing actionable knowledge for successful engagement approaches across a range of community mental health settings. The findings confirm the importance that mental health practitioners place on being person-centred, collaborative and creative with service users in a range of community mental health settings to enable successful engagement. Furthermore, it has been identified how practitioners ensure that engagement is not solely focused on clinical interventions but that also addresses social and practical needs. However, the professional challenge is being able to maintain engagement in a context characterised by issues including time pressures, larger caseloads and risk management considerations and the service user's perception of time and priorities.

Screening Eligibility
Full-text articles excluded, with reasons (n = 21) 2=Not based in community mental health settings 1= Review 6= Quantitative design 2=Not primary qualitative study 1= Service user views only 4= Does not explore engagement in community mental health settings 2=Explores the process of engagement and networking between multiple community agencies rather than with service users 3=Mental health practitioner perspectives of engagement with service users are not explicit