Archard, P.J. and Murphy, David (2015) A practice research study concerning homeless service user involvement with a programme of social support work delivered in a specialized psychological trauma service. Journal of Psychiatric (6).

Introduction : This article presents a qualitative ‘practice research’ study into a pilot programme of social support work delivered in a specialised psychological trauma service to homeless service users that was grounded in non-directive, person-centred approach and staffed by student social workers. Homeless people are a population known to be highly vulnerable to trauma, in triggering events to becoming homeless and the considerable social isolation, discrimination, and adversity suffered when homeless. Currently, there is a paucity of research into mental health service delivery to homeless persons and the influence it imparts in individual lives. Research aim/question : T o research a person-centred support work programme as it was received by service users domiciled in supported housing for homeless persons, encompassing experiencing the programme, worker-service user engagement and contextual influences bearing upon positive outcomes. Method : Narrative interviews gathered the impressions of service users and support workers. The data arising from these interviews was analysed thematically. Results : The service user participants valued support work that combined practical and relational elements, however would have preferred a longer-term involvement. They also spoke of feelings of disconnection and estrangement from others including their peers in the supported housing facility in which they resided. The worker participants valued the flexibility of person-centred work tailored to service users’ individual needs but reflected service user concerns around the short-term nature of their involvement.. The learning experience provided opportunities for developing ability to work with marginalised ‘hard-2 to-reach’ client groups. Implications for practice : Psychiatric nurses carrying out, or supervising, mental health support work with homeless service users should be mindful of potential impact of temporary staffing arrangements and the effects on continuity of care. The perceived benefit of being person-centred and tailoring work to client’s practical needs can help foster rapport and trust. acknowledge the feelings of disconnection and estrangement from other homeless persons living in temporary, supported or hostel type accommodation can experience.


Introduction
This article presents an exploratory, qualitative study into a programme of social support work delivered in a specialized traumatic stress service.
A 'practice research' project (Salisbury Statement on Practice Research 2009; Peake & Epstein 2004), the study arose from the concerns of staffs at a traumatic stress service and had the advancement of professional practice and service development in mind. The study aim was to explore the support work programme as it was received by service users domiciled in supported housing for homeless persons, encompassing experiencing the programme, worker-service user engagement, and contextual influences bearing upon positive outcomes. A narrative interview method was used. Interviews were carried out with service users and support workers (comprising student social workers on placement at the service).
We present the study here for the attention of psychiatric nurses and other mental health professionals to help fill a research gap into mental health service delivery to homeless persons and the influence it imparts in individual lives (Bhui et al. 2006;Taylor 2012;Williams & Stickley 2011). There is currently only a limited amount presently known about non-directive social support provision in individual trauma services (Murphy et al. 2012(Murphy et al. , 2013. The article is presented in line with Tong et al.'s (2007) 'consolidated criteria for reporting qualitative research'. We begin by contextualising the study in a brief overview of the support work programme, including the rationale for piloting the programme and nature of the support work. Thereafter, we present the research methodology and findings, discuss the implications of these findings by reference to other 4 relevant empirical and practice literature, and conclude by setting out some possible directions for more extensive research that might be carried out in comparable initiatives.

Programme overview
The support work programme was introduced at the traumatic stress service to augment psychotherapeutic interventions: to more comprehensively attend to the psychological and social needs of service users, particularly to improve capacity to offer early intervention and outreach help, and help over time; and to provide a more joined up approach to the education of unqualified mental health professionals in the field of psychological trauma.
The programme was open to some statutory service users i.e. those carrying a formal or suspected diagnosis of a trauma-related psychiatric condition and/or referral by statutory health services. However, the majority of service users accessed the programme via outreach work with local organisations serving community populations for whom trauma can be considered part and parcel of daily life, but for whom the uptake of mainstream mental health services remains low, for example refugee groups, sex workers and women suffering domestic abuse. The service worked alongside local providers of supported accommodation for, and outreach work with, long-term homeless and vulnerably domiciled persons on the basis that homeless people are known to be highly vulnerable to trauma, in events triggering periods of homelessness and the considerable social isolation and adversity suffered when homeless (see Glasgow Homeless Network 2003;Goodman et al. 1991;McNaugthon 2008;Scott 1993;Williams & Stickley 2011). In turn, these providers were a source of a number of referrals to the programme. Support work involved providing practical assistance with a range of everyday tasks, benefit and grant applications, the seeking of accommodation with service users, home visits and meetings with family and support networks, and liaison with other 5 agencies and making of referrals, where necessary, to them. The support work was grounded in a principled non-directive (Grant 1990) approach to person-centred care (Rogers 1957). Goals for support were always set by the service user and the support worker worked to facilitate support in the direction the service user desired to progress and care was relationship-based with workers carrying small caseloads alongside whom they could invest considerable professional energy.
Simply stated, a person-centred approach is presupposed by a view of human life that people are intrinsically motivated toward being social constructive. It envisions people as situated in a process of actualisation which is either facilitated or inhibited by contextual factors, notably the presence or absence of an acceptant interpersonal environment. In this vein, psychological distress can be conceived of as a manifestation of impeded actualisation that results in a discrepancy between the experience of the total human organism and that portion of experience that actualizes as the self-concept. This discrepancy between self-concept and organismic experience can be overcome given the right interpersonal and social conditions are in place. This contrasts the philosophical underpinnings of a behavioural approach, which conceives of the client as a product of learned behaviours, and a psychodynamic one which holds a more 'tragic' view of human life in its assumption of some degree of innate destructiveness. Also, a person-centred approach is phenomenological and entails, generally speaking, an emphasis on the perceived reality of a person which contrasts a psychodynamic approach's concern with unconscious motivation and early biographical experience. Psychodynamic approaches are concerned with reflective insight whereas person-centred approaches concern is with actualisation processes. Both psychodynamic and person-centred approaches share a view of the worker-service user relationship as fundamental to the facilitation of positive change and emphasise the need to meet service users' distress with empathic sensitivity (see Gomez 2005, Howe 1995) which contrasts a behavioural approach's emphasis on practical strategies to alter learned behaviours.
Prior to embarking on support work with individual caseloads, students who fulfilled this role undertook a self-directed and facilitated training package. This introduced them to fundamentals of 'trauma work', therapeutic relationship skills in the person-centred approach, the assessment of post-trauma psychological sequelae, philosophical approaches to understanding trauma, different treatment modalities for posttraumatic stress and the role of premorbid and personality factors and beliefs and assumptions in the development and maintenance of traumatic reactions. Student progress was assessed by a resident psychologist at the service who had undertaken training to act as a practice assessor. Students were assessed on a portfolio of reflective work submitted to the practice assessor and academic staff from students' professional training programmes and presentations they made about their work to other professionals based at the centre. The practice assessor carried out formal and informal observations of student practice with service users on their caseloads. Students were continually invited to self-assess their and their peers' practice and development as personcentred trauma support workers.

Methodology
In total, six participants were interviewed for the study. The four service users who took part were all male (on account of residing in a male-only facility), of white British descent and aged between 45 and 58 years, a make-up reflecting some of the typical identity characteristics expected of the UK homeless shelter/hostel resident (see Marshall 1996).
Status as trauma survivor was defined as in Coulter (2011) by involvement with specialist provision. Periods spent living rough, in temporary hostel dwelling, and 'couch surfing' at 7 friends or acquaintances on a short-term basis were taken into account. The two support worker participants were female and had successfully completed placements at the time of interview. One support worker was in the final year of her social work training. The other worker had begun to practice in a qualified post as a social worker.
A convenience sampling strategy was used. Interviews were undertaken by the first author who is white European and male and has prior professional experience working with the homeless community and trauma survivors. At the time of carrying out the interviews, the first author was a trainee social worker in his mid-twenties in the final year of his qualifying course. Service user participants were accessed by way of contact with the organisation running the supported accommodation in which they resided. Two residents' meetings were attended to present the research aims and reasons for carrying out the study, gauge interest in participation and develop familiarity and trust with the participants. Four service user participants volunteered for the study from five living at the supported accommodation who had engaged with the support work programme. No participants dropped out during the research process.
Ethical approval for the study was granted by the University Department Research Governance and Ethics Review Board under the auspices of whom the study was undertaken. Participants were made fully aware of the research aims and intentions, and were free to withdraw at any time. Informed consent forms were completed in all cases and a number of procedures, not discussed here, were put into place to safeguard participants' psychological welfare.
A qualitative approach was utilised due to the study's investment in the perspectives of the subjects of study and a intention to capture the subtleties of the processes of professional helping and the 'cultural whole' of service users' social worlds more broadly (Schwandt & Burgon 2006). Of the eight narrative interviews carried out: 8 two were with two support workers and six with four service users including two follow up interviews with two participants to elaborate on themes nascent in the first six interviews. Interviews took place in service premises and participants' residences with only participant and interviewer present during each interview. Each interview lasted somewhere between 50 and 80 minutes. Interviews sought participants' experiences of working together and any development they felt was made in consequence. To be participant centred and generate experiential data indexed in actual events, these were loosely structured, and a predominantly non-directive stance was adopted by the interviewer. Questions were relayed in an economical way and worded in an open manner, supportive observational prompts and paraphrasing were used, and topics were covered in an order contingent upon what participants spoke of (see Hollway and Jefferson [2000] for overview of a similar interview approach). Two pilot interviews were carried out prior to the eight presented here. Given the majority of the service user participants acknowledged difficulties with literacy, interview transcripts were not returned to participants. Interviews were mostly audio-recorded, though two were documented in note-form due to too much noise occurring outside two service users' residences to satisfactorily audio-record.
The interview transcripts and notes were subject to a thematic analysis. This was informed by what is known as 'the framework method' (see Ritchie & Spencer 1994;Gale et al. 2013) and involved the following steps. First, all recorded interviews were transcribed and familiarisation with the data took place. This involved both listening to the interviews again and reading and re-reading interview transcripts. Each interview was individually coded on a line by line basis. Through this coding procedure, an initially large number of patterns and minor themes were identified across cases, from which gradually smaller sub-themes and the more substantive main themes presented here were generated.
These themes were both spontaneous and related to the research aim. In appraising each of main themes, individual participant contributions were weighed up against one another as were service user impressions with regard to material from support workers and vice versa.
The coding of the transcripts was carried out by the first author whilst the interviews were ongoing and the grouping and later establishment of the themes presented was verified with the second author who worked on the study in a supervisory capacity. The coherence of the analysis according to the interview transcripts was also verified by a third health and social care researcher independent to the study.

Results
Where appropriate, quotes are inserted in the main themes summarised below for illustrative purposes. Each quote is accompanied by a participant identification (SU=service user; SW=support worker) and interview number (I1=interview 1; I2=interview 2).

Rapport, commitment, flexibility and worker autonomy and practical assistance
To put it at its broadest, what the service user participants perceived to be beneficial about the involvement of support workers related to the particular role taken by the workers and the manner in which this role was implemented.
Service users appreciated the face-to-face time support workers could, with relatively modest caseloads, invest in working alongside them. This was contrasted with the amount of time shelter staff could commit, who, whilst cordial and warm, were recognized as juggling the demands of a number of residents and administrative duties. 'I found it very useful to talk to them (...) because they sat and listened and the [shelter] staff don't always have the time, haven't always got the time to sit and listen to us' (SU4, I1).
The effort and energy put into visiting and spending time with them was seen as essentially altruistic, and a personal commitment to them. 'You've got to remember they've got their lives to live as well, their problems and that and they come to work and then have to struggle to listen to us lot' (SU4, I1). Furthermore, the provision of a listening and understanding ear without passing judgment or insisting on a particular course of action while sharing one's difficulties appeared to be experienced by some as emotionally cathartic and to a certain extent, something of a motivational force in catalysing hopes for personal adjustment and change. A description of the support workers as 'fantastic', for example, was accompanied with the remark that 'philosophically they helped me in my head; it was like get it together +++ [refers to own name], you know' (SU2, I1).
During the early stages of intervention, connections between workers and service users appeared to be most strongly predicated on rapport built during workers active involvement with practical matters e.g. spending time in residences and helping with shopping and cleaning, of giving lifts, accompanying service users to professional meetings and court appearances. A piece of work recounted by one support worker surrounded an isolated young pregnant female suffering a foetal demise and with whom contact was made via an agency which provided signposting and healthcare advice to persons involved in street based sex-work. Escorting this service user to hospital appointments at different stages throughout medical treatment -for which the client lacked transport arrangements-her manifest presence, and witnessing what the client was going through, all formed part of 'support[ing] her emotionally through that' (SW2, I1). Recurrent in service user narratives was appreciation for the role workers played in a recent wholesale relocation to newly built premises. Specifically, in this scenario, valued was their help in packing, sorting, and un-packing belongings, as well as helping some service users, down the line, to acclimatise to this transition and become more autonomous and domestically skilled, for example, by '... showing them how to use the washing machine, showing them how to use a cooker, (...) helping them to tidy up' (SU4, I1) and take on greater responsibility for paying rent and individual bills with the new more independent living arrangements.
For the support worker participants, the proximity to service users' everyday realities was seen as characterising a key point of difference between the practice of support workers and the more formal, consulting room confined activities of resident psychology professionals. 'I wasn't just sitting in a room with her, talking with her (I: right). I was saying come on we're going out to such and such. We're doing this, we're doing that' (SW2, I1). This type of flexible work was largely possible, the support workers emphasized, as a result of the independence they experienced and freedom to work outside usual office hours and away from the office, really getting to know service users, for listening to what they wished to derive from involvement and spending prolonged periods

Truncated involvement and incongruities in professional status
Although much was perceived as beneficial by the service users, there were also aspects of involvement that caused the service user participants some concern. Workers' youth (being 'only kids': SU2, I1), comparative to their own identity as middle aged homeless men (described in one instance as 'having been round the block'), was perceived to compromise their ability to adequately relate to their situations and life-histories by two service users, and apprehension about workers' status as professionals 'in-training' rather than 'qualified' was volunteered in one case. Albeit this concern appeared to coincide with an acknowledgement that the status of the support worker as apprentice in the arena of professional helping ('only just starting out') came with a fresh enthusiasm for the work and lack of an aloofness that might be met in other health and social care professionals. Notably, it was awkward disclosing personal information about oneself and attaining a certain rapport and closeness, only for the worker-user relationship to then cease rather abruptly. In certain ways, the problematic quality of this brevity of involvement seemed to be precipitated by service users' adaptation, through the agencies and housing organisations they were involved with, to contact with trainee professionals and agency staff who would only be involved for, essentially, a periodic befriending.
You tell them about your personal life and then they, I mean I've nothing to hide (...) and they know all that; but then they go away and then don't come back, and then someone else comes and they take their place and then they expect you to tell them all over again 13 and then they move on (SU4, I1).
This view conspired with one support worker's frustration of the inherent limitations of minimal periods of involvement. Fleshing out what started as a 'patchy' skeletal understanding of a service user's situation and personal history invariably 'took a fair bit of time' as did brokering a connection with service users accessed through outreach work, who were commonly quite sceptical about involvement (SU1, I1).

Residential arrangements, peer relationships and everyday adversities
In terms of the broader lived experience of the service user participants and its bearing on the support work carried out, a number of things emerged during interviews. The aforementioned move in accommodation was talked of frequently and appeared to have made for a significant transition for most, and the shift from a domestic arrangement of single rooms with shared domestic areas to self-contained flats, mixed feelings. On the one hand, there was some enthusiasm for the greater privacy and space this entailed and evidence of a pride taken in individual flats, particularly as for some this was the first time they had a place of their own. The move was also linked, by some, to opportunities to reengage with former interests, of woodwork, reading and helping at a local church and charity shops. Conversely, though, the increasing independence appeared to have also given rise to fears around tenancy sustainment, for example by falling behind with bills, with a sense that access to temporary accommodation was more difficult now than it had been previously, and that theirs was a diminished capacity, at an established age, to cope with the challenges inherent to a street-based existence. 'I'm not up to going on the streets again. I'm not up to it. I won't survive out there. I know I wouldn't' (SU2, I2). For two service users, there was some preoccupation with the area surrounding the facility (metropolitan, deprived and a high crime rate), of being 'too frightened to open the door' or 'look outside' during the night, and unease with security measures at the facility and potential for other resident's behaviour such as leaving the back door unlocked to lead to incursions, as was suspected to be the case in a recent theft of eighty pounds from one participant's flat (SU3, I1).
Accounts of relationships with neighbouring residents and cohabitees were evoked, at times, in terms of fraternal aspects, of mutual history and companionship, and activities engaged in such as watching films and going for walks together as well as supporting each other in times of need: '+++ [name of other resident] pops up now and again for a chat like, which I like, because me and him go back a long time' (SU2, I1). One service user (SU4: I2) spoke of their role in caring for a 'vulnerable' friend who had been temporarily lodging with him, and who was, for example, put to bed at night when inebriated, cooked for, or bought clothes when in a state of disrepair. However, there was also evidence of distrust and frustration, including suggestions of maintaining a low profile in the presence of peers: 'some of them I watch me tongue with' (SU1, I1), and annoyance with other's standards of personal hygiene. Individual flats were branded as 'disgusting' and 'filthy' and a group outing to a local restaurant was brought up as a source of embarrassment on account of being in the company of residents who 'stank' and were 'eating with they fingers' (SU4, I1). Furthermore, the move appeared to have factored in a subduing of peer relationships (particularly with the lack of a common area) and cooking for/with one another and sharing meals, and there was some concern as to what this shift would mean with regard to arrangements at Christmas and other special occasions.
During interviews, all service users mentioned personal histories of considerable adversity, of relationship breakdown and family estrangement, job-loss, self-harm, grief and loss, substantive periods of street-living and rough sleeping and serving custodial sentences and spells of inpatient psychiatric treatment. They all additionally spoke of some form of ongoing, daily form substance/alcohol misuse and accompanying repercussions. (SU1, I1). In addition, professional and social roles occupied prior to becoming homeless tended to be referred to by the service users as ex -as opposed to extant-identities e.g. 'I used to be a scaffolder' (SU3, I1). Though clearly connected to the fact that they were not, actually involved in such work at present, this tendency could also potentially be related to feelings of 'not counting' like the rest of the population as homeless person or hostel dweller (Harris [1991] and Chard et al. [2009] contemplate this tendency in the life narratives of homeless persons).
With these contextual dimensions of service users' lives in mind, it was interesting to note support workers reflections on the rehearsal of a distinction well-known to trauma professionals between service users suffering trauma of a compound type, i.e. homeless persons and others who 'have years and years of trauma and were still living in trauma' (SW2, I1), and those accessing specialist provision via statutory channels for generally isolated events, e.g. accidents and traumatic bereavements. In line with common assumptions, the latter's professional and family support networks and often reasonable monetary income was perceived to be more straightforwardly conducive to enduring change than 'fighting the tide' with the latter group against the persistent circumstantial and historical currents in their lives. However, whilst there was some frustration inherent to their descriptions of working alongside homeless service users and other hard to reach groups with the latter group, both workers characterised this work as having been beneficial for them as fledgling mental health professionals. Coming face-to-face with the level of the adversity faced by these service users, combined with the reflective space provided during placement fostered a healthy confrontation with preconceptions around professional potency and a sense of professional maturation.
The workers indicated that, as a result of carrying out this work, they felt they had a better awareness of how, amongst other things, alcohol and substance use served as a means of 'brief solace' (SW1, 1), self-medication and desensitizer to emotional pain ('It was just easier to live their lives with the heroin or the crack than without it, because their lives were just so difficult. (...) it made it possible in all they were facing' SW2, I1), and had arrived at a 'less idealistic', more realistic approach to practice and valuing of workeruser partnership in their work. This approach was said to involve a kind of 'there but for the grace of God go I' attitude, and commitment to 'standing alongside' service users rather than emphasizing difference and dysfunction. It incorporated an acceptance that were one subject to similarly detrimental circumstances and life-histories, these would be felt as similarly painful and difficult to withstand and likely lead to succumbing to great distress and/or personal withdrawal (SW1, I1), and an appreciation that improvements made in cases with entrenched problems were, most often, hard-won and modest, but meaningful for service users nonetheless.

Discussion
The study's modest sample means our findings are not necessarily reliably representative of all the homeless service users who came into contact with the support work programme or those who practiced on it. With this caveat acknowledged, what the study does offer is an involved impression of a particular type of social support work. In considering the subjective experience of dwelling in supported living accommodation and the personal histories of homeless service users, the study also sheds light on the psychological impact of homelessness and professional need to acknowledge the various ways in which homeless persons living in temporary, supported or hostel type accommodation can feel 'homeless at home' and feelings and anxieties about being disconnected/estranged from others and themselves can remain for homeless persons (Kirkpatrick & Byrne 2009;Partis 2003;Peled & Muzicant 2008;Riggs & Coyle 2002).
In the UK, the current National Institute for Health and Clinical Excellence (NICE) (2005) guidelines for the treatment of posttraumatic stress acknowledge the reciprocally negative impact trauma can have on trust in relationships with others. In 'situations in which people do not experience a neat termination of a traumatic experience' the guidance recommends the use of Herman's (1997) community based intervention framework and that professionals advocate for meeting social support needs, and involve social-service and voluntary sector organisations. This is a positive strategy in terms of the needs of homeless service users, albeit the specific ways in which this strategy is put into practice with this population needs more specification than is currently included in the guidelines.
Homeless service users are recognised as presenting a number of challenges to community mental health provision and are known to frustrate professionals with wariness about, or resistance to, engagement. It has been recommended that those working with the homeless endeavour to minimise the hoops service users must jump through to access services, allow informal time with homeless clients and offer gestures of concrete assistance in order to forge working partnerships, for instance visiting rough sleepers in places they are known to congregate and, where appropriate, to supply them with essentials such as dry clothes and food (cf. Kuhlman 1996;Marshall & Bhugra 1996). This study, like others with homeless adults accessing mental health provision from the UK and elsewhere (Taylor 2012;Bhui et al. 2006, Thompson et al. 2004, Padgett & Henwood 2012, points to the need for provision that is staffed by those with a personal commitment to working with homeless service users in their environments and integrates practically orientated and relationship based features in seeking opportunities for 'doing and talking' and engaging in talk about the more personal and emotional aspects of service users lives. What this study adds to this research is an account of how this support work might be provided using a person-centred approach and by trainee professionals, and under the auspices of a specialised psychological trauma service. Ballatt and Campling (2011) and Kanter (1990Kanter ( , 2000 emphasise that a regularity of visits, rituals of arriving and going, and meeting spaces of support work can, within a broader package of intervention, contribute to the lessening of service user anxieties about involvement. That is to say in offering a 'holding environment', the predictable and robustly reliably setting created is less liable to provoke or antagonize negative affective states, and can aid in empathizing with service user states of mind, particularly concerning the effect of day-to-day realities on emotional wellbeing. Support workers may even be able to provide, what Lemma (2010) describes as, a kind of 'ad hoc, "on the move" therapy' (p. 414).On the basis of on his own experiences as a mental health support worker, Benjamin (2011) argues that, at their best, support workers can exhibit qualities of transparent relating, empathy and positive self-regard, and thus do things which converge with those of a trained therapist. Yet, unlike the consulting room, support work has the advantage of taking place in contexts which provide opportunities for a depth of talk with users who might otherwise be unsure about psychiatric involvement. This view resonates with the sentiments of psychotherapeutic professionals who have attempted to develop therapeutic provision for homeless populations and described the difficulties they have encountered in engaging homeless service users in clinical work who are preoccupied by more pressing practical concerns and put off by the traditional hour-long counselling or therapy appointment which is characterised as formal and distancing (see, e.g. Kuhlman 1996, Brown et al. 2011, Bentley 1997. Though this may be so, more thought is required to bring support work and psychotherapy more fully into communion, whether via person-centred or other theoretical approaches, and the implications of this for developing a model of support work for practice with homeless service users suffering psychological trauma. Browne et al. (2012) comment that articulating what is specifically meant by a 'therapeutic relationship' outside a psychotherapeutic setting is difficult. The therapeutic relationship is something often evoked in the psychiatric nursing literature, even to which the profession's identity is tied, but it is, nevertheless, a 'difficult to define entity' (p. 841) of which, they say, there appears to be a limited understanding of what it actually looks like in practice beyond a set of general principles. Moreover, as Barker and Pistrang (2002) point out, despite commonalities, there is limited amount of comparative scholarship and empirical exploration of the processes of each vis-à-vis the other that, at present, exists, and therapists may be reluctant to regard their clinical work as transposable to support work and a more general (i.e. run-of-the-mill) manifestation of psychological helping.
The service users participating in this study appeared to benefit from the nondirective person-centred approach implemented. This is based on the premise of Rogers' 20 (1957) therapeutic relationship conditions being both necessary and sufficient for constructive behaviour change to occur and is particularly useful as it is assumed that the service user has the capacity to direct their own life when the right socio-environmental conditions are available. The support worker participants, as fledgling health and social care professionals appeared to benefit from this approach also, as well as the placement experience overall. They worked under the supervision of other professionals, but with a certain freedom to progress their own professional niche, gaining a greater sense of professional identity and led them to a greater valuing of the worker-user partnership as a means of enabling positive change and the importance of starting intervention with a service user's expressed concerns. The person-centred approach is just that, an approach, and is not confined to therapy. The relationship variables lend themselves to being adopted by any area of professional practice or voluntary work and need not be the property of the 'credentialed' professional, they can and indeed have been be applied to civic life (Rogers, 1973).
One recommendation that can be taken from our study by services attempting to implement similar programmes is that whilst student staffing can be beneficial in terms of the enthusiasm for the work that students can bring, the transient nature of their support can disrupt continuity of care. With a present focus on 'cost-effectiveness', including the valorisation of randomized control trial evidence and standardised, targeted, short-term interventions, and the commissioning of services by virtue of client throughput, extended forms of social support can be a difficult provision to maintain even when they yield sustainable outcomes with vulnerable community populations (see e.g. Coulter 2011, Haigh 2005, Pilgrim et al. 2009). Acknowledging this, though, for the service user participants, the truncated nature of involvement seemed to be particularly felt given students worked solely for the duration of a six month placement period and the number of 21 different helpers the homeless service users tended to have contact with. Services implementing similar programmes should contemplate the extent such a staffing arrangement serves service users' best interests. If engagement is to be short term, as a result of funding restrictions, or other organisational constraints, professionals need to strive to avoid recapitulating (and replicating pain derived from) previous losses and estrangements, and reinforcing negative feelings damaged service users have about themselves and may displace onto professional relationships such as being unwanted or prone to rejection (see Huntley 2002). Rendering endings as palatable as possible by being prescient about and sensitively communicating timescales for withdrawal or worker alteration from the outset is something that should be planned in advance. Worker awareness around service user anxieties around separation could, we would suggest, be supported by a therapeutic style of supervision that would provide a personally experienced and attuned model of helping for the supervisee to emulate in their front-line practice, occasion for workers to air their personal anxieties and feelings and avoid burnout, and for workers to build their professional acumen by way of 'scaffolding' consultant expert knowledge on to their own (see Duggins & Veitch 2013, Bennett 2008).

Further research
The qualitative and practice orientated paradigm of this study is suited to examining the subtleties inherent to the strengths and weaknesses of psychosocial initiatives and anchoring exploratory evaluative work in the particular professional and social context of its implementation (Gould 2006). To expand the research data we have been able to provide here, and further understand the lived reality of this type of support work engagement and its function over time, one option open to comparable initiatives in the future would be to carry out ethnographic observations alongside a more extensive batch 22 of longitudinal interviews at different junctures tracing the course of involvement (see Longhofer et al. [2013: Ch.3] for discussion of this type of project). Another less intensive option would be to engage support workers and service users in extensive daily recordings (for example in reflective diaries written or audio-recorded) to chart, over a number of months in detail, their own experiences of involvement and personal and professional development, and carry out cross-case analyses of the data generated therein. The former option would, likely, involve external researchers, whereas the latter would conceivably be partly undertaken by staff themselves and potentially have the consequence of a more effective transfer of research knowledge to service delivery given the personal investment in knowledge production it would entail.

Implications for practice
In light of the study findings and ensuing discussion, psychiatric nurses, and other mental health professionals, should be mindful of the following when developing support work provision for homeless service users suffering psychological trauma.
 Support work with homeless services is at its most beneficial when combining practical and relational elements, and begins with the homeless person's everyday needs and expressed concerns.
 Working in the community with homeless service user and hostel dwellers potentiates scope to empathise with the particular circumstantial constraints bearing on a service user's life and mental health and an opportunity to build rapport and trust in a milieu that is not as formal and alienating as clinical settings can be.
 The high incidence of psychological trauma and relationship difficulties in the lives of homeless persons means they can be particularly sensitive to issues around concluding intervention and separation. Thus, in planning and carrying out support 23 work, professionals should be mindful of the need for worker continuity and the appropriate management of endings.
 Professionals should also consider how evaluation of this type of support work initiative might be evaluated, researched within a practice-orientated paradigm and via approaches that practitioners are able to have a sense of investment in knowledge production.