Notes
Article history
The research reported in this issue of the journal was funded by the HS&DR programme or one of its proceeding programmes as project number 09/1001/51 The contractual start date was in April 2011. The final report began editorial review in January 2013 and was accepted for publication in July 2013. The authors have been wholly responsible for all data collection, analysis and interpretation, and for writing up their work. The HS&DR editors and production house have tried to ensure the accuracy of the authors’ report and would like to thank the reviewers for their constructive comments on the final report document. However, they do not accept liability for damages or losses arising from material published in this report.
Permissions
Copyright statement
© Queen’s Printer and Controller of HMSO 2014. This work was produced by Sweeney et al. under the terms of a commissioning contract issued by the Secretary of State for Health. This issue may be freely reproduced for the purposes of private research and study and extracts (or indeed, the full report) may be included in professional journals provided that suitable acknowledgement is made and the reproduction is not associated with any form of advertising. Applications for commercial reproduction should be addressed to: NIHR Journals Library, National Institute for Health Research, Evaluation, Trials and Studies Coordinating Centre, Alpha House, University of Southampton Science Park, Southampton SO16 7NS, UK.
Chapter 1 Background
Inpatient care as the most problematic component of the mental health system
Community mental health services are increasingly well developed, but acute hospital care remains a central component of the mental health system. 1 We still do not know very much about what constitutes effective inpatient care, nor are the aims of hospital admissions very clearly defined beyond the management of immediate risks. 2 Consequently ‘there is a sense that hospital care is a black box, with people being admitted and discharged, but with little known about what happens to them while they are there’. 3
Mental health problems are one of the biggest cost pressures on the health service. In 2011/12 the government spent £5.5B in direct mental health-care provision for those aged 18–64 years, of which 30% was spent on acute services and a further 11% on access and crisis services. 4 Successive governments have emphasised the importance of giving priority to service users’ experiences of, and satisfaction with, services in the development, provision and auditing of mental health services. 5 It is vital that service users’ experiences of, and satisfaction with, acute care shape the delivery of these costly services.
Studies consistently find considerable service user dissatisfaction with care received during acute admissions. 6 In a survey for Mind, of its members’ experiences of inpatient units, Baker found that around half of the respondents experienced hospital wards as non-therapeutic and felt that the ward had negatively impacted on their mental health. 7 In 2004, Mind conducted a further survey alongside qualitative research with service users. 8 Just 20% of respondents felt that ward staff treated them with dignity and respect. A high proportion of respondents said that staff did not interact sufficiently with them, and that inadequate staffing numbers and overuse of agency staff compromised safety and the therapeutic value of the ward. Many found acute wards frightening places;8 this finding has been confirmed just as emphatically by a more recent report into acute care by Mind,9 and in the evidence gathered by the Schizophrenia Commission. 10
Weich and colleagues11 explored the role of ethnicity in the experience of a range of mental health-care settings. They found that the only place where ethnicity had a mediating effect was inpatient wards.
As power imbalances are most explicit in in-patient settings, it is perhaps unsurprising that this should undermine individual care experiences. This may explain why those who have been admitted to hospital are the least satisfied with mental health services.
They also found that, ‘in-patient care was extremely unpopular – irrespective of ethnicity’. Problems included service users’ loss of control, loss of liberty, witnessing high levels of disturbance and poor relationships with other service users. These problems were exacerbated by conflictual relationships with staff. The team identified service users’ lack of power as a key barrier to developing therapeutic relationships and concluded:
Remedial interventions to improve therapeutic relationships by directly addressing power imbalances in hospital settings should be developed and evaluated.
Defining the therapeutic relationship in mental health services research
Although the notion of ‘therapeutic relationships’ originates in psychotherapy,12 the idea has gained currency in mental health services research over the past decade, with authors describing such relationships as ‘the centre piece of psychiatric practice’. 13 A therapeutic relationship can be defined as:
A collaborative bond between [service users and clinicians] (Greenson, 1967), with the most frequently researched model operationalizing the concept as based on shared ‘goals’, ‘bonds’ and ‘tasks’.
Bordin, 197914
Models of therapeutic relationships tend to additionally consider the personal qualities that clinicians and service users contribute to the relationship, such as warmth and motivation. 14 Consequently, the main measure of therapeutic relationships used in mental health services research, the STAR-P (Scale To Assess the Therapeutic Relationship-Patient),15 assesses both personal qualities and goals and tasks:
Overall, the scale covers goals, tasks, openness, trust and unconditional positive regard. The clinician version also uses the ideas of rapport, listening, sharing a good relationship, sharing expectations, openness and trust.
Catty et al. 16
The importance of the therapeutic relationship has been demonstrated in research which has found associations with outcomes,13 and in studies which have found that service users cite high-quality therapeutic relationships as among their highest priorities for mental health services. 13,17,18
Although the term ‘therapeutic relationship’ is often used synonymously with ‘therapeutic alliance’, Catty19 has argued that the former should be adopted as a generic descriptor of constructive relationships between service users and clinicians in mental health services research. The latter term, therapeutic alliance, is poorly defined in psychiatry but tends to refer to a specific aspect of therapeutic relationships. Our working definition of therapeutic alliance, based on a reading of Catty,19 is the affective bonds between service users and clinicians in which service users feel valued and supported as individuals. We have defined therapeutic relationships as the collaborative bonds between service users and staff based on shared goals and positive personal qualities. In this study we have explored therapeutic relationships in order to explore what it is about such relationships that leads to positive therapeutic alliances.
Disappointment with therapeutic relationships in inpatient settings
A recurring theme in qualitative investigations of the difficulties in meeting service user needs and providing positive experiences in inpatient settings is of disappointment with therapeutic relationships in these settings. In an in-depth qualitative study, Stenhouse20 analysed narrative interviews with 13 service users in order to understand the experience of inpatient care. She found that on admission, an expectation was created that service users would receive help with their difficulties through interaction with nurses, and that they would be able to talk to nurses at any time. However, service users subsequently found that nurses were often too busy to talk to them. This ‘dissonance between expectation and experience’ was frustrating for service users. Owing to the lack of support from nurses, service users often supported and counselled one another. Although this could lead to a sense of camaraderie, it could also mean service users taking on, and being burdened by, someone else’s problems at an intensely vulnerable time.
Cleary and colleagues21 reviewed the international qualitative literature on nurse–service user interaction in acute mental health units. A common finding across studies was that people value the time they spend with nurses but do not have enough time together and consequently, relationships between service users and nurses can be superficial. Similarly, the highly unpredictable and demanding nature of acute wards can result in therapeutically superficial care; this can prevent staff from treating service users as individuals,22,23 and means that staff do not always recognise service users’ basic humanity. 21 This finding was echoed in a Mind independent inquiry into acute and crisis mental health care:9
What people overwhelmingly want is to be treated in a warm, caring, respectful way irrespective of the circumstances in which they come into contact with services. In other words, all of us would like to be treated with humanity. Some do this but many of the experiences the panel heard suggested that mental health services have lost touch with basic humane principles when dealing with people in crisis – as shown by dirty wards, lack of human contact, a lack of respect often bordering on rudeness by staff, and a reliance on force. This does not produce the relationships and conditions that help people recover.
The inquiry recommended that action is taken to ensure that acute care is based on humane values including warmth, care and respect. Research with service users has found that service users expect good inpatient care to include formal and informal opportunities to talk to empathic staff and develop relationships with them. 24 Stenhouse20 found that service users want staff to initiate conversations with and make time for them; service users also value effective communication, trust, and feeling listened to and understood. 25 In their review of the UK literature, Quirk and Lelliott3 found that a high value is placed on staff being ‘active listeners’ and demonstrating basic ‘humane qualities’ including ‘empathy, tolerance and respect’. Similarly, in their review of the international literature, Cleary and colleagues21 defined an integrated triad of ‘listening-understanding-responding’ as essential to building therapeutic relationships. The personal qualities they identified as important to service users included a sense of humour, patience, perseverance, calmness, respect, imagination and being non-judgemental.
There are a number of barriers to achieving therapeutic relationships between staff and service users. The recent Schizophrenia Commission found that staff are often overwhelmed, feel unable to provide basic levels of care and support, and have little time to spend with service users. 10 Totman and colleagues26 found that staff identify a number of barriers to therapeutic relationships including: bureaucratic demands; staff shortages; lack of staff continuity; and uncertainty about which therapeutic models to use. Additionally, Stenhouse20 reports that numerous studies have found that the time nurses can give to working therapeutically is limited due to external constraints such as staffing levels and the prioritisation of other tasks. This can be interpreted by service users as disinterest. Stenhouse20 concluded that staff need to be aware of this, manage expectations and communicate effectively in order to avoid frustrations.
Alternatives to standard acute inpatient care
Given the above, it is unsurprising that the recent Schizophrenia Commission found that reforming acute and crisis care was the highest priority of service users and front-line workers. 10 One strategy for improving acute care has been to establish residential alternatives to acute wards in community settings, often called crisis houses. The Alternatives Study (TAS), the National Institute for Health Research (NIHR) Service Delivery and Organisation (SDO) study which was the precursor to the current investigation, examined nationwide availability of crisis houses. In 2006, 40 such services were operating in England, offering a range of service models. 27 They are typically smaller than inpatient wards, with no more than 10 beds, and are often based in converted houses in residential streets. Crisis houses vary in the degree of medical cover provided and the extent to which they employ qualified mental health staff. They also vary in the extent to which they adopt a medical model of mental health and service provision. The UK has both voluntary sector and NHS-run services: regardless of who manages them, most are highly integrated with the local NHS mental health system, with crisis resolution teams very often providing additional support, gatekeeping beds and, in some cases, even managing the services. Populations served by these community crisis residential services overlap considerably with standard acute wards, with high proportions having previous histories of acute hospital admission and with current experiences of psychosis, although few admit people detained under the Mental Health Act. 28
Despite a long history dating back at least to the 1960s, relatively little research has examined such alternatives. However, a recent systematic review of predominantly US studies suggested that service user satisfaction may be greater than with standard wards. 29 One reason may be that, as suggested in a paper from a qualitative study of service user experiences of a women’s crisis house and a neighbouring hospital, ‘crisis house staff are seen as more informal, approachable and thought to have more time than hospital staff’. 30 More recently, Lichtenberg31 discussed alternative residential models of crisis care. He described how the quality of contact between staff and service users is given priority, and there is a ‘supportive and caring milieu’ with ‘intense and regular therapeutic contact with staff’. He also found a de-emphasis on medication, a destigmatising approach to psychosis, and a home-like atmosphere. A crucial aspect of crisis residential alternatives is that compulsion, force, restraint and seclusion are not routine aspects of ‘treatment’.
Explaining greater satisfaction in crisis house settings
The NIHR SDO programme commissioned TAS in response to the need for new evidence about residential alternatives to standard acute inpatient care to inform NHS policy and service planning. 27–29,32–37 It was a multisite mixed-methods investigation of residential alternatives to standard acute inpatient care. The study followed previous research in confirming substantially greater service user satisfaction at community-based crisis houses than on acute wards. A number of aspects of service provision identified as potential explanations for greater service user satisfaction in crisis houses were examined. These candidate variables included differences in amount of staff–service user contact, in content of care, and in service outcomes. However, quantitative analyses examining these variables found no significant differences between crisis house and hospital settings.
Qualitative interviews with service users suggested that the quality of relationships with staff might be key to service users’ experiences of admission and to their satisfaction. Relationships with peers, coercion, safety and exposure to very disturbed service users also emerged as important, and ratings of ward atmosphere suggested less anger and aggression in crisis houses than acute wards. However, therapeutic relationships with staff and peer relationships were not explored quantitatively in TAS, so the hypothesis that they might be important in accounting for greater satisfaction in these settings could not be investigated. An editorial discussing findings from TAS recommended the investigation of these areas as a next step in research on crisis houses. 38 Similarly, Cleary and colleagues21 recommended that future research should focus on the factors which affect a therapeutic relationship and how the setting is related to these conditions.
Assessment of therapeutic relationships
Measures of therapeutic relationships typically focus on the extent of goal-orientated collaboration between an individual therapist/worker and the service user. The relationship can be rated from service user and staff perspectives. Initially developed for use in psychotherapy settings, a psychometrically robust measure of therapeutic relationships has now been developed specifically for use in mental health service settings. 15 Research in community settings has found associations between service user-rated therapeutic relationships, engagement with services and service user satisfaction. 39 A study conducted two decades ago in Scandinavia indicated that good therapeutic relationships may be correlated with better outcomes at discharge from hospital. 40 However, there has been little systematic investigation of therapeutic relationships on inpatient wards, nor have empirical comparisons been made between crisis residential and standard inpatient services.
One finding from TAS was that considerable difficulties had been encountered in implementing and sustaining innovative working practice practices (e.g. the Tidal Model41 specifically designed to improve the quantity of contact and quality of relationships between staff and service users). The difficulties often encountered in implementing new initiatives in health settings have been well documented. 42 Successful initiatives to enhance staff–service user relationships in acute settings are likely to be facilitated by a greater understanding of the factors that affect their quality. We have not, however, found substantial studies exploring the contextual and organisational factors that promote good relationships and therapeutic relationships in acute mental health settings. The current study addresses these gaps. It aims to provide new evidence on the critical ingredients of alternatives – explaining service users’ preference for them – and to enhance overall understanding of how to optimise care and service user experiences in inpatient settings.
Chapter 2 Aims and objectives
In this study, we aimed to explore potential mechanisms underlying a key finding from TAS: that service users are more satisfied with crisis residential services than with standard acute wards. Our qualitative data suggested that the key differentiating variable in satisfaction might be the strength and quality of therapeutic relationships. Consequently, our aim was to investigate service users’ experiences of relationships with staff in community-based crisis residential services that offer an alternative to acute hospital admission and in standard acute mental health wards. We explicitly aimed to explore the factors that influence therapeutic relationships and experiences of admission in both settings. The specific objectives, addressed using a mixture of quantitative and qualitative methods, were:
-
to test the hypothesis that stronger therapeutic relationships are achieved in community crisis residential alternatives (such as crisis houses) than in standard acute psychiatric wards
-
to understand factors that impede and facilitate good staff–service user relationships in acute settings
-
to develop a model to explore the factors that influence service user satisfaction with acute services, exploring how far stronger therapeutic relationships may explain our previous finding that service users are more satisfied with crisis residential services than with standard acute inpatient care.
Chapter 3 Methods
Design and conceptual framework
Service user satisfaction is influenced by contextual and compositional factors. 43 Compositional explanations focus on differences in service user characteristics whereas contextual explanations call attention to organisational, social, cultural and physical factors of the care setting. This study assessed a number of contextual and compositional factors not measured in the original TAS, their relationship to service user satisfaction and whether or not they can explain greater service user satisfaction in crisis houses. The primary contextual factor on which we focussed was relationships experienced by service users during their stay. The main relationships that service users have in psychiatric inpatient settings are with staff and with other service users. Staff–service user relationships involve both goal-orientated engagement, assessed by measures of therapeutic relationships, and more informal engagement.
Our primary hypothesis was that therapeutic alliances between staff and service users are stronger in crisis houses than in standard inpatient settings. As well as measuring this, we compared peer relationships, recovery and experience of specific negative events between the two settings. These are all factors that emerged from TAS, especially its qualitative components, as potentially important to service user experiences, but that were not empirically measured in the original study. In this study, we explored the influence of these factors in a model of service user satisfaction and tested their explanatory value in understanding the mechanisms underlying greater satisfaction in crisis houses.
This study thus addresses a key question to which we were not able to provide a clear answer in the original TAS: why service users are more satisfied with crisis houses than standard inpatient services? This greater satisfaction did not appear in the original study to relate to greater improvement in symptoms or social functioning: the alternatives were characterised by shorter stays during which improvement was less than for hospital admissions. There were also few differences in the amount of staff–service user contact, or in the content of care between wards and crisis houses, and these variables did not appear to account for differences in satisfaction. An exploratory model of satisfaction in TAS found that three-quarters of the variance in satisfaction remained unexplained with a model containing service users’ characteristics, the amount of contact with staff during admission or the types of interventions provided.
The qualitative elements of this study are intended to further enhance our understanding of therapeutic alliances by yielding evidence on factors which influence the quality of relationships. Qualitative data were collected to explore both service user and staff perspectives on the factors that impede or enhance relationships. We elicited experiences and perspectives on positive and negative relationships and factors that impede and facilitate better staff–service user relationships, as well as contextual factors such as service organisation and management, environment, atmosphere and culture, staff teams, procedures, protocols and routines, and staff training and support, and if and how these influence therapeutic relationships.
Service user involvement throughout the study
The study was a collaboration between service user, clinical, qualitative and mixed-methods researchers. As described, the study derived from previous research into acute residential services, which included a service user-led study. 34 A highly experienced service user researcher was involved in study design, and the qualitative interview guide derived in part from themes identified as important to service users in the service user-led study. The main study researchers responsible for collecting both quantitative and qualitative study data were service users. Participants in the qualitative element of the study were generally aware of this, and we believe this enhanced the quality of the data obtained by increasing the candour of respondents and reducing inhibitions about commenting negatively about services or relationships with staff. The service user researchers were also involved in the analysis of qualitative data, enhancing the validity of the data analysis by ensuring that a service user perspective contributed to data interpretation. 44 Overall, we feel that the meaningful involvement of service user researchers within a multidisciplinary team has enhanced the quality and relevance of our findings.
Collaboration with the Protected Engagement Time study
This study is closely linked with a study funded by the NIHR Research for Patient Benefit Programme on Protected Engagement Time (PET) (PB-PG-0808-17014) (lead Fiona Nolan). The PET study is an evaluation of an initiative, PET,45,46 designed to increase the quantity and quality of staff–service user interaction in inpatient wards. PET involves ring-fencing time during which ward staff are required to focus solely on service user contact for fixed periods of the day: visitors are not admitted to the service and administrative duties and meetings are not allowed. For the PET study, data were collected from service users in acute inpatient services in three sites: Norfolk, Camden and Islington, and East London. Our study utilised quantitative data from the PET study for acute wards in the Camden and Islington and East London NHS Foundation Trusts, with which we compared data collected using the same measures from four crisis houses in the same Trusts. This allowed a comparison of therapeutic alliances and factors influencing service user satisfaction at crisis houses and acute wards. There was a large overlap between research teams for the two studies, and we co-ordinated measures and planned sites for the two studies. We sought consent from participants to share data between the two studies. This enabled us to collect data from a larger number of participants than would otherwise have been possible within the study’s resources. The value of both studies was thus enhanced by this collaboration, allowing linked investigations of therapeutic alliances in different acute settings and of the impact on service users’ experiences both of an innovation within standard inpatient wards and of community alternatives.
Setting/context
Quantitative and qualitative data were collected at four residential alternatives, all crisis houses, in Camden and Islington and East London NHS Foundation Trusts and (as part of the PET study) on 16 local acute inpatient wards in the same Trusts. The crisis houses comprised a range of service models, including statutory services staffed mainly by qualified mental health clinicians and voluntary-sector run services employing mainly social care staff. One service where we planned to collect data closed during the study period: we substituted another relatively recently opened service in the same Trust. The services were:
-
a long-established crisis house provided by an NHS mental health trust for women only, with a particular specialism in working with people with personality disorders
-
a long-established crisis house commissioned by the local mental health trust and local authority from a voluntary sector organisation that is mainly active in the provision of sheltered housing. This was the only service to have been included both in the original TAS and in the present study.
-
a crisis flat provided by an NHS mental health trust and very closely linked to the local crisis resolution team which gatekeeps its beds (this has been established for around 5 years)
-
a recently opened crisis house jointly provided by a voluntary sector organisation and local mental health trust, with close links with the local crisis resolution team.
The comparison acute wards were local acute inpatient wards participating in the PET study in the same Trusts; we used quantitative data collected in the course of the PET study for our comparison between crisis houses and acute wards. Our initial intention was to compare crisis houses and those PET study wards which were not using PET but were providing standard acute inpatient care. However, as investigation into the provision of PET revealed a greater similarity between standard acute and PET wards than anticipated we were able to also include wards offering PET in the comparison. Some qualitative data for our study were also collected on standard acute wards participating in the PET study.
Quantitative component of the study
Aim
Aims of the quantitative component of the study were to test the hypothesis that stronger therapeutic alliances are achieved in community crisis residential alternatives (such as crisis houses) compared with standard acute psychiatric wards, and to develop a model of factors that influence service user satisfaction with acute services, exploring how far stronger therapeutic alliances may explain our previous finding that service users are more satisfied with crisis residential services than with standard acute inpatient care.
Sampling and inclusion criteria
Regarding required sample size, we calculated that a sample of 85 service users per arm would provide 90% power to detect a medium standardised effect size of 0.5 at the 5% significance level. The initial sample size calculation was based on recruitment from four wards and four crisis houses. When an intracluster correlation of 0.01 was filtered into the calculation, to account for clustering, it was estimated that approximately 27 service users per cluster were needed overall, giving a design effect of 1.28 and a cluster-adjusted sample of 108 per arm [calculations carried out in Stata Version 10.1 (StataCorp LP, College Station, TX, USA) using sampsi and sampclus commands]. As the study in fact recruited individuals from a greater number of clusters, the power for the comparisons will have been greater than assumed in this power calculation.
The inclusion criteria for the quantitative arm of The Alternatives Study 2 (TAS-2) were:
-
level of English sufficient to complete the study assessments without linguistic difficulty
-
capacity to provide informed consent
-
resident in the crisis house for at least 5 days (in one site with a particularly short average stay) or 1 week (all remaining crisis residential alternative sites).
The PET study, with which we pooled data, yielded interviews with 247 participants from acute wards, selected using similar criteria except that the minimum length of stay was 2 weeks – adopting this criterion in the crisis house setting would have resulted in a markedly unrepresentative sample due to shorter length of stay in crisis houses. Both service users detained in hospital under the 1983 Mental Health Act47 and voluntary service users were included in the PET study.
There were no diagnostic inclusion or exclusion criteria from either arm of the study. The usual admissions policies of the participating studies resulted in a sample that was largely aged between 18 and 64 years. Both participating mental health trusts had inner London catchment areas characterised by high levels of ethnic diversity and considerable social deprivation alongside pockets of considerable affluence.
Recruitment and consent
A senior staff member on duty at each service identified service users who were suitable for interview. Service staff initially approached service users to ask about their willingness to participate in the study. A study researcher then met each person who expressed willingness to be approached. The researcher provided an information sheet and a full verbal explanation of the study, and answered any questions that arose. Where service users were willing to participate, their written informed consent was sought. Service users’ consent was also sought for a study researcher to access their clinical notes; however, service users were able to refuse this access and still participate in the study. Our aim during the data collection was as far as possible to interview all capable and consenting service users who were admitted to the participating services during the study data collection period for the minimum length of time required by the study. The number of service users participating and declining to participate were recorded by researchers. Data collection continued until the target number of participants had been recruited from each service. Participants were paid £15 each in cash on completion of study measures as an acknowledgement of their contribution to the study.
Measures
The following measures were employed both in crisis houses and – via the PET study – on acute wards:
-
Therapeutic relationships STAR-P is a 12-item measure comprising three subscales that measure collaboration, positive clinician input, and non-supportive clinician input. It has been developed specifically for rating clinician–service user relationships in mental health settings and good psychometric properties have been established. 15 Relationships were rated separately for three staff nominated by the service user as important in their care. Several service users at one crisis house only felt able to complete STAR-P ratings for two staff members. The primary outcome measure for the study was the mean of all STAR-P ratings completed by each service user. Our intention was that this would provide an overall measure of service users’ relationships with staff, allowing for variation in service users’ relationships with different staff members.
-
Satisfaction with services The Client Satisfaction Questionnaire (CSQ)48 is an eight-item instrument which provides a global measure of service user satisfaction with services. It has been widely used in mental health service settings and has good psychometric properties. It was also used in the original TAS as the main measure of service user satisfaction in crisis houses and hospitals.
-
Peer support Peer support from other residents was measured using an adaptation of the Interpersonal Relationship Inventory (IPRI). 49 The IPRI was developed as a measure of informal social support designed to be used in a range of health settings. It has demonstrated acceptable psychometric properties. In order to gauge support received by service users from other service users, we used the 26 items from the support and conflict subscales of the IPRI, which elicit information from participants regarding their relationships with other users of the same mental health service.
-
Recovery The Recovery Assessment Scale (RAS),50 a 41-item measure with established psychometric properties, quantifies concepts derived from the recovery model including hope, empowerment, connection and self-identity. In keeping with this conceptualisation of recovery,51 the RAS focuses not on changes in symptoms or social impairment, but on whole-person improvement from the perspective of the service user, with the domains measured being those that service users tend to prioritise.
-
Negative events during admission A checklist was developed for the current and PET study, measuring negative experiences during admission, some related to staff, others to service users. A specific instrument was drafted for the study as we were unable to find a suitable measure. Sources were the types of event described by participants in the qualitative component of TAS as having a negative impact on their experience of admission, and input from the PET and TAS teams, who include service users, inpatient, crisis house and community clinicians, and researchers. The resulting checklist was piloted before the study and modifications made based on service users’ comments, but no formal testing of psychometric properties was possible in the context of this or the PET study. Items included a range of negative events involving staff, including restraint, forced medication, assaults or threats by staff, and being deliberately ignored, and events involving other service users including violence or intimidation, or pressure to use drugs. Service users were asked whether or not they had experienced these events, and how severe the impact on them had been.
-
Participants’ sociodemographic characteristics Participants were asked their sex, age, ethnicity and country of birth, and also about their diagnosis and history of admission, including compulsory admissions under the Mental Health Act. 47 Where participants consented, data were extracted from clinical records to confirm length of stay, diagnosis, Mental Health Act status and previous admission history; length of stay was available at all services but records kept in some crisis houses did not consistently yield the other data.
Data management and analysis
Researchers entered data from completed forms onto SPSS databases (Statistical Product and Service Solutions; SPSS Inc., Chicago, IL, USA). The same data templates were used by TAS-2 and PET study researchers. Following entry, a second researcher checked the data against the paper sources. The database for crisis houses (this study) and acute wards (the PET study data with the East Anglia wards participating in the PET study excluded) were then merged electronically, with data cleaning carried out by checking frequency distributions for all main variables in the resulting merged database. Data were then transferred into Stata 12 statistical software (StataCorp LP, College Station, TX, USA) in order to carry out analyses.
Tables of descriptive statistics for TAS-2 and PET samples were first constructed. Linear regression with therapeutic relationship (using the mean of three staff questionnaires) as the outcome measure and service setting (crisis house vs. ward) as the sole explanatory variable was then carried out as an initial test of the primary hypothesis (that therapeutic alliances are stronger in the crisis house), unadjusted for potential confounders. For this and all subsequent analyses, adjustment was made for clustering within the data, clusters being all the service users admitted to a particular crisis house or ward.
Following this initial unadjusted test, an adjusted estimate of the association between service setting and therapeutic relationships was obtained by carrying out a cluster-adjusted linear regression with mean therapeutic relationship as the outcome variable and age, ethnic group, sex, length of stay prior to the study interview, history of previous admission, whether or not detained under the Mental Health Act47 during this admission, and diagnosis, as further explanatory variables alongside service setting.
Subsequent main steps in the analysis involved exploring variables associated with service satisfaction. Initially, mean score for service satisfaction was compared between crisis houses and hospital in an unadjusted analysis. This was then adjusted for the service user-related demographic and service use variables listed above. We then used cluster-adjusted linear regression analysis to explore the associations between service user satisfaction and therapeutic relationship, relationship with peers, extent of recovery and negative events. Each of these was first individually entered into a linear regression with service user satisfaction as an outcome variable. Those variables that were significantly associated with service user satisfaction with at least a p = 0.10 level were then entered together into a linear regression for service user satisfaction with the above demographic and service use variables also included. This allowed us to investigate independent predictors of satisfaction and the relative influence of different variables and the amount of variance in service user satisfaction explained by the model.
Qualitative study component
Aim
The aim of the qualitative component of the study was to explore experiences of relationships between staff and service users in acute residential settings, focusing especially on the factors that impede and facilitate good staff–service user relationships. We also aimed to use the qualitative findings to help us understand the quantitative findings, and, in particular, to add depth, richness and enhanced understanding to the model of factors that influence service user satisfaction with residential services.
Sampling and inclusion criteria
We sought to interview four service users from each service (32 in total), with analysis proceeding alongside the latter stages of the interviewing to allow assessment of whether or not thematic saturation had been reached. The inclusion criteria for this component of the study were:
-
a good level of English
-
able to provide informed consent
-
resident in the crisis house for 1 week, or on the ward for 2 weeks to ensure experiences of establishing and maintaining relationships with staff and peers
-
fit with the purposive sampling framework.
Purposive sampling was used to identify people with a wide range of clinical, sociodemographic and service experience characteristics. To create the purposive sampling framework, demographic and service use data for all service users over the previous 12 months were requested from each ward and crisis house (although there was some variability in the time frames of the data eventually provided). Data were specifically requested on sex, age, ethnicity, broad diagnosis, and number of previous hospital and/or crisis house admissions.
For each service, the percentages of service users within each demographic and service use category were calculated. We were then able to take an average percentage for each demographic and service use category across crisis houses and across wards. This enabled us to generate broad purposive sampling targets. We also wanted to ensure diversity in our sample and so erred on the side of over-recruiting categories of people who are often under-represented in research and who may have particular experiences and insights regarding staff–service user relationships, such as black and minority ethnic service users.
We aimed to interview two staff members from each service (16 in total). Again, staff were sampled purposively to reflect a range of relevant perspectives including professional background and seniority. We specifically examined the following demographic and professional features and set targets based on ensuring a wide variety of experiences and characteristics: sex; age; ethnicity; level of seniority; clinical compared with non-clinical professional backgrounds; time spent working in that environment; and experience working in both psychiatric ward and crisis residential alternative settings.
Recruitment and consent
For service user interviews, the study researchers liaised with senior staff members on duty at each service to identify participants who met the inclusion criteria and the profile required for purposive sampling and who were able to participate in an interview. Service staff initially approached participants, as in the procedures described above for the quantitative part of the study. A researcher then met with participants who were interested in taking part and provided further information and opportunities to discuss the study. Written informed consent was sought, as described above. Participants were paid £15 on completing interviews as an acknowledgement of their time and contribution. Most service user interviews (93%) were conducted by service user researchers.
For staff interviews, the study researchers liaised with senior members of staff, typically the ward or crisis house manager, to obtain details of staff teams. The purposive sampling framework was used to identify suitable interviewees who were approached regarding their willingness to participate in the study. In addition, the research was discussed at some staff team meetings and in communal staff offices. Where staff members indicated an interest in participating in the study, their fit with the purposive sampling framework was assessed. Interested staff members were provided with a written information sheet, given the opportunity to discuss the study and asked to provide written consent.
Measures
The study researchers conducted semistructured interviews with service user participants. The interview guide was developed to address the main study questions; the questions were grounded in topics that were identified as important in determining people’s experiences of psychiatric hospitalisation in a user-led study,34 expert opinion in the study group and relevant literature. The interview guide was piloted with two service users and was finalised in collaboration with the service user researchers and the other members of the study research team. The service user topic guide included:
-
expectations and preferences regarding their relationships with staff
-
views on the barriers to and facilitators of positive relationships with staff
-
a number of factors that, based on TAS, we hypothesised might affect the nature and quality of service user and staff relationships, including the ward/crisis house atmosphere, the personal qualities of staff, the skills of staff, the professional background of staff, the numbers of staff available, how staff spend their time, how readily available staff are, choice of worker, the named worker system, team factors, the ward/crisis house routines, other people staying in the ward/crisis house and freedom
-
recommendations for improving staff–service user relationships
-
comparisons of staff–service user relationships in wards and crisis houses based on experiences, where participants had relevant experience allowing them to respond to this.
The study researchers also conducted semistructured interviews with staff. The topic guide was piloted with two members of staff and was finalised in collaboration with the service user researchers and the study research team. As in the service user interviews, the topic guide included:
-
views on the characteristics of positive working relationships
-
views on the barriers to and facilitators of positive relationships with service users
-
again, we asked staff about factors that, based on TAS, might affect the nature and quality of service user–staff relationships, including the atmosphere or culture, the management of the service, supervision and support, the organisation of the service, team factors, the named worker system, the numbers of staff available, the professional background of staff available, the amount of time spent with service users, the ways in which staff spend their time with service users, the personal qualities of staff, the skills of staff, the routines of the service, changes in people’s crises, service users’ freedom and other service users staying on the ward or in the crisis house
-
recommendations for improving staff–service user relationships
-
comparisons of staff–service user relationships in wards and crisis houses based on experiences, where relevant.
Data management and analysis
Service user and staff interviews were digitally recorded and transcribed using a professional transcription service before being uploaded to the qualitative software.
Qualitative data were analysed using thematic analysis. The analysis was approached from a contextualised perspective with the primary aim of understanding the factors that impede and facilitate staff–service user relationships in acute settings. NVivo 9 (QSR International, Warrington, UK) qualitative data analysis software was used to facilitate and ensure systematic analysis, storage and retrieval of data. A service user researcher (AS) generated the initial coding frame based on detailed reading and re-reading of a selection of transcripts. A number of interviews were then coded to test the feasibility and fit of the coding frame with further transcripts. AS then wrote a long memo containing early thoughts about the key higher-level themes arising from the data. This document was circulated to the study group. Concurrently, members of the study group analysed a selection of transcripts. All study group members analysed the same interview with a service user on an acute ward. Half the study group members then analysed an interview with a staff member from a crisis house while the other half analysed an interview with a staff member working on an acute ward. Finally, time permitting, study group members each analysed a different interview with service users resident in crisis houses. Analysts then brought their descriptions of and reflections on the data to a study group meeting. Discussions focussed on the patterns and connections between our analyses and also points of divergence and disagreement. The meeting was carefully recorded and the results used to revise the coding frame and raise the level of abstraction. Following study group discussion, two researchers (AS and EB) coded the transcripts based on the revised coding frame. However, this coding frame was not static but evolved as data coding progressed. The final coding frame addressed our initial research questions regarding the factors that influence therapeutic alliances, as well as exploring themes that emerged inductively from the data. Themes included factors that influence relationships; the impact of situations/relationships on staff and service users; experiences of positive or negative relational incidents; wanted and unwanted relationship attributes; staff attitude and approach; perceptions and descriptions of the staff role; issues of power and control; and survival strategies. The account in this report is based on an initial analysis of the data and focuses principally on the themes that are relevant to an understanding of therapeutic alliances in acute residential settings. We aim to present a more detailed account of the qualitative findings in further papers arising from the study.
Chapter 4 Results
We begin by describing the findings of the quantitative study, followed by an account of relevant themes arising from the qualitative part of the investigation.
Quantitative findings
Sample numbers and response rates
For the TAS-2 study, we interviewed 108 service users in crisis houses. The PET study recruited 247 participants in East London, and Camden and Islington (Norfolk participants in the PET study were excluded from the analyses reported here, as no Norfolk crisis house was included in the study). Following adjustment for clustering within wards, this fulfilled the study aim of yielding 90% power to detect a difference of half a standard deviation between crisis houses and standard acute wards in mean scores on our measure of therapeutic relationship, STAR-P. Table 1 describes the composition of the sample by service type and the numbers of refusals at each site.
| Quantitative study component: participant numbers per site with numbers of refusals | |||||
|---|---|---|---|---|---|
| Crisis house | Participants (n) | Refusalsa (n) | Acute ward | Participants (n) | Refusals(n) |
| Service 1 | 38 | 7 | Site 1 (wards implementing PET) | 61 | 26 |
| Service 2 | 27 | 8 | Site 1 (wards not implementing PET) | 61 | 30 |
| Service 3 | 27 | 2 | Site 2 (wards implementing PET) | 64 | 16 |
| Service 4 | 16 | 2 | Site 2 (wards not implementing PET) | 61 | 22 |
Thus a good response rate of 85% was obtained for the crisis house arm of the study. For the acute ward arm (from the PET study) a response rate of 72% was achieved.
Sample characteristics
Table 2 shows demographic and service use characteristics of the study sample.
| Characteristic | Overall (n = 355) | Crisis house (n = 108) | Acute ward (n = 247) |
|---|---|---|---|
| Sex, n (% male) | 179 (50.4) | 38 (35.2) | 141 (57.1) |
| Age, mean years (SD) | 40.2 (13.2) | 41.3 (13.0) | 39.7 (13.3) |
| Ethnicity, n (%) | |||
| White British | 138 (39.0) | 63 (58.9) | 75 (30.4) |
| White other | 41 (11.6) | 17 (15.9) | 24 (9.7) |
| Black Caribbean or African | 79 (22.3) | 9 (8.4) | 70 (28.3) |
| Asian | 38 (10.7) | 4 (3.7) | 34 (13.8) |
| Mixed | 24 (6.8) | 12 (11.2) | 12 (4.9) |
| Other | 34 (9.6) | 2 (1.9) | 32 (13.0) |
| Time in service centre prior to the interview, median weeks (IQR) | 3.3 (1.9–7.9) | 1.3 (1.0–1.9) | 5.6 (3.0–10.0) |
| Admitted to psychiatric hospital in the past, n yes (%) | 274 (77.8) | 81 (77.1) | 193 (78.1) |
| Previous admissions to psychiatric hospital, n (%) | |||
| 0 | 78 (22.2) | 24 (22.8) | 54 (21.9) |
| 1 | 55 (15.6) | 11 (10.5) | 44 (17.8) |
| 2–5 | 138 (39.2) | 33 (31.4) | 105 (42.5) |
| 6–10 | 54 (15.3) | 21 (20.0) | 33 (13.4) |
| > 10 | 27 (7.7) | 16 (15.2) | 11 (4.5) |
| Total admissions to crisis houses (including current), n (%) | |||
| 1 | – | 45 (42.5) | – |
| 2–4 | – | 37 (34.9) | – |
| 5–10 | – | 13 (12.3) | – |
| > 10 | – | 11 (10.4) | – |
| Mental Health Act status at admission, n detained at admission (%) | 166 (46.9) | 1 (0.9) | 165 (66.9) |
| Current/most recent clinical diagnosis, n (%) | |||
| Schizophrenia/schizoaffective disorder | 138 (43.7) | 18 (17.5) | 120 (56.3) |
| Bipolar affective disorder | 53 (16.8) | 17 (16.5) | 36 (16.9) |
| Other psychosis | 7 (2.2) | 3 (2.9) | 4 (1.9) |
| Depression | 42 (13.3) | 31 (30.1) | 11 (5.2) |
| Personality disorder | 41 (13.0) | 25 (24.3) | 16 (7.5) |
| Other | 35 (11.1) | 9 (8.7) | 26 (12.2) |
Several substantial differences emerge from these findings. The age profile was very similar between the crisis house and acute ward samples but with markedly higher proportions of women and of white British service users in the crisis houses. The sex disparity is likely to be more marked because one crisis house admitted only women. People from black Caribbean, black African and Asian backgrounds were under-represented in the crisis house group compared with the ward. The majority in each group had a previous history of at least one psychiatric hospital admission, with greater numbers of the crisis house participants having had a large number of admissions (at least five admissions for 35% of crisis house service users and 18% of ward service users). In keeping with service policies, two-thirds of hospital service users, but only one crisis house service user (transferred on leave from the ward) had been compulsorily detained in hospital at any stage in their current admission. Diagnostic mix was also markedly different, with just 56% of the ward sample, but only 17% of the crisis house sample, having a diagnosis of schizophrenia or schizoaffective disorder, and depression (24% in the crisis houses vs. 8% on the wards) and personality disorder (24% in crisis houses vs. 7% on wards) more highly represented in the crisis houses.
Ratings of therapeutic relationships, satisfaction and other aspects of inpatient experience
Table 3 shows scores on the main measures of service users’ views and experiences of admission for the crisis houses and the inpatient wards. We also present data separately for the PET and the non-PET wards: these show relatively similar scores for the two arms of the PET study, justifying our decision for TAS-2 to include both groups as a pooled hospital ward group. Service user satisfaction scores indicate that the ward group had a mean satisfaction score indicating satisfaction just more positive than the neutral point on the instrument (mean 21, where 20 indicates that overall the service user is neither satisfied nor dissatisfied). The crisis house group mean of 27.3 (maximum possible – 32) indicates that on average service users were between the satisfied and very satisfied points on the scale. Likewise, therapeutic relationship scores indicated a markedly positive view among crisis houses, and a more neutral one among ward service users. Ratings for the crisis house staff were similar to those for long-term care co-ordinators in community mental health teams in a recent London study. 52
| Characteristic | Overall (n = 355)a | Crisis house (n = 108) | Acute ward (n = 247) | PET ward (n = 125) | Non-PET (n = 122) |
|---|---|---|---|---|---|
| Client satisfaction: CSQ total,b mean (95% CI) | 23.0 (22.3 to 23.7) | 27.5 (26.6 to 28.3) | 21.0 (20.2 to 21.8) | 21.5 (20.4 to 22.7) | 20.5 (19.3 to 21.7) |
| Therapeutic relationship: average STAR-Pc (maximum three questionnaires),d mean (95% CI) | 30.9 (29.9 to 32.0) | 37.2 (35.5 to 38.8) | 28.3 (27.1 to 29.5) | 28.3 (26.6 to 30.0) | 28.3 (26.5 to 30.0) |
| Therapeutic relationship: general staff STAR-P,e mean (95% CI) | 28.8 (27.5 to 30.0) | 36.5 (34.7 to 38.2) | 25.6 (24.2 to 27.0) | 25.9 (24.0 to 27.8) | 25.2 (23.1 to 27.4) |
| Extent of recovery: RAS total,f mean (95% CI) | 115.3 (112.1 to 118.6) | 102.3 (97.2 to 107.4) | 120.9 (117.1 to 124.8) | 119.2 (113.6 to 124.8) | 122.9 (117.5 to 128.2) |
| Support from other service users: IPRI total,g mean (95% CI) | 60.3 (58.5 to 62.2) | 68.4 (65.5 to 71.3) | 57.1 (54.9 to 59.3) | 55.9 (52.8 to 58.9) | 58.4 (55.3 to 61.4) |
Table 4 shows the results of the primary analysis comparing therapeutic relationships (average STAR-P score between ward and crisis house). Both the unadjusted and adjusted tests show a highly significant difference in the strength of therapeutic relationship that favours the crisis house group. Wards have a mean −8.7 (95% confidence interval −12.3 to −5.2) lower average STAR-P score compared with crisis houses after adjustment for demographic and service use characteristics in the multivariable analysis. Age was the only other variable to show a statistically significant association with therapeutic relationship, with a 0.4 increase in STAR-P score for every 5 years increase in participant’s age. Ethnic group, whether or not compulsorily detained and clinical diagnosis were significantly associated with therapeutic relationship on initial univariable testing, but not significant in the final multivariable model.
| Characteristic | Unadjusted analysis | Multivariable analysis | ||
|---|---|---|---|---|
| Coefficient (95% CI) | p-value | Coefficient (95% CI) | p-value | |
| Service type, ward vs. crisis house | –8.86 (–11.53 to –6.19) | < 0.0001 | –8.74 (–12.30 to –5.19) | < 0.0001 |
| Sex, female vs. male | 0.42 (–3.00 to 3.85) | 0.80 | –1.73 (–3.74 to 0.29) | 0.09 |
| Age, per 5 years older | 0.37 (–0.00 to 0.74) | 0.05 | 0.40 (0.05 to 0.15) | 0.03 |
| Ethnicity | 0.01 | 0.23 | ||
| White British | Ref | Ref | ||
| White other | –1.44 (–5.33 to 2.45) | –1.91 (–5.54 to 1.72) | ||
| Black | –2.97 (–6.56 to 0.62) | 0.73 (–1.90 to 3.36) | ||
| Asian | –5.34 (–8.51 to –2.17) | –1.98 (–4.91 to 0.94) | ||
| Mixed | 0.65 (–4.03 to 5.34) | 0.54 (–3.65 to 4.74) | ||
| Other | –5.31 (–9.55 to –1.08) | –1.70 (–4.68 to 1.28) | ||
| Time in service centre prior to the interview, per week longer | –0.15 (–0.33 to 0.03) | 0.11 | 0.03 (–0.09 to 0.15) | 0.63 |
| Admitted to psychiatric hospital in the past, yes vs. no | –2.07 (–4.71 to 0.56) | 0.12 | –1.26 (–3.70 to 1.17) | 0.29 |
| Mental Health Act status at admission, detained vs. not detained | –3.05 (–5.68 to –0.42) | 0.03 | 0.24 (–2.30 to 2.79) | 0.84 |
| Current/most recent clinical diagnosis | < 0.0001 | 0.73 | ||
| Schizophrenia/schizoaffective disorder | Ref | Ref | ||
| Bipolar affective disorder | 0.57 (–3.13 to 4.27) | –0.49 (–3.80 to 2.81) | ||
| Other psychosis | 4.03 (–6.39 to 14.44) | 2.63 (–5.72 to 10.98) | ||
| Depression | 7.26 (4.05 to 10.48) | 1.45 (–1.19 to 4.10) | ||
| Personality disorder | 2.80 (–0.25 to 5.85) | 0.02 (–3.09 to 3.12) | ||
| Other | 1.03 (–4.08 to 6.15) | 0.54 (–4.43 to 5.50) | ||
Service satisfaction: comparison between crisis house and inpatient ward
The overall aim of the next step in the analysis was to examine the relationship between setting and satisfaction, and identify variables influencing this relationship. We began by examining the unadjusted relationship between satisfaction and setting, and then the relationship adjusted for individual demographic and service use variables. The results are shown in Table 5 . A highly significant relationship was found between satisfaction and setting in both adjusted and unadjusted analyses, with only a small reduction in mean difference between hospital and crisis house from 6.5 points on the CSQ to 5.3 points when adjustment was made for demographic, diagnostic and service use variables (95% confidence interval –7.6 to –2.9). None of the demographic and service use variables remained significant in this multiply adjusted regression model.
| Characteristic | Unadjusted analysis | Multivariable analysis | ||
|---|---|---|---|---|
| Coefficient (95% CI) | p-value | Coefficient (95% CI) | p-value | |
| Service type, ward vs. crisis house (reference) | –6.46 (–8.05 to –4.86) | < 0.0001 | –5.26 (–7.59 to –2.94) | < 0.0001 |
| Sex, female vs. male (reference) | 0.44 (–1.70 to 2.58) | 0.67 | –0.81 (–2.12 to 0.51) | 0.22 |
| Age, per 5 years older | 0.25 (0.00 to 0.49) | 0.05 | 0.13 (–0.10 to 0.35) | 0.26 |
| Ethnicity | 0.002 | 0.50 | ||
| White British | Ref | Ref | ||
| White other | –0.80 (–2.69 to 1.09) | –0.91 (–2.85 to 1.03) | ||
| Black | –2.72 (–5.02 to –0.42) | –0.82 (–2.85 to 1.21) | ||
| Asian | –3.23 (–5.28 to –1.19) | –1.41 (–3.55 to 0.73) | ||
| Mixed | 0.57 (–1.73 to 2.87) | –0.39 (–1.93 to 1.16) | ||
| Other | –4.45 (–7.87 to –1.04) | –3.02 (–6.28 to 0.24 | ||
| Time in service centre prior to the interview, per week longer | –0.14 (–0.28 to –0.01) | 0.04 | 0.00 (–0.08 to 0.08) | 0.99 |
| Admitted to psychiatric hospital in the past, yes vs. no | –1.34 (–2.89 to 0.21) | 0.09 | –0.67 (–2.26 to 0.93) | 0.39 |
| Mental Health Act status at admission, detained vs. not detained | –2.85 (–5.28 to -0.42) | 0.02 | –1.45 (–3.25 to 0.35) | 0.11 |
| Current/most recent clinical diagnosis | < 0.0001 | 0.06 | ||
| Schizophrenia/schizoaffective disorder | Ref | Ref | ||
| Bipolar affective disorder | –0.52 (–2.72 to 1.68) | –1.91 (–3.63 to –0.19) | ||
| Other psychosis | 5.66 (2.41 to 8.91) | 3.65 (–1.02 to 8.32) | ||
| Depression | 4.16 (2.11 to 6.21) | –0.66 (–2.68 to 1.36) | ||
| Personality disorder | 1.57 (–1.10 to 4.23) | –1.75 (–4.00 to 0.49) | ||
| Other | 0.05 (–1.91 to 2.01) | –1.06 (–3.15 to 1.03) | ||
Relationship between service setting and service user ratings of recovery and peer support
A secondary aim of the analysis was to test whether or not self-rated recovery and peer support differed between settings, and, if so, whether or not they might contribute to explaining differences in satisfaction between crisis houses and hospital wards. Table 6 shows unadjusted and adjusted comparisons between hospital and crisis houses in RAS score. Table 7 shows these data for the IPRI items from the study.
| Characteristic | Unadjusted analysis | Multivariable analysis | ||
|---|---|---|---|---|
| Coefficient (95% CI) | p-value | Coefficient (95% CI) | p-value | |
| Service type, ward vs. crisis house | 18.59 (11.56 to 25.63) | < 0.0001 | 2.36 (–4.90 to 9.62) | 0.51 |
| Sex, female vs. male | –7.79 (–15.66 to 0.08) | 0.05 | 0.39 (–8.11 to 8.89) | 0.93 |
| Age, per 5 years older | 0.34 (–0.87 to 1.54) | 0.57 | 0.78 (–0.26 to 1.81) | 0.13 |
| Ethnicity | 0.002 | 0.09 | ||
| White British | Ref | Ref | ||
| White other | 8.61 (–3.43 to 20.64) | 6.28 (–5.69 to 18.25) | ||
| Black | 19.60 (10.90 to 28.29) | 11.21 (2.66 to 19.76) | ||
| Asian | 9.28 (–0.80 to 19.37) | 1.28 (–9.35 to 11.91) | ||
| Mixed | 1.45 (–14.36 to 17.25) | –0.49 (–11.69 to 10.72) | ||
| Other | 9.52 (–3.18 to 22.21) | –0.66 (–9.55 to 8.24) | ||
| Time in service centre prior to the interview, per week longer | 0.47 (0.03 to 0.92) | 0.04 | –0.07 (–0.44 to 0.29) | 0.69 |
| Admitted to psychiatric hospital in the past, yes vs. no | –0.85 (–9.97 to 8.28) | 0.85 | –2.02 (–9.99 to 5.96) | 0.60 |
| Mental Health Act status at admission, detained vs. not detained | 9.11 (–1.96 to 20.18) | 0.10 | 7.46 (0.71 to 14.21) | 0.03 |
| Current/most recent clinical diagnosis | < 0.0001 | < 0.0001 | ||
| Schizophrenia/schizoaffective disorder | Ref | Ref | ||
| Bipolar affective disorder | –5.86 (–14.98 to 3.26) | –1.29 (–11.54 to 8.95) | ||
| Other psychosis | –1.03 (–22.33 to 20.28) | 3.11 (–18.10 to 24.33) | ||
| Depression | –27.74 (–38.65 to –16.84) | –20.72 (–31.91 to –9.53) | ||
| Personality disorder | –38.46 (–47.19 to –29.74) | –31.41 (–41.36 to –21.47) | ||
| Other | –16.08 (–27.39 to –4.77) | –12.68 (–23.04 to –2.31) | ||
| Characteristic | Unadjusted analysis | Multivariable analysis | ||
|---|---|---|---|---|
| Coefficient (95% CI) | p-value | Coefficient (95% CI) | p-value | |
| Service type, ward vs. crisis house | –11.31 (–15.27 to –7.36) | < 0.0001 | –12.08 (–18.53 to –5.63) | 0.001 |
| Sex, female vs. male | 0.12 (–4.77 to 5.01) | 0.96 | –2.07 (–6.39 to 2.24) | 0.33 |
| Age, per 5 years older | 0.19 (–0.60 to 0.99) | 0.62 | –0.15 (–1.17 to 0.87) | 0.76 |
| Ethnicity | 0.13 | 0.54 | ||
| White British | Ref | Ref | ||
| White other | –3.55 (–8.75 to 1.66) | –0.89 (–7.55 to 5.77) | ||
| Black | –6.19 (–13.01 to 0.64) | –0.62 (–7.60 to 6.37) | ||
| Asian | –6.31 (–13.71 to 1.09) | –4.49 (–12.50 to 3.52) | ||
| Mixed | –7.20 (–16.95 to 2.55) | –4.31 (–12.13 to 3.51) | ||
| Other | –0.82 (–6.29 to 4.64) | –4.83 (–14.11 to 4.44) | ||
| Time in service centre prior to the interview, per week longer | –0.30 (–0.54 to –0.06) | 0.02 | –0.07 (–0.29 to 0.16) | 0.53 |
| Admitted to psychiatric hospital in the past, yes vs. no | –5.65 (–9.39 to –1.91) | 0.005 | –4.23 (–7.79 to –0.68) | 0.02 |
| Mental Health Act status at admission, detained vs. not detained | –2.61 (–7.77 to 2.55) | 0.30 | –0.55 (–6.71 to 5.62) | 0.86 |
| Current/most recent clinical diagnosis | 0.0005 | 0.15 | ||
| Schizophrenia/schizoaffective disorder | Ref | Ref | ||
| Bipolar affective disorder | –1.24 (–9.58 to 7.10) | –2.30 (–11.10 to 6.51) | ||
| Other psychosis | 8.72 (0.95 to 16.49) | 5.61 (–3.36 to 14.58) | ||
| Depression | 6.45 (1.41 to 11.49) | –2.63 (–8.47 to 3.20) | ||
| Personality disorder | –0.45 (–6.08 to 5.19) | –6.46 (–12.76 to –0.17) | ||
| Other | 0.91 (–5.30 to 7.11) | –1.50 (–8.10 to 5.10) | ||
In the unadjusted analysis comparing crisis house and hospital wards in Table 6 , a highly significant difference in self-rated recovery emerges between crisis house and ward, with much better ratings of recovery in the ward. However, with adjustment for demographic, diagnostic and service use variables, this difference was no longer statistically significant, diminishing from a mean difference of 18.6 to 2.4 favouring the ward. In the multivariable analysis, the main variable significantly associated with self-rated recovery was diagnosis, with people diagnosed with personality disorder and depression having adjusted mean scores 31.4 (95% confidence interval −41.4 to −21.5) and 20.7 (95% confidence interval −31.9 to −9.5) below the reference group who were people diagnosed with schizophrenia/schizoaffective disorder respectively. Mental Health Act status also remained significant, with people detained under the Mental Health Act rating their recovery as 7.5 points better than those not detained.
With regard to peer support (see Table 7 ), both adjusted and unadjusted analyses indicated significantly better peer support in the crisis houses than the hospital wards, a mean difference of 11.3 points (95% confidence interval −15.3 to −7.4) on the 104-point scale, rising to 12.1 (95% confidence interval −18.5 to −5.6) in the adjusted analysis. The only other variable reaching significance in the multivariable analysis was hospital admission, with people with a previous history of hospital admission rating peer support less positively than those without such a history.
Satisfaction’s relationship with therapeutic relationship, peer support and recovery ratings
In order to fulfil the study aim of examining whether or not better therapeutic relationships potentially account for differences in satisfaction between crisis house and hospital, therapeutic relationship ratings were added to the model for service user satisfaction shown in Table 5 . Secondary study aims were to examine how far peer support and extent of self-rated recovery may contribute to satisfaction: we therefore also investigated these relationships at this step in the analysis ( Table 8 ).
| Characteristic | Unadjusted analysis | Multivariable analysis | ||
|---|---|---|---|---|
| Coefficient (95% CI) | p-value | Coefficient (95% CI) | p-value | |
| Service type, ward vs. crisis house | –6.46 (–8.05 to –4.86) | < 0.0001 | –2.44 (–4.76 to –0.13) | 0.04 |
| Sex, female vs. male | 0.44 (–1.70 to 2.58) | 0.67 | –0.22 (–1.37 to 0.93) | 0.70 |
| Age, per 5 years older | 0.25 (0.00 to 0.49) | 0.05 | 0.03 (–0.17 to 0.23) | 0.76 |
| Ethnicity | 0.002 | 0.59 | ||
| White British | Ref | Ref | ||
| White other | –0.80 (–2.69 to 1.09) | –0.29 (–1.93 to 1.36) | ||
| Black | –2.72 (–5.02 to –0.42) | –0.91 (–2.96 to 1.14) | ||
| Asian | –3.23 (–5.28 to –1.19) | –0.39 (–2.10 to 1.33) | ||
| Mixed | 0.57 (–1.73 to 2.87) | –0.35 (–2.06 to 1.37) | ||
| Other | –4.45 (–7.87 to –1.04) | –2.21 (–5.30 to 0.87) | ||
| Time in service centre prior to the interview, per week longer | –0.14 (–0.28 to –0.01) | 0.04 | –0.00 (–0.08 to 0.08) | 0.94 |
| Average STAR-P score, per 10 units higher | 3.79 (3.10 to 4.48) | < 0.0001 | 2.49 (1.51 to 3.48) | < 0.0001 |
| RAS total, per 10 units higher | –0.07 (–0.36 to 0.22) | 0.63 | – | – |
| IPRI total, per 10 units higher | 1.68 (1.27 to 2.09) | < 0.0001 | 0.74 (0.16 to 1.31) | 0.01 |
| Admitted to psychiatric hospital in the past, yes vs. no | –1.34 (–2.89 to 0.21) | 0.09 | –0.02 (–1.28 to 1.24) | 0.97 |
| Mental Health Act status at admission, detained vs. not detained | –2.85 (–5.28 to -0.42) | 0.02 | –1.32 (–2.81 to 0.17) | 0.08 |
| Current/most recent clinical diagnosis | < 0.0001 | 0.05 | ||
| Schizophrenia/schizoaffective disorder | Ref | Ref | ||
| Bipolar affective disorder | –0.52 (–2.72 to 1.68) | –1.54 (–2.95 to –0.13) | ||
| Other psychosis | 5.66 (2.41 to 8.91) | 2.75 (–2.42 to 7.91) | ||
| Depression | 4.16 (2.11 to 6.21) | –0.91 (–2.70 to 0.89) | ||
| Personality disorder | 1.57 (–1.10 to 4.23) | –1.29 (–3.70 to 1.12) | ||
| Other | 0.05 (–1.91 to 2.01) | –0.87 (–2.83 to 1.10) | ||
Again, the central column (see Table 8 ) shows unadjusted relationships for each variable with satisfaction. Both therapeutic relationship and peer support emerge in this analysis as having highly significant relationships with satisfaction, with a 10 point higher average STAR-P score associated with a CSQ score 3.8 points higher, and a 10 point difference in peer support associated with a 1.7 point higher CSQ score. Self-ratings of recovery, however, showed no association with service user satisfaction: recovery was not therefore included in the multivariable model. In the multivariable model, the association between therapeutic relationships and satisfaction remained significant at the p < 0.0001 level, with the mean difference per 10 points on the scale now 2.5. For peer support, mean difference per 10 points on the scale fell from 1.7 to 0.7, but the association remained significant at the p = 0.01 level. With these variables in the model, a marked reduction in the adjusted association between service setting and satisfaction was observed, with the mean difference between crisis houses and wards falling from 5.3 in the analyses adjusted only for demographic and clinical variables (see Table 5 ), to 2.4 in the new model (now just statistically significant at p = 0.04). Thus these findings are compatible with the idea that better therapeutic relationships and peer support may be important factors in accounting for greater service user satisfaction with crisis house care.
Exploratory analyses regarding the influence of negative events
The final step in our analysis involved exploring the frequency and possible impact on service user satisfaction of negative events experienced in residential care. This analysis is very much exploratory in nature given that the checklist of negative events used was generated specifically for the study and has not been subject to any form of psychometric testing.
We obtained reports of two types of negative event, those related to the behaviour of other service users and those related to staff. Table 9 shows the frequency of the service user-related negative events in the crisis houses and wards, along with the results of asking participants to rate their impact on them. Most of the negative events were more frequently reported in the hospital. The most frequent adverse events related to other service users in hospital were witnessing disturbed behaviour (74%), theft of personal belongings (32%) and verbal abuse (28%). In the crisis house, the most frequently reported events were witnessing disturbed behaviour (35%), being dismissed or ignored (23%) and being verbally abused (10%). Thirty hospital ward service users (12%) reported having been physically assaulted by another service user during their current stay. Twenty-one ward service users (9%) had been sexually harassed and five (2%) had been sexually assaulted by another service user during their stay. In the crisis houses, one person (1%) reported physical assault and one person (1%) reported sexual harassment by another service user; none reported sexual assault.
| Characteristic | Overall (n = 355) | Crisis house (n = 108) | Acute ward (n = 247) | p-value |
|---|---|---|---|---|
| Theft of personal belongings, n yes (%) | 83 (23.5) | 4 (3.8) | 79 (32.0) | < 0.0001 |
| Offered illicit substances or alcohol, n yes (%) | 37 (10.5) | 7 (6.6) | 30 (12.2) | 0.12 |
| Verbal threats, n yes (%) | 71 (20.1) | 6 (5.7) | 65 (26.3) | < 0.0001 |
| Verbally abused, n yes (%) | 81 (23.0) | 11 (10.4) | 70 (28.3) | < 0.0001 |
| Physically assaulted, n yes (%) | 31 (8.8) | 1 (0.9) | 30 (12.2) | < 0.0001 |
| Sexually harassed, n yes (%) | 22 (6.2) | 1 (0.9) | 21 (8.5) | 0.007 |
| Sexual assaulted, n yes (%) | 5 (1.4) | 0 (0.0) | 5 (2.0) | 0.33 |
| Victim of religious, racial or homophobic discrimination, n yes (%) | 39 (11.1) | 6 (5.7) | 33 (13.4) | 0.03 |
| Forced to do something, n yes (%) | 25 (7.1) | 4 (3.8) | 21 (8.5) | 0.11 |
| Dismissed or ignored, n yes (%) | 76 (21.6) | 24 (22.9) | 52 (21.1) | 0.71 |
| Witnessed disturbed behaviour, n yes (%) | 219 (62.0) | 37 (34.9) | 182 (73.7) | < 0.0001 |
| Other, n yes (%) | 11 (3.1) | 3 (2.8) | 8 (3.3) | 0.84 |
Table 10 shows reports of staff-related negative events. The most frequent in hospital were being dismissed or ignored (41%), being given medication against one’s will (38%) and being physically restrained (22%). In the crisis house, 17% reported that they had been dismissed or ignored, 5% that they had been forced to do something, and 4% that they had spent time in the quiet room. Thirteen (5%) ward service users reported that they had been physically assaulted by a staff member, six (2%) reported sexual harassment and two (1%) reported sexual assault by a staff member; none of the crisis house residents reported this.
| Characteristic | Overall (n = 355) | Crisis house (n = 108) | Acute ward (n = 247) | p-value |
|---|---|---|---|---|
| Theft of personal belongings, n yes (%) | 25 (7.1) | 0 (0.0) | 25 (10.1) | < 0.0001 |
| Offered illicit substances or alcohol, n yes (%) | 0 (0.0) | 0 (0.0) | 0 (0.0) | |
| Verbal threats, n yes (%) | 25 (7.1) | 2 (1.9) | 23 (9.3) | 0.01 |
| Verbally abused, n yes (%) | 26 (7.4) | 2 (1.9) | 24 (9.7) | 0.01 |
| Physically assaulted, n yes (%) | 13 (3.7) | 0 (0.0) | 13 (5.3) | 0.01 |
| Sexually harassed, n yes (%) | 6 (1.7) | 0 (0.0) | 6 (2.3) | 0.18 |
| Sexual assaulted, n yes (%) | 2 (0.6) | 0 (0.0) | 2 (0.8) | 0.99 |
| Victim of religious, racial or homophobic discrimination, n yes (%) | 45 (12.8) | 3 (2.8) | 42 (17.0) | < 0.0001 |
| Forced to do something, n yes (%) | 51 (14.5) | 5 (4.7) | 46 (18.6) | 0.001 |
| Dismissed or ignored, n yes (%) | 119 (33.7) | 18 (17.0) | 101 (40.9) | < 0.0001 |
| Physically restrained, n yes (%) | 59 (16.8) | 2 (1.9) | 57 (23.2) | < 0.0001 |
| Spent time in quiet room, n yes (%) | 58 (16.5) | 4 (3.8) | 54 (22.0) | < 0.0001 |
| Given medication against will, n yes (%) | 96 (27.4) | 2 (1.9) | 94 (38.2) | < 0.0001 |
| Other, n yes (%) | 29 (8.3) | 18 (17.0) | 11 (4.5) | < 0.0001 |
Table 11 summarises the data on numbers of negative events experienced per service user in ward and crisis house, and also shows the results of weighting the impact of these events on the basis of service user’s ratings, with events described as having no impact rated as 0, and those with a great deal of impact rated 4.
| Characteristic | Overall (n = 355) | Crisis house (n = 108) | Acute ward (n = 247) |
|---|---|---|---|
| Service user-related negative events,a median (IQR) | 1 (0–3) | 0 (0–1) | 2 (1–4) |
| Staff-related negative events, median (IQR) | 1 (0–2) | 0 (0–1) | 1 (0–3) |
| Service user-related negative events (weighted by their impact),b median (IQR) | 2 (0–6) | 0 (0–3) | 3 (1–8) |
| Staff-related negative events (weighted by their impact),c median (IQR) | 2 (0–7) | 0 (0–2) | 3 (0–9) |
The final step in our analysis was to explore the impact of including negative events in the model for service user satisfaction. Table 12 shows the results, again, with unadjusted relationships in the central column and adjusted relationships in the right-hand column. Regarding staff-related negative events, unadjusted analyses both using number of negative events experienced and weighted number of negative events indicate a strong relationship with service user satisfaction. Such a relationship is also found for service user-related negative events. The weighted negative events scores were also added to the model for service user satisfaction previously obtained in Table 8 .
| Characteristic | Unadjusted analysis | Multivariable analysis | ||
|---|---|---|---|---|
| Coefficient (95% CI) | p-value | Coefficient (95% CI) | p-value | |
| Service type, ward vs. crisis house | –6.46 (–8.05 to –4.86) | < 0.0001 | –1.38 (–3.16 to 0.40) | 0.12 |
| Sex, female vs. male | 0.44 (–1.70 to 2.58) | 0.67 | 0.54 (–0.51 to 1.60) | 0.30 |
| Age, per 5 years older | 0.25 (0.00 to 0.49) | 0.05 | –0.02 (–0.24 to 0.21) | 0.87 |
| Ethnicity | 0.002 | 0.56 | ||
| White British | Ref | Ref | ||
| White other | –0.80 (–2.69 to 1.09) | –0.14 (–1.68 to 1.40) | ||
| Black | –2.72 (–5.02 to –0.42) | –1.19 (–3.03 to 0.65) | ||
| Asian | –3.23 (–5.28 to –1.19) | –0.60 (–2.45 to 1.25) | ||
| Mixed | 0.57 (–1.73 to 2.87) | –0.15 (–2.15 to 1.84) | ||
| Other | –4.45 (–7.87 to –1.04) | –1.01 (–3.90 to 1.88) | ||
| Time in service centre prior to the interview, per week longer | –0.14 (–0.28 to –0.01) | 0.04 | –0.00 (–0.07 to 0.06) | 0.97 |
| Average STAR-P score, per 10 units higher | 3.79 (3.10 to 4.48) | < 0.0001 | 2.19 (1.39 to 2.98) | < 0.0001 |
| Service user-related negative events, per additional negative event | –1.27 (–1.77 to –0.77) | < 0.0001 | – | – |
| Staff-related negative events, per additional negative event | –1.85 (–2.20 to –1.51) | < 0.0001 | – | – |
| Service user-related negative events (weighted), per additional weighted event | –0.33 (–0.52 to –0.15) | 0.001 | 0.05 (–0.13 to 0.23) | 0.56 |
| Staff-related negative events (weighted), per additional weighted event | –0.51 (–0.63 to –0.40) | < 0.0001 | –0.35 (–0.52 to –0.18) | < 0.0001 |
| IPRI total, per 10 units higher | 1.68 (1.27 to 2.09) | < 0.0001 | 0.50 (0.08 to 0.92) | 0.02 |
| Admitted to psychiatric hospital in the past, yes vs. no | –1.34 (–2.89 to 0.21) | 0.09 | 0.28 (–0.84 to 1.40) | 0.61 |
| Mental Health Act status at admission, detained vs. not detained | –2.85 (–5.28 to –0.42) | 0.02 | –1.26 (–2.72 to 0.19) | 0.09 |
| Current/most recent clinical diagnosis | < 0.0001 | 0.15 | ||
| Schizophrenia/schizoaffective disorder | Ref | Ref | ||
| Bipolar affective disorder | –0.52 (–2.72 to 1.68) | –1.12 (–2.73 to 0.50) | ||
| Other psychosis | 5.66 (2.41 to 8.91) | 2.23 (–2.14 to 6.61) | ||
| Depression | 4.16 (2.11 to 6.21) | –1.16 (–2.93 to 0.62) | ||
| Personality disorder | 1.57 (–1.10 to 4.23) | –1.44 (–3.77 to 0.88) | ||
| Other | 0.05 (–1.91 to 2.01) | –1.30 (–2.97 to 0.37) | ||
In this adjusted analysis, shown in Table 12 , the relationship between staff-related negative events, but not service user-related events, remained highly significant, with a fall in CSQ score of 0.4 per additional weighted event. In the resulting model, the difference between ward and crisis houses in satisfaction has fallen to 1.4 and is no longer statistically significant. This is compatible with the idea that therapeutic relationships, peer support and negative experiences related to staff may be important factors in accounting for the greater satisfaction found in crisis houses than in wards.
Qualitative findings
In this section we present findings from a thematic analysis of the qualitative data collected, aimed at identifying main themes relevant to understanding the nature of therapeutic alliances in crisis house and in hospital settings, the key influences on them, and the major ways in which therapeutic alliances differ between these settings. It is important to note that this is an initial analysis: our aim has been to present an overview of themes of particular relevance to our research questions and to interpretation of the quantitative findings. A more fine-grained analysis of some key themes arising from the data set will be presented in some further papers from the study. Below we focus principally on service user perspectives, with service user researchers having led on the identification of major themes from interviews with these participants, as described in the methods section. In the latter part of this section we will present some major themes from the staff interviews with the aim of understanding the dynamics and processes involved in staff–service user relationships from both perspectives.
Total numbers recruited fell slightly short of original targets: we had aimed for 32 service users and interviewed 29, whereas with staff we had aimed for 16 but interviewed 13. Difficulties arising from one crisis house closing and our need to find a substitute for it, and from staff absence, contributed to a slightly slower recruitment rate; however, analysis had been proceeding alongside the final stages of data collection and on reviewing this our view was that on the important main themes, thematic saturation appeared to have been reached and the main aims of the purposive sampling frame seemed to have been achieved. Our assessment was thus that this slightly reduced sample size was acceptable.
Sample for service user interviews in crisis houses and wards
Twenty-nine service users were recruited, 14 from crisis houses and 15 from acute wards. For service users from acute wards there was a near even split between men and women (seven women, eight men). The majority of participants were aged between 25 and 55 years, with two younger than this and one older adult. Seven participants identified themselves as being from a white British or white other background, three as black African or black Caribbean, two as Asian, two as being from mixed heritage backgrounds and one as ‘other’. There was a spread of previous experiences of inpatient wards: four people were experiencing their first admission, whereas six people had experienced more than six admissions. Four people had previously been admitted to crisis houses, one of these on numerous occasions.
Table 13 shows the characteristics of the service users recruited from inpatient wards, comparing these with the purposive sampling framework that guided participant recruitment, based on recent routine data regarding the characteristics of service users. Table 13 indicates that the broad characteristics aimed for were, for the most part, achieved. The service use data for wards showed a roughly even split between male and female service users (47% female and 53% male), and this was borne out in our sample, with 7 out of 15 interviewees (47%) being female and 8 out of 15 (53%) being male. The majority of previous service users (73%) were aged between 25 and 54 years, as were 12 out of 15 (80%) of our sample. For diversity we were able to include two people aged < 25 years and one aged > 55 years. Forty per cent of previous service users in wards were white British, as were six (40%) of our interviewees. A significant minority (23%) were of black Caribbean or black African origin, as were three (20%) of our interviewees. Seventeen per cent of previous service users were of Asian origin, reflecting the large Asian population in East London. We were able to mirror this in our sample with two Asian interviewees (13%). We also included two people of mixed heritage and one person of other white heritage. Regarding prior hospital admissions, 26% of service users had no previous admissions, 22% had one previous admission, 33% had two to five previous admissions and 19% had six or more previous admissions. In our sample four people (27%) had no previous admissions, one (7%) had one previous admission, five (33%) had two to five previous admissions and five (33%) had six or more admissions. Of these last five, three had also had previous crisis house admissions and so were well placed to compare the two. Seventy five per cent of ward service users had a diagnosis of a psychotic disorder, 25% had one of a non-psychotic disorder. In our sample 12 people (80%) had a diagnosis of a psychotic disorder and three people (20%) had one of a non-psychotic disorder.
| Characteristic | Purposive sampling targets (n = 16) | Actual figure (n = 15) |
|---|---|---|
| Sex | Female = 8 | 7 |
| Male = 8 | 8 | |
| Age (years) | Majority = 25–55 | 12 |
| < 25 = 1 | 2 | |
| > 55 = 2–3 | 1 | |
| Ethnicity | White British = 6–7 | 6 |
| White other = 2 | 1 | |
| Black Caribbean or black African = 3–4 | 3 | |
| Asian = 2–3 | 2 | |
| Mixed heritage = 1 | 2 | |
| Other/do not know/not stated = no target | 1 | |
| Number of previous hospital admissions | No previous admissions = 4 | 4 |
| One = 3 | 1 | |
| Two to five = 5 | 5 | |
| Six or more = 3 | 5a | |
| Diagnosis | Psychosis = 12 | 12 |
| Non-psychosis = 4 | 3 |
Table 14 presents the same data for the crisis house group, again indicating that the sample achieved was in most respects close to the intended purposive sample in its composition. Data regarding clinical populations in crisis houses showed that 63% of crisis house service users were female, possibly a reflection of the fact that one service admits only women. Our crisis house qualitative sample broadly reflected this with eight women (57%) and six men (43%). As with the wards, the majority of service users (75%) were between the ages of 25 and 54 years. Seventy-one per cent of our sample (10 people) were within this age range. With regards to ethnicity, 55% of crisis house service users were white British according to routine data. Nine out of 14 (64%) of our sample were white British. We were also able to include one person of other white background, three people of black Caribbean or black African background and one person of Asian background. With regard to previous hospital admission, the routine data on which we based our sample indicated that no service user without a prior hospital admission had gone to a crisis house during the period for which we obtained data, whereas 42% had one previous hospital admission, 53% had two to five admissions and 5% had six or more admissions. In our sample five people (36%) had no previous admissions, two people (14%) had one admission, four people (29%) had two to five admissions and three people (21%) had six or more admissions. Where diagnosis was concerned, routine data indicated a lower proportion (54%) with a psychotic diagnosis than in hospital. Fifty-four per cent of service users had a diagnosis of psychosis, 46% one of non-psychotic illness. In our sample five people (35%) had a diagnosis of psychosis; thus, people diagnosed with psychosis were slightly under-represented compared with our original planned sample.
| Characteristic | Purposive sampling targets (n = 16) | Actual figures (n = 14) |
|---|---|---|
| Sex | Female = 10 | 8 |
| Male = 6 | 6 | |
| Age (years) | Majority = 25–55 | 10 |
| < 25 = 1 or 2 | 2 | |
| > 55 = 2 or 3 | 2 | |
| Ethnicity | White British = 9 | 9 |
| White other = 2 | 1 | |
| Black Caribbean or black African = 2 | 3 | |
| Asian = 1 | 1 | |
| Mixed heritage = 1 | 0 | |
| Other/do not know/not stated = 1 | 0 | |
| Number of previous hospital admissions | No previous admissions = data suggests 0, but decided to include some people for diversity | 5 |
| One = 7 | 2 | |
| Two to five = 8 | 4 | |
| Six or more = 1 | 3 | |
| Diagnosis | Psychosis = 8 or 9 | 5 |
| Non-psychosis = 7 or 8 | 7 | |
| Missing | 1 not asked; 1 not known |
Staff sample
Thirteen members of staff were recruited to the study, seven from the hospital wards and six from the crisis houses. For the staff sample, we did not fix defined numbers to be interviewed from each category; instead, we aimed for a range of seniority, including at least one service manager from each setting, and considered whether or not staff were clinically qualified, their length of experience, their age and their ethnic group. Table 15 describes the sample recruited in crisis houses and in wards, indicating that a broad range of characteristics was indeed achieved.
| Characteristic | Ward staff (n = 7) | Crisis house staff (n = 6) |
|---|---|---|
| Role | Ward manager = 1 | Crisis house manager = 1 |
| Clinical practice lead = 1 | Senior project worker = 2 | |
| Ward nurse = 3 | Project worker = 3 | |
| Health-care assistant = 1 | ||
| Social therapist = 1 | ||
| Professional background | Mental health professional qualifications = 6/7 | Qualified counselor = 1 |
| Nurse = 1 | ||
| Without clinical qualifications = 3 | ||
| Missing = 1 | ||
| Number of years experience working in current type of service | < 1 = 0 | < 1 = 2 |
| 1–5 = 3 | 1–5 = 3 | |
| > 5 = 4 | > 5 = 1 | |
| Experience working in other setting | Had worked in crisis house = 2/7 | Had worked on inpatient ward = 3/6 |
| Age (years) | 25–34 = 3 | 25–34 = 1 |
| 35–44 = 2 | 35–44 = 2 | |
| 45–54 = 1 | 45–54 = 2 | |
| 55–64 = 1 | 55–64 = 1 | |
| Sex | Male = 3 | Male = 1 |
| Female = 4 | Female = 5 | |
| Ethnicity | Black African or black Caribbean = 5 | White British = 3 |
| Asian = 1 | White other = 2 | |
| Mixed heritage = 1 | Black African or black Caribbean = 1 |
Results of thematic analysis: service user interviews
Appendix 4 shows the topic guide used to interview service users. Although the focus of questions throughout was on relationships with staff, a wide range of material relevant to experiences of using the service emerged. In the themes reported below, we have focused principally on themes that allow interpretation of the quantitative findings by allowing us to understand what aspects of relationships in crisis houses may be preferred by service users. A service user perspective informs the analysis in this section: although several members of the research team contributed to the analysis through reading transcripts and discussing themes (see above), a service user researcher (AS) led on the process of analysing, interpreting and writing up the service user data, with input also from a second service user researcher (SF). Illustrative quotes appear in italics, with IV denoting the interviewer and IE the interviewee (participant).
To facilitate interpretation of our quantitative findings, a further focus in this account of the thematic analysis of our findings is experiences of peer relationships: a discussion of these and of their impact on relationships with staff follows the main account of service users’ experiences and views of relationships with staff.
What do service users want from their relationships with staff?
Participants were asked to describe how they would like their relationships with staff to be, and to compare this with the reality they experienced. There was a strong degree of convergence between acute ward service users and crisis house residents in the relationship attributes they valued. Key for both groups was that staff have the right personal qualities:
What kind of things do you want from your relationships with staff?
Well, I just want . . . I just want a caring person.
Uh hmm.
A person that’s going to listen and that’s what they do do; they listen.
Acute ward service user
The next section explores in more detail the attributes valued by service users in both settings. Put simply, people want staff to be decent and kind, ‘the ones that are being nice are best suitable for the position’ (acute ward service user). A service user in a crisis house wanted staff to walk the line between helping and not being intrusive, whereas a service user on an acute ward wanted staff to find the balance between kindness and taking control, but not being so controlling that people lose the ability to take responsibility for themselves. Many people described wanting staff to be friendly or even to act like friends, for example chatting about one another’s lives. This was particularly true for people on acute wards, perhaps because this group tend to stay longer and so come to see staff as significant members of their networks.
I’d like them to stop treating me like a prisoner and let me go back into the world . . . I don’t want them to be my warders.
Okay.
I want them to be my friends.
Acute ward service user
A second prominent theme emerging was that service users in both crisis houses and on acute wards wanted staff to talk to them more, to listen to them more, and to demonstrate therapeutic counselling skills. For some people, this was the most important attribute of the relationship between staff and service users. This talk could be either general chat to find out how your day is going, or more structured therapeutic talk to help people understand themselves and their traumatic experiences. One service user described the need for more than ‘just tea and sympathy’.
One-to-one time for specific emotional support, problem solving, specific worries, um, or, you know, if you’re in a patch when you’re very depressed, and you want to talk about how you’re feeling and, um, but also general ward interaction, so there would be that, um, kind of hello as you went by or a little chat while somebody was opening the laundry room, and I would like more of that.
Acute ward service user
A third emerging theme to arise was that both crisis house and acute ward service users wanted a focus on recovery and hope and support in achieving this, although this theme was not as prevalent as the personal qualities of staff or the desire for more therapeutic relationships. Examples included help to move forward, to identify support outside of hospital, support at night, goal-setting for positive change, help to feel better about yourself, and being given the chance to recover.
If someone's really positive about your recovery it's so reassuring and gives you hope as well because sometimes the thing you need more than anything in these places is hope.
And attitudes they have about mental health. I think some people see you as an ill person and more interested in just managing your illness rather than promoting your recovery.
Acute ward service user
It’s like a prison and this shouldn’t be treated as a prison. It should be treated as if it’s here to rehabilitate people and to help people to get them back into their lifestyle so that they can do the things and get through life the way they should.
Acute ward service user
In the crisis house in particular, participants described the importance of having staff who had an in-depth understanding of their difficulties and situation. For some, this meant employing staff members with relevant life experience, including people with experience of acute mental breakdown or distress.
I find that actually the staff who, who have been service . . . service users are the ones who tend to be the one who I can, who I can work with better because of their, their, their backgrounds, um, I think that’s very true, um. There’s a depth, there’s a depth, you know, someone’s been through, ah, has been, is still, you know, a, a mental health user, um, then there’s bound to be in their experience, ah, a depth that maybe someone who hasn’t, who’s come through
Crisis house service user
It also meant being understood as an individual, rather than a diagnosis or just another person moving through the system, ‘I really want to be seen for who I am, not just a number’ (crisis house service user). Some participants also wanted staff to have a greater understanding of what it’s like to be experiencing acute mental distress and residential crisis care. Acute ward service users placed less emphasis on staff’s personal experiences, but often described wanting relationships that were trusting and where their word was believed, where they were listened to, and not provoked or baited by staff.
Key values and personal qualities that promote positive relationships
Interviews explored accounts of relationships and interactions with staff, aiming for an in-depth understanding of the qualities, behaviours and attitudes that resulted in service users feeling that staff had got relationships with them right.
Basic human qualities lie at the heart of all therapeutic relationships
Underpinning all positive therapeutic alliances between staff and service users are the basic human qualities of staff. Service users in both environments valued relationships with staff who are caring, kind, honest, trustworthy and trusting in return, non-judgemental, reassuring, warm, friendly, mature, helpful, real or genuine, relaxed and calm, sincere, approachable, respectful (of people and confidentiality), empathic, and who have a sense of humour.
It’s sense of humour, calmness, ah, inner serenity, um . . .
So it’s really about the personal qualities of staff?
I think it’s more the personal quality, rather than what qualifications they’ve got, exactly, rather than what qualifications they’ve got. Definitely, definitely, yes.
Crisis house service user
How would you describe the right attitude?
Being caring and understanding, giving the people a chance that come here that have a issue with their mental health. And are experiencing like difficulties with their thoughts and all of that.
Mm. Did you say, give them a chance?
Yeah. The ones that give . . . like the ones that give you a chance to recover, and always say good things to you and be nice to you, instead of getting on your nerves, and making, provoking you to become aggressive, and ah, get upset and that, make you want to defend yourself against them. The ones that are being nice are best suitable for the position.
Acute ward service user
Most service users felt unable to build trust or rapport with staff members who did not demonstrate these positive human qualities. As the quote above suggests, these staff members were often seen as having the wrong attitude or approach for the job. Inappropriate personal qualities described by service users included staff who are uncaring, rude, disrespectful, hurtful, confrontational, belittling, infantilising, unconcerned, untrustworthy, anxious, distracted, insincere and who take advantage of service users.
It would be nice to be recognised and not judged, analysed and scrutinised and criticised
Crisis house service user
Encountering and engaging with staff members who are seen as wrong for the job could have a deeply negative impact on service users.
So what’s more important? Is it their professional background for example?
No, it’s about, I think it’s about they really care about what they do, yes. And some of them don’t really care. They just do it. I don’t know why. I don’t feel they care.
Hmm, how does that make you feel?
It makes me feel um, that even, like even more lonely.
Yes.
Yes, because people who are meant to help me are not the right people for the job.
Acute ward service user
A vocation, not just a job
In both crisis houses and on acute wards, service users described encountering staff who were dedicated to their profession and who, as a consequence, service users felt able to trust and form strong therapeutic relationships with. These dedicated staff members were often described as dependable, deeply compassionate, reliable, responsible, keeping to their word, knowledgeable, and as going out of their way to help.
But I must say they’re quite lovely, and they go the extra mile. Ah, they do more than they should do, over and above their, their normal duties.
Crisis house service user
Because, don’t want to do it out of the sake of knowing that, like, they’re doing it for a job to get their wages to live their life . . . They’re doing it because they care.
Acute ward service user
Conversely, there were a number of staff members that service users felt unable to form relationships with because they were seen as simply doing a job. These staff members were described as uncaring, unbothered, going through the motions and simply being there to collect their wages. This experience of staff was far more prevalent in hospital wards, although one or two staff members in crisis houses were also described in this way. Service users tended to avoid these staff members, although they were still at times able to exert a negative influence on them:
If it’s someone who’s just sort of laying down the rules and that’s it, uh, or just doing their job, then you think well, you know, I won’t bother talking to them at all.
Crisis house service user
Sometimes, ah, certain members of staff, they can be a bit difficult, and then with . . . because of their attitude; they don’t have the right attitude for the job, really. They're supposed to be a carer, and the guy doesn’t care. He just cares about himself; he doesn’t care about the people, he just wants to get his money and go home. He wants to keep his job, basically, but makes your life hell while he's doing it.
Acute ward service user
The ambivalence of professionalism
Service users in both environments valued relationships with staff who were able to be professional and maintain professional boundaries.
You’re here for mental illness, it’s a very lonely experience, and you want support from professionals, you want them to be professional, regardless of what you think.
Acute ward service user
Being professional meant respecting confidentiality, treating people in the same way regardless of one’s personal views and the views of the service user, being dedicated and dependable, and being able to respond to distress and agitation in a calm and containing manner:
I don’t know. Like, you can have . . . you can have a laugh with them, and even when, like . . . because I’ve got a short temper . . . even when . . . shout, you know, going a bit off on one, they stay very calm and it’s . . . yeah.
Mmm. What effect does that have on you, if they stay calm?
Calm. It makes me think . . . I . . . it doesn’t rile me up even more; it makes me think . . . it makes me stop and think, and think, calm down, it’s not that person’s fault . . .
Yeah. What about . . .
So why . . . yeah, because I’m shouting at them, then . . . it’s not their fault, it’s for other reasons, and because they stay so calm and understanding it makes you . . . it makes me stop and think and it’s like, hold on.
Crisis house service user
At the same time, service users in both crisis houses and acute wards valued staff who were at times also able to step beyond the boundaries of their professional role in order to respond in a straightforward human way to individual situations. This could be as simple as giving someone a cuddle, exchanging personal information or engaging in general chat.
I said to her, you know, you’re telling me I do too much, what about you? She laughed, and she said, yes, you know. So I think, again, there’s an acknowledgment of, you know, you don’t have to put up this professional wall, but actually you’re really quite natural and comfortable about talking about yourself as much as the other person.
Crisis house service user
right now I'm seeing the best consultant I've ever seen. The rest of them are all not as good as this one.
Mm? What's good about this one?
This one is very understanding. She's very down to earth, like, and she, she's a normal person, basically. She's not a . . . she's not a job. She just be’s herself. And she, she knows how to talk to people.
Acute ward service user
Similarly, relationships with staff members who were able to be professional but who lacked basic personal qualities such as warmth, kindness and understanding were not valued:
I think some people, almost like mental illness kind of repulses them a little, just they don't really want to engage with someone that's mental ill or talk to them in a friendly way. They're professional and they'll do their duties but not really want to be warm and inclusive like some of the other members of staff.
Acute ward service user
Thus service users valued aspects of professionalism such as ability to respect confidentiality, treat all equally regardless of personal views and respond non-judgementally to all life situations and difficulties, yet at the same time they very much valued the capacity of some staff to maintain but also go beyond professionalism, using their own human qualities and experiences to form positive relationships.
The importance of having staff who are engaged, interested and understand service users as individuals
Service users in crisis houses and on acute wards were able to form more positive relationships with staff members who they felt were engaged with and interested in them. These staff members listened carefully to what service users had to say, and often went out of their way to touch base and check on people.
So what kind of things make your relationships with staff better, do you think?
Staff coming to you and talking to you, like rather than you have to go and talk to them, and say hello, or stuff.
And does that happen?
Yeah, some do.
Crisis house service user
Being engaged also meant showing an interest in people and their lives, which could be as simple as asking someone how their day is going. Although this may seem like a small act, to service users it signalled whether or not the staff member cared and had a big impact on how they felt about themselves and their stay in that residential crisis facility.
When they're being really nice, what kind of things might they be doing or saying?
Um, just when I ask for things they'll be getting me things, sometimes coming up to me asking how my day is and how am I getting on and stuff like that.
And how does that feel when they come up to you . . . ?
It feels that you're, you're important, yeah, and you've got a reason for being in here.
Acute ward service user
Conversely, staff members who did not engage with or demonstrate an interest in people – that is, those who were unresponsive, who didn’t listen, didn’t make time for people, didn’t engage in small talk, and didn’t get to know people – were the staff members who people often felt unable to form relationships with. In crisis houses, such staff members were often seen as aloof or standoffish. However, on acute wards, rather than being aloof, these staff members were sometimes seen as actively ignoring service users.
The ones that are quite distant and I don't have such a good relationship with, they don't really ask how my day has been or they don't really interact with me, so, yeah, they just ignore me, kind of thing.
Acute ward service user
What kind of things do you think make your relationships with staff worse?
Um, just being ignored. Um, just little things like, um, just them not . . . just, like I say, asking for things and them not being given or not being given the time of day
Acute ward service user
Staff members who are engaged and interested are able to get to know service users better and, therefore, can understand them as individuals. This means that these staff members are better able to meet service users’ needs and to vary their approach according to individual circumstances.
They’re very sentimental and understanding if you’re upset and, um, like if you want a break you can go outside for a cigarette or go and get a coffee and come back or whatever, so I think they’re quite therapeutic.
Mmm. What do you mean by sentimental?
Understanding, they don’t just treat everyone the same, you’re all treated individually. So if somebody needs more time because they’re more emotional, like me, in the house then I tend to get more talk time.
Crisis house service user
Some people who had experienced both crisis houses and acute wards felt that staff in crisis houses were able to get to know people better because of the smaller ratio of staff to service users. These service users therefore felt more understood when they stayed in crisis houses. Some people in crisis houses also felt more understood by peer workers who shared their life experiences.
Staff members who were disengaged and disinterested were often unable to understand service users and their needs. Sometimes service users felt that staff didn’t understand mental health or the service they were supposed to be providing. Feeling misunderstood could exacerbate distress and cause people to withdraw, undermining both communication and the therapeutic alliance. There were examples of this in both crisis houses and on acute wards.
I have my days when I just don’t even want to have a one-to-one sometimes because I just feel like they don’t understand what I’m trying to say to them, so each staff member is really different.
Crisis house service user
This service user felt frustrated by her failed attempts to get staff to understand her.
I don’t know how to make them see or make them understand. They always say, tell you how. . . tell us how you feel. I write down how I feel. I gave it to them. It didn’t make a difference. It’s just like, what can I really do?
However, one service user in a crisis house explained that staff weren’t able to understand her because she felt unable to open up to them, while a service user on an acute ward also suggested that this was a two way process.
Other factors that influence therapeutic relationships
A substantial number of other factors, both contextual and related to individual staff behaviour, emerged as important influences on the perceived quality of the therapeutic relationships developed by staff and service users. An overview of the most prominent of these themes follows.
Freedom
Service users in crisis houses deeply valued the freedom that they had to come and go and to use different parts of the crisis house such as the kitchen and gardens. This was of particular importance to people who were new to the system and who didn’t know what to expect, and to service users with experiences of involuntary detention.
You know those old films you can, well I’ve never been to a place like this before so it’s kind of difficult so, um, for example I’ve got this idea of what a kind of institution, I don’t want to say mental institution, do you know what I mean, what they’re like, they’re not very nice places. They’re very cold and, um, not very nice places to be in. This is completely the opposite, you know, it’s not like you’re quarantined anywhere, this is more like freedom, it’s more like you can walk in, walk out, if I want a cigarette I can just go out and smoke a cigarette, if I want to go and have something to eat then I can make some food. I can travel from the second floor to the first floor, first floor to the second floor, I can join the classes if I want. If I don’t want to I don’t have to.
Crisis house service user
These freedoms exerted a powerful positive influence on the dynamics between staff and service users and, therefore, their capacity and potential to form therapeutic alliances. For some people who had previously been admitted to hospital, the novel freedoms of the crisis houses directly affected their perceptions of staff.
I think the staff are better here, I don't know. Yes. Because it's more relaxed here, you don't have to keep asking if you want to go out for a cigarette or something, and that. Because in hospital you get certain times you want to go out on, like morning, lunchtime, or afternoon.
Crisis house service user
The majority of service users we interviewed on acute wards had restrictions placed on their freedom associated with being detained under the Mental Health Act. 47 However, even voluntary service users felt that their freedoms were limited, for example their ability to leave the ward was dependent on a staff member being available to open the door, and access was also limited to various areas of the ward, notably the kitchen. Most service users felt that this lack of freedom immediately established a negative dynamic between staff and service users.
Because I resent them for not trusting in me after they know me. And they know I’ve got a kind heart and they know I’m really not . . . a non-harmful person. Why can’t I walk down now and get a [chocolate] bar? I think, I want to go and get a [chocolate] bar now. Would they let me? No, because of my leave, I’m not to be escorted.
Acute ward service user
As this quote suggests, freedom was bound up with notions of trust, respect and responsibility, with many service users feeling as though staff did not trust or respect them if they weren’t given basic freedoms.
Start trusting the patients that you’ve got in the wards instead of like basically treating them as prisoners
Acute ward service user
Conversely, those with leave, or at least more leave than they had anticipated having, felt as though they were trusted and this had a positive impact on the dynamic between staff and service users.
I’ve already got leave so that’s incredible, isn’t it? . . . They trust me that much that they’ve given me leave already, straightaway.
Acute ward service user
The frustration and anger caused by people’s lack of freedom also had a major impact on the dynamics between staff and service users. Many service users described themselves or their peers as ‘kicking off’ because of their deep frustrations on the wards. People were particularly angry where agreed leave did not occur because there were insufficient staff available to escort people outside or on cigarette breaks. For some people this meant that staff were seen as prison wardens and it appeared to undermine the nursing role, impeding the development of therapeutic alliances.
You can't really have a relationship with the people you see as holding you captive, even though it's done with the best of intentions. I think with me it made me disengage and not want to . . . I felt I was being wronged and I should be free.
Acute ward service user
Many acute ward service users felt that increased freedoms would bring huge benefits. First, it would increase recovery as people would have access to fresh air, light and space, rather than being pent up in untherapeutic and boring spaces (see below regarding activities and environment).
There wasn't really a garden accessible or someone had to take you down and they were too busy, and not having the fresh air I think was bad for my mental health: not having any sun, not having any fresh air, not having any exercise. I think they worsen your mental health. I think those are things that people say you must have, whatever kind of mental health problems you're suffering from, but the very environment that you're trying to get better is like you lack those things. Obviously there are safety concerns and stuff but it would be something I think would make people calmer as well and get better more quickly.
Second, service users would benefit from the sense of autonomy, which they believed would positively impact on their relationships.
Well, you just feel more of a, more of an adult, more of a sort of cogent, responsible adult because you’re going off on your own, so that affects the way you feel about yourself, and therefore the relationship with the nurses is different.
Acute ward service user
Third, increased freedoms would mean that individuals would feel less angry and frustrated, and consequently less irritable and irate with staff, thereby improving relationships. Finally, increased freedoms would mean that the environment would not be as volatile as people would feel less trapped, and this would directly benefit the relationships between staff and service users because:
There would be a calmer environment on the wards so I think the staff would spend less time putting out the fires and arguments and people getting stressed and getting irate. So I think it would be beneficial to the staff, and then there's a knock-on effect where the staff have less to deal with on the ward, like altercations or people being aggressive, then we can have a better . . . they can spend more time with the other patients and stuff, like therapeutic interactions instead of just stepping in where there's a problem.
Acute ward service user
I think the ward atmosphere can feel a bit like being inside a pressure cooker, and, um, if you don’t have the freedom to get out, it can lead to explosions.
Acute ward service user
However, not every service user agreed that freedom had a major impact on relationships: one participant felt that staff were simply doing their job in limiting people’s freedoms, meaning that the therapeutic alliance was largely unaffected.
Interestingly, some people in crisis houses described negotiating their freedoms on a daily basis, discussing their plans for outside visits with staff. Staff then ensured that the plans were safe and wouldn’t put the service user at risk. This negotiated freedom was viewed positively, and sometimes meant that service users felt safe and supported, rather than angry and incarcerated. Conversely, service users in acute wards did not feel listened to regarding their requests for more freedom. Instead, one acute ward service user described gaining leave as being ‘like playing chess’, with another stating that you have to play the game in order to be awarded leave. Where service users feel that they cannot be honest and open, the therapeutic alliance will inevitably be undermined.
It makes you feel like you have to convey a certain impression to them in order to win your freedom or whatever. So sometimes I think you feel like you have to engage with them in a certain way or you often hear patients here say, oh, you've got to play the game. To get out you have to play the game.
Acute ward service user
Autonomy and responsibility
The extent of personal autonomy and responsibility experienced by service users appeared to vary between crisis houses and hospital wards. In crisis houses, service users sometimes felt that they were expected to take responsibility for themselves, for example identifying support that they could access once they had left the crisis house. However, some service users found this difficult to do when in crisis and felt that they needed additional support from staff.
I feel like I came . . . I know it’s a crisis house, but I feel like I’m doing everything myself, which if I was out there I would, but because I’m here I feel like I’m having to refer myself to certain different places and I need just that extra, just that extra boost, kind of thing, just knowing that they’re, that they’re there, because sometimes I feel like they’re not.
Crisis house service user
However, others welcomed the personal autonomy they had in crisis houses. The following service user compared her experiences favourably to those on a hospital ward where she had no say in her care and treatment.
The people responsible for your care weren’t interested in what you had to say. They didn't talk to you and explain what they thought was wrong and how they were going to treat you. They just dump you there and pump you full of massive doses of drugs, uh, the wrong ones. Here it’s uh, it’s completely different. I have a certain amount of control over what, what medication I take insofar as I can refuse it if I, if I don’t want to take it, or I have a certain amount of autonomy.
Crisis house service user
Indeed, many service users on acute wards described lacking freedom and autonomy. This sometimes left people feeling controlled by staff. Others described being dependent on staff for everything, down to the littlest things such as asking for a cup to make a cup of tea. This dependency could be infantilising.
I like to be quite self-sufficient and be able to do things for myself than to constantly have to . . . I found it frustrating to have to be interrupting people and knocking. It puts you in quite a . . . like a subordinate position when you have to knock and knock and wait for someone to look up from what they're doing, and sometimes they don't look up from what they're doing.
Acute ward service user
One service user felt strongly that people on wards should be given more responsibilities to help prepare them for the outside world, such as cleaning and cooking. Another service user felt angry that her lack of autonomy extended to her personal space, where she felt that she wasn’t trusted enough to have privacy, a key to her bedroom or somewhere to lock her personal possessions. Such rules undermined personal responsibility and were also seen as overly bureaucratic and lacking in common sense.
I think they could, I think they could relax some of the rules and give a bit [or lot?] more trust to us because I think some of the rules, like, you’re not allowed a lighter or you’re not allowed to have your phone charger in your room, or . . . All these things are a bit ridiculous because if I, I could take my belt off and strangle myself if I wanted to. So, it’s like, you know, they’re just creating . . . I hate the silly rules. I don’t know whose fault it is but apart from that they’re wonderful.
Acute ward service user
However, for one service user the lack of personal autonomy she experienced was a source of comfort and safety.
Almost I think at times a sense of staff as parents. Um, I feel that somebody’s in a quite maternal role when you’re feeling small and lost and needy . . . at those times when you’re feeling quite vulnerable, that sense that parents are around . . . there’s somebody, um, responsible, you know, and, and in some way adult who’s looking out for you.
Acute ward service user
This same service user later described the benefits of having responsibility taken away:
Sometimes when I feel very lost I want to be told what to do, I want to be told . . . at some level I want to be told no, you can’t go out, you know, you’re having a lot of urges and you aren’t safe to do that at the moment because I can’t make that decision myself because I’m so torn between urges and not wanting to, and if the urge part takes over I’m going to bloody well go and do it. Um, but if somebody sort of puts a spoke in the wheel and says actually no, you’re not going out by yourself today, so being quite directive and taking over really, taking over responsibility for a while.
Acute ward service user
Atmosphere and environment
There were important differences in the ways that service users described the atmosphere and environment of crisis houses compared with acute wards. The majority of crisis house service users described the atmosphere in positive terms, such as pleasant, homely, relaxed, peaceful and quiet. Similarly, the physical environment was often described positively, with many appreciating the space, freedom and privacy, and their access to the kitchen and gardens. Space in the crisis houses was also to a greater extent perceived as shared between staff and service users. Many service users felt that environment, atmosphere and sharing of space had a positive impact on their relationships with staff.
It’s just calming, it’s relaxing.
Yeah.
It’s safe.
Yeah.
It’s like, it’s safe. As soon as you go out there, it’s not safe.
Yeah, yeah. And that . . . the fact that it feels quite calm and relaxed and safe here, does that have any effect, do you think, on relationships with staff?
Yeah, because I think they must have a good working relationship between them, and you know, you see staff walking around and it’s not like they’re just kept away and that’s your space, that’s their space; I mean, it’s nice, you see staff sitting at the table with people and stuff like that, and that feels [unclear] . . . it’s just it makes it a nicer atmosphere.
Crisis house service user
Service users on acute wards were evenly split regarding whether they experienced the atmosphere and environment positively or negatively. For some the acute ward was relaxed, quiet, easy-going, merry or friendly, although the facilities were only occasionally described favourably. The following acute ward participant described the atmosphere as ‘good’, and went on to explain:
When the atmosphere’s good everyone feels happy; there’s a virtuous circle, absolutely.
Yes. Can you describe a good atmosphere?
A good atmosphere, we’re all having a laugh, chatting, making jokes. A bad atmosphere is somebody’s really upset and angry because they haven’t got what they wanted, normally a patient. They haven’t got their leave or their cigarette or they’ve been refused something. That can create a bad atmosphere. But we all know that they’re having a difficult time so we just give them their space.
Acute ward service user
As the above quote suggests, other service users play a crucial role in creating the atmosphere. This was equally true in crisis houses and on acute wards. Participants often described the atmosphere as variable or volatile, which was partially dependent on which service users were staying there and how they were feeling. This has a big impact on relationships with staff: where the atmosphere is calm, staff have more time to talk to people and attend to their needs.
How would you describe the atmosphere here?
It’s very, variable. It depends on who the patients are, how well or unwell the patients are, who the staff are, um, how calm or chaotic the ward is, um, all those things play enormous, affect enormous . . . enormously affect the atmosphere in the ward.
Hmm, and again, could the, those changes in the atmosphere; can they affect how your relationships with staff are?
Yes, because the staff become very busy, they don’t have time, their immediate priority is crisis intervention. They don’t have time for the people who are getting by.
Acute ward service user
Both crisis house and acute ward participants described some negative aspects to the atmosphere and environment, although it was relatively uncommon for crisis house participants to do so. The few who were negative about crisis houses described the environment or atmosphere as busy, tense and rundown, ‘a bit grim’.
What would you say the atmosphere is like here?
A bit tense.
Really?
Yeah. And I even get scared to knock on the office door sometimes for, like, they’re thinking, oh, what does she want now or, like, some of their looks when they see me, and it is a bit awkward. It doesn’t feel like it . . . how it should be like, kind of just comforting. It doesn’t really feel like that.
Crisis house service user
On acute wards, the most common descriptions were of an unrelaxing atmosphere with constant and intrusive noise: this is a direct point of contrast with the descriptions of crisis houses. For some participants, an inability to escape the noise was very damaging to their recovery.
So you said it's not relaxed. How else would you describe the atmosphere here?
[Pause] As they keep banging on the window, on the doors, like too many doors being opened and shut, opened and shut, you know. You can hear it, too many doors being slammed and maybe if it was bigger you wouldn’t hear the doors being slammed, ‘cause it does get to you.
And how does it make you feel with the noise going on?
It just gives you a headache and just I want to sss . . . there's things stopping me from screaming out.
Yeah, yeah, so it makes you feel like screaming, screaming out?
Helpless, yeah.
Acute ward service user
The physical environment of acute wards was also described unfavourably. Participants sometimes referred to a lack of space, lack of fresh air, lack of access to outdoor and green areas, and small claustrophobic rooms. One participant summed this up in the following way:
The hospital, it feels, it makes, it makes you feel like you're inside of a bottle of, uh, medicine, you know
Acute ward service user
Again, this was a direct contrast with crisis houses and it had a negative impact on relationships between staff and service users where service users became tense, angry and frustrated. This was compounded by the lack of things to do on the ward.
Most people are just bored. It’s not, it’s not a particularly happy place, unless people have leave, and they’re much healthier, they’re a lot better in themselves. But mostly people are just bored.
And that sort of atmosphere of being bored and not having enough to do, does that affect relationships with staff?
Yes, of course it does, people get frustrated and agitated.
Acute ward service user
Unsurprisingly given the above, service users in crisis houses and on acute wards who had experienced both types of residential care tended to describe crisis houses far more favourably.
Activities
The availability of activities for service users, and the extent to which staff and service users engaged in activities together, were further key factors influencing relationships. This was particularly true on acute wards where many service users felt trapped in a sterile environment and described boredom as one of their biggest problems. For many, the only activity on offer was watching television.
Hardly anything to do except for watching TV.
Acute ward service user
Can you tell me a little bit about what the staff are like here?
Most of them are pretty good, but I think there, uh, uh, it’s very boring here, there isn’t much put on here, especially if you’re a, I was two weeks without anything to do, where I wasn’t allowed to do any occupational therapy, I wasn’t allowed off the ward, I wasn’t given any reasons for that, and so it’s very boring, it’s just the TV on all the time.
Acute ward service user
The constant noise emanating from the television was sometimes experienced as antitherapeutic. It was clear that service users felt that the lack of activities, beyond watching television, caused intense frustration and damaged staff and service user relationships.
It's just that constant presence on the ward, trying to fill 12 hours a day awake or whatever. It's frustrating, it makes you frustrated with the staff.
Acute ward service user
One service user described the Early Intervention Service as far more helpful and therapeutic than the ward because of the activities on offer. Staff from the service attended the ward to enable this person to participate in activities.
Early intervention, they will um, arrange badminton sessions, picnics, sports activities, stuff like that, and that kind of helps me forget about whatever I’m having, you know, and just enjoy for the moment, and every week I look forward to those things.
Acute ward service user
Some participants in crisis houses also described a lack of activities.
I seem to have spent most of my time here just drinking tea and smoking cigarettes.
Crisis house service user
Although this lack of activities was less damaging than in acute wards, largely because people have the freedom to leave, one participant nonetheless felt that it was damaging because it meant that service users had nothing to distract them from their difficulties. Furthermore, participants in both crisis houses and on acute wards felt that shared activities between service users and staff would have a positive impact on relationships, increasing interaction, helping people get to know one another, building a sense of community and alleviating boredom.
Yeah, the patients and the nurses, and have more, like, games and stuff like that so you can get to know the person.
Acute ward service user
So what kind of things do you think makes your relationships with staff better?
When I’m more occupied, when I’m doing things. I’ve started sketching, reading, playing table tennis, and when I’m more occupied, that makes my relationship with staff better.
Why’s that?
Because I’m getting along with things, I’m not bored. Because they’re quite isolated in there, there isn’t much interaction, and there isn’t any put on on the ward, to, to do as such, there’s a table tennis table, everything else is broken, there’s a keyboard there that’s broken. There is a xylophone, that we can play, but table tennis and xylophone and TV, that’s the only, they’re the only entertainment things that there are, or things to do. I’ve asked countless times, can I help with the washing up, can I help with cleaning, can I help with something, they say, no no no, it’s not your job, it’s not your job.
Acute ward service user
Equally, it was notable that when describing good working relationships with staff, service users often referred to shared activities. On the contrary, poor working relationships were often characterised by a lack of activities, with staff and service users largely ignoring one another.
Yeah, I go out with them, um, go to the park and play football, um, just have a good interaction with them. Um, they take me, yeah, to the park, to the shops, stuff like that. They've got time for me. It's like as if they've got time for me, so . . .
Acute ward service user
In one recovery-orientated crisis house, service users were very positive about the range of activities available.
Obviously there’s um, there’s a wide array of choices of activities to engage in um, while here. They um, the floor below this one is full of such, different rooms or different groups and activities, which the residents are encouraged to partake in, although there are not too many residents um, who make use of the um, facilities provided.
Crisis house service user
This service user felt that the activities were not well attended. This may partially be explained by a lack of information, as a further service user reported not being told about activities until well into his stay. This lack of information was frustrating.
In some crisis houses staff and service users shared meals together, and this was warmly welcomed by the majority of (but not all) service users who often likened it to being a part of one big family spending time together and discussing the day’s events.
It’s quite good, because then you feel like you are all together, you are all one big family. To, to be honest, I’ve been going to the dining table, not that . . . for food, or, obviously, the medication has been given, not for the food, but somehow it reminds me of home; it’s taking me back to, um, you know, family. I come from a large family; it’s taking me back to my young days.
Crisis house service user
The importance of this in building relationships shouldn’t be underestimated, with some service users on wards recommending meal sharing as an important way of improving staff and service user relationships.
Staff visibility and availability
Participants from both crisis houses and acute wards often felt that staff were, on the whole, readily available to them if they had an immediate need for help: ‘They’re available at anytime you need them’ (acute ward service user); ‘most of them are approachable’ (crisis house service user). This meant that service users typically felt able to approach staff, usually by knocking on the office door (as this is where staff spend most of their time). Knowing that someone would be available to talk was often reassuring.
They’re very, in my opinion [background noise], very approachable. You’ve just got to knock on the door if you need to talk. They’ve given me one booked keyworking session a day for about 45 minutes, but if I need to talk any other time or I get urges to self-harm I can go and knock on the door any time of day or night and speak to the member of staff.
Crisis house service user
However, many service users also felt that staff were working under considerable pressures and consequently couldn’t always make themselves available to them. Although service users understood that staff needed to prioritise their workload, this could nonetheless cause some difficulties, such as feeling disempowered or discarded.
It affects the relationships with staff because they have more on their minds. They're trying to manage more and tricky situations and things, and you feel less like you assert yourself and ask for help with things that you need because there's too much going on around that you can see requires more immediate attention.
Acute ward service user
There is that little bit of feeling of being discarded, but like I say, you . . . because I’ve been here now, like, over two weeks, it’s . . . I understand . . . it’s understandable, like, if I haven’t got time.
Crisis house service user
Additionally, in some crisis houses, only assigned workers were available to talk to service users. Again, service users were pragmatic about this.
If you were really in a really crisis, and your worker wasn’t about, I’d put my hand on my heart and say, I don’t think they would ignore you. But other than that, they would tell you to wait, because when you’re . . . when there’s staff changeover and they come on, they look for you and they try to arrange a time, um, so you’d have to stick by that, unless you were really in a crisis. I don’t think they’d turn you away, but that would be the only time.
Crisis house service user
Most crisis house participants had one-to-one time, and very few felt unable to approach staff. This contrasts with some acute ward service users who felt that staff ignored them and rarely gave service users one-to-one time. There was also a sense on acute wards that the extent of availability was entirely dependent on the staff member in question.
Interestingly, in both crisis houses and on hospital wards most staff were said to spend the vast majority of their time in the office.
Do you know how they spend their time?
They’re in the office a lot.
Crisis house service user
Do you know how staff spend their time when they’re here?
Um, they’re usually in the office. Mostly in the office.
Acute ward service user
However, this was sometimes less pronounced in crisis houses because service users and staff shared dedicated one-to-one time as well as meals and, on occasion, activities. On acute wards, the persistent lack of staff presence outside of the office could have a profound impact on service users, the ward atmosphere and the relationships between staff and service users.
I don’t want them locked in the office, I want them out here helping me socialise back to society.
Acute ward service user
No, the atmosphere, it can be that like two different worlds under the same roof . . . So there’s one world that the patients know about and another world that staff know.
Acute ward service user
I don’t think they should be behind that glass booth. I think they should be out and about; I think they . . . It’s too much like a prison. It’s like One Flew Over the Cuckoo’s Nest.
Acute ward service user
You can't build a relationship if you're always in the office. And, like I say, I can't really build a relationship if I'm always in my room.
Acute ward service user
For some, though not all participants, the lack of staff visibility affected their sense of safety as staff were not witnessing incidents as they unfolded, and so were unable to intervene to prevent situations from escalating. However, one service user felt that staff did intervene when necessary.
I think the staff do step in to protect your interests a little bit, so they are . . . usually there is someone on the floor. They're not just in office observing but someone is there to step in if needed.
Acute ward service user
Staff who spent their time in communal areas were described very favourably, with one acute ward service user explaining that relationships with these staff members were much closer. The following quote describes a health-care assistant; these workers were often valued because of the time they spent with and talking to service users.
She’s good because she spends most of her time with us, so she knows, like she’ll be doing people's hair, if they’re down, you know, she gives them a haircut, she gives them um, like she blow dries hair. You know, just, you know, just that can lift the spirit of the person in their day. She applies makeup on people, you know, just little things, and she talks to them, which is the most important. You know, you just want someone to talk to.
Acute ward service user
Length of stay
Service users in crisis houses sometimes described problems being caused by fixed and relatively short lengths of stays and the subsequent impact this had on their relationships with staff. The maximum length of stay varied, but was often 2 weeks (although in some crisis houses this could be longer); typically the discharge date was negotiated. Many service users felt that this wasn’t enough time to deal with the difficulties that had led to the crisis, and felt deeply scared by the prospect of imminent discharge.
Because it’s short term, you can’t work through your, your problems, and unless you get them assisting you with getting your stuff outside, then your problem . . . Well, some people’s crises might go, you know, everyone’s different, but like with my problem, it will not go.
Crisis house service user
Many participants felt that staff did not understand or demonstrate empathy about how scared they were at the prospect of returning home, often to the situation that had contributed to or caused the crisis. This lack of understanding was a source of anger and frustration, and negatively impacted on people’s relationships with staff.
I know it’s only a short time but I feel like I’m being . . . I’ve only been here two weeks I think, but my leaving date is meant to be Monday but I don’t feel ready right now, but it’s like they don’t want to hear that. So I can’t really do much.
Two weeks, in, out. Ask for an extension, you don’t get it, you’re not supposed to be upset or anything about it, you’re supposed to accept that the normal stay is two weeks.
And how are you feeling about it?
Rejected, negative . . .
Crisis house service user
Other influences on therapeutic relationships
Themes that were prominent in a substantial number of interviews are described above, but other influences on relationships were also identified. Service users were well aware of the constraints and pressures to which staff were subject, and funding and staff numbers featured in a number of interviews as restrictions on the development of therapeutic relationships. Service users were also aware of the impact on them of staff morale and stress, with greater difficulty in forming positive relationships when staff were demoralised. Continuity of care and the ways in which allocation of workers was organised was a further theme, with consistency of worker valued, especially where some choice of worker was available so that they could choose to engage more with staff whose personal qualities and approach they valued. Staff personal qualities were more prominent than their therapeutic approach, as already described, but some service users also described valuing contacts with staff whose training allowed them to take a focused therapeutic approach to helping them to recover from the crisis.
The tactics service users employ to survive difficult relationships with staff
We found that service users in both crisis houses and acute wards had developed strategies to help them survive difficult relationships with staff and peers. In both environments, service users had learnt which staff to approach and which to avoid. There were numerous reasons for avoiding staff. Most typically, service users avoided staff whom they perceived as having had poor levels of understanding about either mental health or them as individuals (more typical in crisis houses), an unhelpful or damaging personal approach, or who were experienced as difficult (the latter two being more typical of acute wards). Staff were also avoided where they interpreted service users comments as being indicative of an illness, rather than listening to and believing what service users had to say.
It was I, I realised there's no point in me talking to them because they just use it as a mental health issue, instead of a real life issue.
Acute ward service user
Others avoided staff simply because they were trying to survive their own crisis.
She’s busy working and I’m busy smoking cigarettes trying to get through the day.
Crisis house service user
The second major survival tactic used by service users was retreat. For those in crisis houses, this meant retreating to get through the crisis, or to prevent oneself from becoming aggressive. However, retreat was used far more extensively as a tactic by people on acute wards. Service users here often retreated to their bedrooms in order to avoid all staff, as well as intimidating peers. This often meant that it was impossible to form relationships with staff.
Me personally, I want to stay in my room a lot because, ah, there's no . . . there's no common ground with, ah, some of the staff, so I just stay in my room and don't really do much. Um, yeah, so that could affect . . . Ah, if I came out of my room a bit more maybe I could get to know them a bit better and, um, a relationship could, yeah, happen after that.
Acute ward service user
You come to see the staff as being sort of the prison guards and you're in the prison. It changes how you view the situation, and then you can't really have a relationship with the people you see as holding you captive, even though it's done with the best of intentions. I think with me it made me disengage and not want to . . . I felt I was being wronged and I should be free.
Acute ward service user
One acute ward service user retreated when they found that their needs weren’t being met, and another because they were (in the early stages) too confused and disorientated to engage in conversations with staff and peers.
Likewise, some service users on acute wards describe deliberately withdrawing from situations with staff they found provocative. This enabled them to deescalate potentially destructive situations. However, the following service user questioned whether the onus should be on her to deescalate aggression:
And the only thing I find is when someone raises their voice to me, it raises my aggression. So my best idea now is, just walk away and not speak. I just go like that, sorry. If you didn’t see that, that was the prayer sign, and then I hold my hands, okay, with the back off sign, and then I walk off and I’ll come back later when they’re in a better mood. But I am the patient, so is that really right?
Acute ward service user
Another participant described using both avoidance and flattery as tactics to avoid being targeted by a difficult member of staff:
I try to, I tend to try and avoid her. Actually, what I do is I sort of suck up to her a bit. And [laughing] she’s particularly bad with other people, and I just watch, and think, shit, [laughing] I’m not going to go down that road.
Acute ward service user
Another acute ward service user described drawing on her inner strengths and concentrating on positive relationships with staff members, whereas a second service user described becoming more assertive so that staff could no longer ignore him:
I ask for things a bit more louder and clearer. Um, um, yeah, and I ask for, um . . . yeah, a bit more louder and clearer. Yeah.
Acute ward service user
What are service users’ recommendations for improving therapeutic relationships in residential crisis care?
Service users in crisis houses and on acute wards were asked what residential services can do to improve relationships between staff and service users. The most frequently mentioned theme was for staff and service users on acute wards to spend more time together.
I don’t want them locked in the office, I want them out here helping me socialise back to society.
Acute ward service user
This included having more interaction, engaging in more activities together, and getting to know one another. Two service users in crisis houses also recommended that service users have more structured time with staff through set one-to-one times.
A number of suggestions were given, each by a few service users, with none seeming to have priority. Both acute ward and crisis house service users recommended more therapeutic input, through, for example resident psychotherapists, trauma counsellors or therapeutic groups. Both groups recommended extra training for staff and greater investment in services. For acute wards, two suggestions were made: (1) that staff should be trained to understand better what the experiences of acute distress and receiving acute care are like in order to help them think differently, and (2) that they should receive further training in dealing with aggressive service users. This reflects a more general concern with safety on acute wards. In crisis houses, service users felt that staff needed the support, training and funding to fulfil their role effectively.
One acute ward service user had an interesting suggestion about how to improve relationships both between service users and between service users and staff.
I think they could give the staff extra training in how to have therapeutic relationships because I think I've definitely seen in the course of my admission that there's been an improvement in my mental health through better engagement with staff and better relationships with staff, so I think if staff could be trained on how to promote that change of helping people engage. I think mental health wards could be like a therapeutic community . . . I think the staff could encourage patients that are more well, that are maybe close to discharge, to be able to interact with and support the ones that are still a bit sick or frustrated or can't really see the light at the end of the tunnel or whatever. For there to be . . . because I think that also helps create independence in the person that might be about to leave or whatever. Then there's more of a sense of things going full circle, maintaining a community that's sort of . . . that has a positive impact on people that are often forced to join whether they want to or not.
Acute ward service user
Crisis house service users made a number of additional recommendations including that:
-
greater continuity of staff would result in more consistent and sustained therapeutic relationships and would make people feel more secure
-
staff should encourage peer support and a sense of community
-
it is crucial to employ the right staff members: staff need to be of a higher calibre
-
staff should prepare people for discharge, including structured time to discuss anxieties about going home (the reasons driving people to the crisis house should be discussed so that people are not simply returned to the same circumstances).
Acute ward service users also made a number of additional recommendations:
-
staff should focus on rehabilitation, preparing service users to resume their usual roles and to be able to function outside the hospital
-
more staff should be on duty to increase safety and security
-
staff should make sure that people have the greatest levels of freedom possible
-
staff should respect service users’ privacy and confidentiality, for example not having conversations about service users in public spaces.
Service users’ relationships with peers
A considerable amount of material also emerged from the service user interviews regarding their relationships with peers, of particular interest given our quantitative finding of an independent association between positive peer relationships and service user satisfaction. The following is a summary of the main themes regarding relationships with peers, their impact, and the ways in which they interact with relationships between staff and service users.
Positive relationships between service users
Descriptions of solidarity between service users pervaded reports of both hospital and crisis house care. There was a sense of service users being able to support each other through shared experience. Sometimes this support was in addition to that offered by staff, sometimes it was seen as superior, service users having an understanding of each other that was not shared by staff.
The users are very supportive of one another . . . So you find that you can actually open up quite easily, maybe, to someone who’s [a] service user, rather than [the] them and us, of staff.
Crisis house service user
It’s like the residents help each other out, even though we don’t know each other, but we can talk to each other.
So what kind of, what kind of conversations might you have with other residents that you would find helpful?
Just talking about our feelings, our emotions, just comforting stuff and they always make you feel better in a kind of a way rather than going to staff.
Crisis house service user
We do encourage each other, the patients . . . Yes. So sometimes, um, because I think we know each other more than the nurses know us, we are like in different two worlds under the same roof.
Acute ward service user
The same service user went on to say:
Because we are taking the exact [same] medication, we know what they’re going through . . . So we can call them and say oh, do you feel restlessness? Oh yes, that’s the word I was looking for, they would say something like that.
Yes. So you feel like the patients know more about each other than the nurses know?
Yes, and because we are under the same experience, we can help, do you know what I mean, we can help each other that way.
Acute ward service user
As the above quote shows, service users were aware both of feeling understood, and of being able to help others. The below quote also illustrates how helping others can help an individual.
The fact of the matter is that I’m not the only person with the issues I’ve got, um, and the fact that, you know, um, I can listen to people’s problems and try to help them with them and vice versa . . . But, yes, you know, it helped me help other people or try to help other people with mental health issues and how to get them to feel better. I think that helps me.
Acute ward service user
There was also value in simply being around others.
Just being here I feel good simply because, you know, I’m not alone. I live by myself so, um, you know, company is important to me.
Crisis house service user
In addition, one service user spoke of a relationship with a member of staff who had also experienced mental distress. In this relationship there appeared to be an added dimension, an extra layer of integrity and understanding which occurred while maintaining boundaries:
There’s an honesty, I suppose, he’s gone through all these things . . . and, of course, he’s been through and through the . . . and he’s very, you know. I said, oh no, I need a cuddle, and he said, oh okay, just a boundaried one, he said, quite a luke-warm, you know . . . But, you know, you get someone who’s a nurse and I wouldn’t even think about saying, oh, I need a cuddle, you know. It’s, sort of just that . . . and there is that extra sensitivity and awareness.
Crisis house service user
Staff ignoring other service users
In keeping with the theme of solidarity between service users there were occasional descriptions from acute ward service users, of not feeling that other service users were being looked after. This also occurred in crisis houses.
I don’t feel they took enough time to find out why somebody walked out just simply over the food.
Crisis house service user
It makes me feel that they don’t really care about her getting better . . . There is no strategy for her, there’s no goal setting for her, no one, and only one nurse takes her out.
Acute ward service user
Indifferent relationships between service users
Solidarity between service users was not, however, the sole experience. In both crisis houses and hospitals relationships with other service users were sometimes experienced as just ‘okay’. They were not seen to affect the service user’s overall experience. Similarly, there were descriptions of people keeping very much to themselves.
How do you find the other people staying here, the other residents?
They're okay. I've spoken to one or two. Yes, it's okay.
They're all right. Can, can the other people staying here affect how you feel?
No.
Crisis house service user
People here are decent. Ah, the people that are kind . . . Ah, they keep a lot to themselves.
Crisis house service user
Difficult relationships between service users
It was also possible for service users to be deeply irritated by each other. This occurred in both crisis houses and in hospital although descriptions of irritation were slightly more prevalent in hospital. Although the quote below, from a crisis house service user, shows irritation, it also shows some, perhaps forced, solidarity in the words ‘we have to ask each other’.
People that are really loud, like, and there’s a woman who’s just come, she just doesn’t shut up, do you know what I mean? And you know when you just . . . you’re outside and you sit and have a fag . . .
Mm . . .
. . . and you just want a bit of peace and, like, there’s other people that will come along and everybody’s always like, are you all right? You know, we have to ask each other, but like, there’s particular people and this woman, especially, she just doesn’t shut up, and it gets . . . it riles me.
Crisis house service user
You know, to me personally, the, the only sort of bitchiness is, is among the patients.
Acute ward service user
Because XXX out there, he drives me mad; he always puts the [TV] on. But there are a lot of people here, and he’s very bullshit about it, and he turns it very loud. And it’s lots of things he does that annoys me, but I don’t let him get to me, because I think he likes getting to people.
Acute ward service user
You know, you can stay in your room and stay away from the other patients because it does, some of the patients, they drive you nuts.
Acute ward service user
The issue of boundaries between service users was also raised by two crisis house service users.
I get a bit uncomfortable when people talk about, like, what they’re going through, their problems, really, like . . . out loud.
Crisis house service user
A further service user spoke of a particularly close relationship between themselves and another (crisis house) service user. This was, according to the service user, met with some concern by the staff of the crisis house, as they felt that it was important that service users concentrate on their own issues and not get too drawn into those of others. This issue did not arise among acute ward service users.
Exposure to violence and disturbed behaviour
Two crisis house service users felt that they would be, or indeed had been, exposed to a higher level of disturbance in their peers when in hospital. Despite this perception there were instances where disturbed and disturbing behaviour was reported in the crisis house. This was usually around service users who the participant felt should have been in hospital, and one of these was eventually transferred to hospital.
I certainly wouldn't go to a mental hospital for that, because there’s no containment there. You just get shoved in a ward with a load of really disturbed people.
Crisis house service user
Um, sticking me in a dormitory with a lot of really disturbed people. Not just disturbed but downright violent people, uh, completely insane.
Crisis house service user
Violence and disturbed behaviour was also a prevalent theme for acute ward service users. However, discussions of violence and disturbed behaviour could either indicate its presence or its absence, or the feeling that it could happen. Thus a mixed picture emerged regarding violence and exposure to disturbed behaviour on hospital wards. Some acute ward service users appreciated the absence of aggression whilst others were troubled by its presence.
Like, there’s no violence, there’s no aggression.
Acute ward service user
People are quite quiet and there isn’t that much kicking off.
Acute ward service user
You hear that – there’s a banging up there. She and them went mad, and me and you remain calm. But we’re trying to control our heart rate, which has just risen in both of us.
Acute ward service user
There's some of the patients, you know, who are coming in from, uh, prison and things like that, and it’s quite scary, you know. Some of them might be violent.
Acute ward service user
If a patient decides to start a fight with you, or swear at you or pick on you or something.
Acute ward service user
Theft and the presence of drugs was reported only by ward participants:
Like when I go home, on Monday, then I'll have more of a chance of not smoking cannabis than I do now, because there are still some people that are smoking in the hospital.
Acute ward service user
I had a girl walk in my room today. I told her in no uncertain terms to get out, because she was looking to steal stuff from me.
Acute ward service user
One crisis house service user had a hypothesis about the reason for the greater degree of violence and disturbed behaviour in hospital.
I've not seeing any trouble here like you see in [hospital ward] I've not seen any trouble here so far.
What kind of trouble do you get in [hospital ward]?
Patients, ah, sometimes they sort of kick off or something. I haven't seen that here, so . . .
Why do you think there's a difference? Why do you think people kick off there and not here?
I don't know.
Do you think it's because of the place that they're in, the environment they're in, or . . . ?
Yes, yes. Yes. Because in hospital there's less, it's less freedom, isn't it?
Crisis house service user
Staff intervention in relationships between service users
From acute ward service users, but rarely from crisis house service users, there was an expressed need for the staff to intervene in difficult relationships. This need was sometimes met and sometimes not. The difference between the two settings on this dimension may arise from the difference in the degree of freedom and responsibility afforded to service users in crisis houses and in hospitals, with service users in crisis houses permitted greater freedoms and offered a greater degree of responsibility. Differences in level of disturbance between service users in the two settings may also contribute.
If there’s stuff going on in the living room, and things are getting out of hand, and staff don’t respond, I get very angry.
Acute ward service user
It’s not monitored at all, it’s not supervised, there’s no supervision here, which I find really frustrating, especially with some people.
Acute ward service user
There have been a couple of patients on the ward that have been a bit more vocal and aggressive in their manner than the rest of us, and they've been quite good at stepping in to . . . I think the staff do step in to protect your interests a little bit, so they are . . . usually there is someone on the floor. They're not just in office observing but someone is there to step in if needed.
Acute ward service user
By contrast:
And staff are too, staff are in their office, offices, and before their response, before anything takes place, they actually unknown to see what's happening and it’s too late before violence happens.
Acute ward service user
The effect of other service users on relationships with staff
Around half of the participants saw relationships between staff and service users, and between service users and peers as separate issues.
What I was wondering was whether other people staying here can affect you, can affect how you feel, can affect your experiences, can affect how your relationships with staff are?
No, I don’t think so, no. Because, like, the staff and patients are two different people.
Acute ward service user
However, other service users felt that disruption or a bad atmosphere could affect relationships with staff in a negative way.
What about the atmosphere on the ward? Does that affect your relationships with staff?
It affects the relationships with staff because they have more on their minds. They're trying to manage more and tricky situations and things, and you feel less like you assert yourself and ask for help with things that you need because there's too much going on around that you can see requires more immediate attention. So, yes, it does affect.
Acute ward service user
Um, possibly because you, you feed off the energy. Well, I'd say you feed off the energy. Um, you feel what the other person is feeling, so if they was happy you'd, kind of, be in more of an upbeat mood. It is quite, um . . . it's quite, um, a downbeat mood I'd say, personally, so that would affect my relationship with the staff in a way that my responses might be different, um, depending on my mood. Yeah.
Acute ward service user
Conversely, a quiet atmosphere improved relationships with both staff and service users.
Yeah. Does it also . . . does the sort of quiet atmosphere here, does that also have an effect on your relationships with staff?
Ah . . . it makes them more readily available to talk to you whenever you need to talk to someone. You can always go and talk to the staff, or, or the patients, whoever you think is more appropriate to what you've got to say.
Acute ward service user
So I think if the other patients aren't so demanding then it gives the staff the opportunity to spend time with you when you're discharged or helping you with things.
Acute ward service user
One crisis house service user felt that the overall atmosphere and also the staff to service user ratio made a big difference to their relationships with staff.
Do you think there are any differences, um, in the staff that you met in hospital and the staff that you’ve met here?
Oh definitely, yes. I mean, you get to really know, because of staff ratio, you really get to know what’s . . . you get the chance to know staff here much better than on the ward. Um, and also, you know, there’s more crisis . . . you don’t get that chilled, nice, peaceful, spa-like quality . . . on the ward it’s just different. It’s too hectic, and not the chance to get to know the staff as well as here.
Crisis house service user
The effect of atmosphere and the extent to which other service users were demanding and the impact that had on an individual service user’s relationships with staff was more pronounced in hospital wards. This is consistent with the assertion that the atmosphere is more unstable and that there is more disturbance on hospital wards.
It was also acknowledged that relationships between service users could interfere directly with relationships between staff and service users.
Are there any other ways that the other people staying on the wards can affect your relationships with staff?
If they decide to pick on you.
Yeah?
If a patient decides to start a fight with you, or swear at you or pick on you or something. And then the staff might think it's you . . . and lay the blame on you, instead of the person who’s starting it, because it’s your word against his. Unless they weigh this, ah, for themselves.
Acute ward service user
The service user describing this incident went on to say that they would try to explain their side of things to staff but would not bear a grudge if the staff read the situation wrongly.
Interestingly, the one crisis house service user who acknowledged the possibility of this sort of thing happening felt that the crisis house staff would not apportion blame incorrectly.
No, I don’t think so. The only time that possibly could happen is if someone goes, um, if there’s either any sort of . . . or someone makes up an incident, because that’s more likely to happen, and they go to the staff. I would like to think XXX would play fair, if someone goes with an allegation, they don’t just pass judgement.
Crisis house service user
Results of thematic analysis: staff interviews
Initial analysis of transcripts from the seven interviewed hospital staff and six crisis house staff also focused on identification of the main broad themes relevant to therapeutic relationships, with examination of commonalities and divergences with service user perspectives.
What do staff think service users want from their relationships with staff?
Staff’s views of what service users want from relationships with staff showed many parallels with the equivalent responses from service users. Their perceptions of this centred around a cluster of interpersonal skills in which listening (often described as ‘active listening’), understanding, warmth, empathy and respect are central. Other personal characteristics that are frequently raised are being patient, being ‘generous’ with time, and showing interest in the person as an individual.
Having that sense of being listened to and understood, and treated as an individual, rather than, kind of, just a, another bed, another bed space.
Crisis house staff member
A respectful attitude of accepting the person as they are, being open-minded, friendly and non-stigmatising was also thought to be important. One crisis house staff member described this as ‘willingness to kind of be open to people and meet them where they are at’ (crisis house staff member). Trust and honesty are also key attributes commonly cited by staff:
There cannot, and will never be, a good working relationship without honesty and transparency. Transparency in that you answer questions as honestly as you can. Honesty in that you don’t patronise them, because they are not children.
Acute ward staff member
‘Professionalism’ and a cluster of similar characteristics such as reliability (e.g. doing what you have said you will do, or if not explaining why), clear communication (e.g. around rules of the service or treatment plans) and setting boundaries were also thought to be important. Staff think service users want ‘someone who they think knows what they’re doing’ (acute ward staff member). However, professionalism on its own is not seen to be sufficient: there is an acknowledgement that the value of a professional approach comes from its combination with the above mentioned interpersonal skills and personal attributes. This mirrors what service users say they want (see The ambivalence of professionalism).
Although there were no clear differences between crisis house and ward staff in what they think service users want from staff, one interesting difference was that some crisis house staff talked about staff’s personal skills in relation to broader features of the staff team or the service as a whole.
When I first started working here it was obvious that, you know, there wasn't anyone who wasn't talking, instantly approachable, and instantly, like, affable, you know, I think there's a very strong sense of affability within the team, um, when people come in, um, and no-one with a very, kind of, harsh [ . . . ] harsh demeanour, I suppose, you could say. Everyone's very, kind of, open and welcoming, and affable.
Crisis house staff member
Finally, although staff and service users are generally in accord in their views of what service users want from relationships with staff, there was only one mention among staff respondents of instilling a sense of hope and optimism about recovery. This was a stronger theme among service user participants.
Staff views of their working style
Not surprisingly, much of what staff thought important in their working styles overlapped with their views of what service users want from relationships with staff, discussed above. Warmth, humour, caring, empathy, a non-judgemental attitude, fairness, honesty, understanding, and good communication and listening skills were commonly mentioned characteristics. Respect in relation to various aspects of relating to service users was commonly mentioned. Staff talked about respecting service users’ privacy (e.g. by knocking on service users’ doors before entering), beliefs and preferences, and about relating in a way that validated service users’ experiences. The ability to inspire hope was also mentioned.
I think a lot about isolation and trying to break down that isolation and having, kind of, lasting memory of somebody that actually, you know, listened to them, took them in, cared in some way, helped, you know, had helped them to, kind of, hold onto some hope, thought about, you know, that they’re going to get out of a crisis, that things can change.
Crisis house staff member
Seeing and trying to understand the person as a whole and unique individual was also seen as important. This idea was described variously as working ‘holistically’ or using a biopsychosocial model. The following extract exemplifies this in relation to a service user who was encouraged to use existing skills as a guitarist to teach other residents to play the guitar:
It’s actually finding out more about the person rather than the crisis and taking that forward and trying to get them involved. He said that he wanted to give something back but, um . . . and obviously this is an opportunity. So I think the rapport comes from the engagement, um, and you know, as I say, familiarity, consistency [ . . . ] So it’s tapping into those things that make them feel well once they leave. So it’s, I think it’s by getting to know people a bit more really, about who they are and what kind of people they are.
Crisis house staff member
In these general comments about the important characteristics required when relating to service users, no significant differences between crisis house and ward staff were detectable. However, some other related issues discussed below do appear to differentiate between respondents in crisis house and ward environments.
‘Being boundaried and being warm’
There's, kind of, an issue around us with regards to, you know, being boundaried . . . This is an issue we have with regards to people coming in who, they like the place so much, it becomes a [door bangs] big wrench for them if they then leave. So, although it's a very warm environment, we are quite guarded about making it too cosy and it, it's, it can be quite tricky sometimes, but you've really got to try and keep those boundaries up and keep the relationship as professional as possible. You know, you're not trying to become mates with them, you're just trying to be a, kind of, friendly person. And, um, I suppose that's a, that's a challenge in the way of that, you don't, getting that balance right between being boundaried and being warm, I suppose, because too much either way, you're, you're losing something.
Crisis house staff member
Respondents from three crisis houses raised the issue of setting appropriate boundaries in their relationships with service users. Often this was discussed in relation to the challenges of finding a balance between a ‘boundaried’ approach and a warm or human way of being. Similar issues were discussed by ward staff, although less frequently, and they tended to describe these issues in different terms (‘I’m firm but I’m fair’). The need to find an appropriate balance from the beginning of a relationship with a service user is highlighted. These issues show similarities with discussions of ‘professionalism’ combined with human qualities, already identified as key issues in the responses of both staff and service users. The importance of being self-aware or reflexive was highlighted, helping staff to question their own prejudices or assumptions, or to maintain the balance between caring and professionalism:
I guess it’s about being reflective and it’s about always trying, you know . . . being reflective in your practice and checking yourself in terms of how . . . am I being actually therapeutic, in terms of, you know, am I just doing it, I’m going through the motions and I think this is about evaluating yourself, am I just going through the motions or am I actually having . . . am I really caring?
Acute ward staff member
Consent not coercion
A theme that appeared in interviews with crisis house staff was the value of service users’ consent to engage in treatment. This relates to the fact that crisis houses do not take people detained under the Mental Health Act, as well as their policies of allowing self-referrals. The lack of service users detained under the Mental Health Act was seen as important for the development of therapeutic alliances based on mutuality and consent:
The mutual engagement here is, is the only way that’s really effective to work with women. Um, they need to want to be here. We . . . we don’t have any women who are on a section here, for example. Everyone is considered informal, um, and the chances . . . chances for any intervention being effective are much greater if, if we’ve agreed to work together on, on things, when we have an agreement plan, and that’s . . . that’s the basis of our work that we do together.
Crisis house staff member
Witnessing poor practice by colleagues
A small number of comments were made about the working practices of colleagues. Only one of these came from a crisis house worker: ‘they’re not burnt out; they seem to be very focused and enjoy their work which is positive’ (staff member crisis house). In comparison, comments about colleagues were made by three ward staff (all from different wards) and these were all negative. These comments showed parallels with the views of inpatient ward service users discussed above (see A vocation, not just a job). Some colleagues were identified as lacking empathy or patience, not wanting to engage with service users or not sharing workloads with colleagues. One person linked this to both ward culture and to long-serving staff:
There's a culture in the wards where they . . . it's a bit of a separation, you know, you're the [unclear] nurse. And I don't think there's much, unless they have to, but I find, um, how do you, how do you comment on a patient's mental state if you're not interacting much with them. How do you update the system if you've not observed what's happened [ . . . ]. So yes, I think a lot of staff here have probably got to that stage where they've been here too long and they're just doing the job just to have a job rather than doing the job because they enjoy the relationships they have with the patients and the actual what the job entails, patient care.
Acute ward staff member
Another person highlighted situations in which colleagues ignoring service users’ requests led to the escalation of a conflictual situation:
It makes you feel bad. If you see a patient who has got a general concern, something which can be done and get on with it, you know, the patient is upset, that sort of thing, which, you know, if you correct it then things are just fine. This is the main type of thing. Why can't you just do it, ‘cause, you know, you can just do it and hold [?] things up, rather than it might escalate. If you kind of, uh, they don't pay attention, it might start the patient get angry, we have now aggressive, they become aggressive.
Acute ward staff member
When asked about whether or not he has challenged colleagues about their interactive styles, this person (a newly qualified nurse) replies that hierarchies within the nursing team mean that this is not possible.
Influences on therapeutic relationships
Time and workloads
Staff in both settings highlight the importance of having sufficient time to spend with service users in developing a strong therapeutic alliance.
I think the more time you spend with a service user then the more you have a stronger bond, the more you kind of I guess understand the service user and there is kind of a stronger therapy building relationship.
Crisis house staff member
In hospital wards, staff feel that the heavy workload and limited numbers of staff act as barriers to relationships with service users.
There’s a lot going on. You can feel like, you know, you can’t give everybody everything that they need and so you know, frustration can creep in, you know, with certain service users and you know, if that happens they can, your working relationship with them can break down.
Acute ward staff member
There is pressure to meet the needs of service users, attend to more practical ward duties and complete paperwork. Staff report that this results in stress, which in turn impacts on the therapeutic alliance.
Because I’m not relaxed, I’m not able to kind of give them the time or the attention that I would like to because I’m being pulled other ways.
Acute ward staff member
This contrasts with the experience of staff in crisis houses, who reported that they had more time to spend with service users, thus facilitating therapeutic alliances.
From my experience only working on a ward, I’ve got more time I feel here to develop relationships with service users. I’ve got more time to certainly like have one to ones with them.
Crisis house staff member
Activities
Staff across both settings emphasised the importance of both spending time together during one-to-ones and more informally, by eating together or engaging in general activities. They report that this time was paramount in staff and service users getting to know one another better, leading to service users feeling able to disclose personal details and interact with staff. They also highlighted the importance of ‘general’ chat in order to create a warm, comfortable atmosphere.
We sit and have our meals together, you know, so that's quite a nice, friendly, calm atmosphere as well, you know, and we can talk about things in general, over the dinner table, so that that's quite a . . . It's like a family, really.
Crisis house staff member
Atmosphere and environment
The physical environment was identified as important in facilitating relationships between service users and staff. A distinction was drawn between the calm atmosphere of the crisis house and the hectic one of the ward.
Different clients from . . . on the wards, who've come here, to the Crisis House, and, because of the environment, which is a lot more peaceful, a lot, it's quieter, there aren’t many people around.
Crisis house staff member
A quiet, peaceful environment impacts positively on service users.
It feels safer, I think people feeling that they can relax more.
Crisis house staff member
In contrast, the ward environment is busier with greater numbers of service users who are more distressed, which leads to an atmosphere which is less relaxed.
I suppose being a treatment ward, being, you know, having a lot of people here that have restricted leave they only gradually build up it’s a slightly more, it can be slightly more claustrophobic feeling for certain people. So that probably impacts on, you know, the culture on the ward, you know, is partly contained, and you know, you can have people that are very unwell as well, you know. You know, I think apart from that though we do I think try to make it as relaxed as possible. You will be very clear about, you know, the fact that, you know, people in threes they have to be contained, you know, and build up to escorted or unescorted leave.
Acute ward staff member
Freedom and autonomy
Staff report that the freedom service users have within their environment also impacts on the therapeutic alliance.
I think people probably feel a bit happier over the fact that they’ve got that freedom and maybe it avoids some tensions.
Crisis house staff member
Again, there was a contrast between crisis houses, where service users have more freedom to come and go as they please, and wards, where service users are more restricted. This links with the mental health of service users; in a crisis house environment, service users are less distressed than in a ward setting. Staff highlight how the level of crisis experienced by service users has a marked impact on their ability to form relationships, and how relationships change when the crisis changes.
Once they've worked through that, then there should be gradual improvement, and with that, I guess the, the relationship improves as well, because, just the general, general state of mind that individual is, it may be more conducive to forming stronger relationships and that rapport with staff.
Crisis house staff member
Differences in the freedom of service users were coupled with differences in the extent to which autonomy could be promoted. Participants described trying to strike a balance between promoting a sense of responsibility and ensuring safety. They identified advantages from the service user perspective of having responsibility removed, particularly when very distressed. However, they also felt that lack of service user autonomy could have a negative impact on relationships.
Well no-one likes, kind of, being preached to, do they? You know, people don't really like that. And, um, you know, I think they're just, that kind of approach, um, forges an atmosphere of, kind of, almost, resentment, and increases the likelihood, I suppose, of, well, it just increases that, um, patient/expert dichotomy. And, I know in certain cases that can be very useful, and when people are at very low ebb and they, they need, kind of, that, that kind, that feeling that someone's taken the responsibility of them for a period, just for that short period, because they came, you know, can cope with more.
Acute ward staff member
Individual and team morale
Staff talked about the importance of a good team, reporting that this leads to more confidence that difficult situations can be dealt with and higher morale, which impacts positively on relationships with service users.
Because we have a team approach, um, and, and because we work so closely together, um, hopefully our residents feel held by the expectation that things can get better and that we believe in them as, as individuals.
Crisis house staff member
Uniformity and coherence of approach within the staff team were described in both settings as important both for staff morale and for service users to feel contained. Situations where staff are inconsistent often result in service users becoming stressed and ‘kicking off’ thus impacting on the therapeutic alliance.
I think there’s something about having consistency in the staff here that, ah, makes the project feel like a safe place for women to use.
Crisis house staff member
Staff discussed the importance of good support, both from management and colleagues. They emphasised the way in which feeling supported and encouraged maintained their motivation to communicate with service users and meet the demands of the role.
It’s a very demanding job and I think that if you don’t have good support and feel, you know, you have to have stamina in this job, I think, and if you don’t feel supported, um, it can definitely have an impact on, you know, what you’re willing to give, how much patience you might have for people in their presentations.
Acute ward staff member
Changes in funding and cutbacks have resulted in staff having less time to spend with service users which impacts on the therapeutic alliance.
I think it’s less, kind of, clinical time with women, um, and less time to do, kind of, case studies with one another, you know, in the team. Um, so I think that’s . . . that’s where the negative, um, impact on relationships, ah, has come about.
Crisis house staff member
Furthermore, the quality of service available has been affected, due to less training opportunities for staff.
Not being able to do training in this last financial year has . . . has had a massive impact on, on our morale.
Crisis house staff member
One staff member working in a crisis house voiced concerns that, as cutbacks result in the closure of wards, crisis houses will have an intake of service users who are more distressed, taking up more resources and making staff less available to give sufficient time to support service users.
Pushing the boundaries a little bit with regards to the people that we accept, just because it's something we may have to do. But I'm concerned that, that, that'll be just pushed too far, because I think when you do that, then you lose that therapeutic environment.
Crisis house staff member
The survival strategies that staff employ
Survival strategies to manage difficulties in work situations were mentioned by a small number of staff (four ward staff and one crisis house respondent). These acknowledged the often emotionally demanding nature of the work in strategies such as self-reflection about one’s emotional responses in interactions with service users, taking breaks away from the ward during a shift to refresh oneself, and trying not to take things too personally.
What I found has got me through the 16 years of my career is not to get too attached, so just to keep everything professional – have empathy, but keep it very professional. And when I leave work, I leave my work behind me, and then I come the next day with the approach to pay attention, do the best I can in work.
Acute ward staff member
One ward staff member described how when service users were violent or abusive there was a need for additional boundaries which could then limit the amount of therapeutic interaction that is possible. Finally, one staff member on a ward described interactional difficulties and strategies to survive these entirely in terms of his relationships with other members of the ward staff team. Within what is described as an incohesive and hierarchical team, this respondent employed strategies of trying not to make enemies, not taking sides, and ignoring other members of staff (newly qualified nurse acute ward).
Staff recommendations for improving therapeutic relationships in acute care
Staff responses to queries about what residential crisis services could do to improve therapeutic relationships were varied with no very consistent themes detected. Some were about making better use of positive practical strategies:
I’m a big fan of using computers I think, you know, well, we can use many ways to sort of support people, but it’s such an important thing to use these days in terms of a tiny thing getting, you know, one of our patients about two hours ago some information on the natural history museum, travel, etc. It’s little things like that can help to, you know, build or maintain a good relationship and it isn’t separate from their health, you know, it’s doing something like that is really important for somebody, to getting really good information on medication.
Acute ward staff member
Other comments were about avoiding negatives, by ‘not just pay(ing) lip service to user feedback’ (staff member crisis house) and taking complaints seriously so that ‘people know that should there be an incident, that it’s not going to be swept under the carpet’ (staff member acute ward). One ward staff member acknowledged that in the practically oriented ward environment it is rather difficult to train for and ensure that rather intangible personal qualities are enacted – putting in place practical strategies to improve relationships is no guarantee of interpersonal style changes:
How do we cultivate empathy? I think empathy . . . I’m thinking of ways and, I mean, recommendations, how can you . . . how can you motivate someone to be empathetic? That’s . . . that is a difficult . . . yes, that is difficult. [ . . . ] Most of the stuff that we do here at the moment is practical and I understand, you know, most of them in terms of meeting, giving them a welcome pack, those are practical stuff that we do to kind of build the relationships. So yes, so we do . . . we do do those stuff but maybe I’m looking at the difficulties or the times when people just actually do it going through the motions.
Acute ward staff member
One staff member on a ward raised the issue of respect for staff by managers, suggesting that this could improve both staff morale and lead to good working practices:
Well, bottom line is treat the staff with . . . which word shall I use . . . with recognition, and they will recognise that they need to treat the clients with recognition. If they’re given recognition, then they’ll pass it on. Yes, that’s the bottom line.
Acute ward staff member
In many responses it was unclear exactly how the factors mentioned might improve therapeutic relationships or whether or not respondents were raising more general areas for improvement. These include comments about reducing paperwork; working flexibly; increasing clarity about what the service is able to offer; better signposting to other services; and increased staffing levels. Perhaps not surprisingly, compared with service users’ recommendations, staff appear to place greater weight on the organisational context. Interestingly, there is no mention by staff of strategies to enable more time or shared activities with service users. These were two priorities identified by service users as recommendations for improving therapeutic relationships.
Chapter 5 Discussion and conclusions
Main findings
How do therapeutic relationships differ between crisis house and hospital, and how may this contribute to service user satisfaction?
The primary hypothesis to be tested in the study was that an important difference between the ward and crisis house environments may be in the strength of therapeutic alliances between staff and service users in the two settings. Our quantitative study confirmed that a large and statistically significant difference favouring crisis houses is indeed present, with the finding persisting strongly with adjustment for various potential demographic and clinical confounders. This is present even though most staff in crisis houses are not qualified mental health professionals and despite the shorter length of stay in crisis houses (length of stay was not in fact associated with therapeutic relationships or with satisfaction).
Previous research, including the earlier TAS, in which several of the study team participated, has demonstrated markedly greater service user satisfaction in crisis houses than in acute wards,29,35 confirmed again in the present study. However, these investigations have not yielded quantitative data allowing a potential mechanism for this greater satisfaction to be confidently identified; for example, in TAS, content of care and time spent on service user–staff contacts were very similar between crisis houses and hospital. 36 The current study does yield evidence on potential mechanisms for greater satisfaction: better therapeutic relationships are strongly associated with greater service user satisfaction, a parallel with previous findings in physical health care. 53 Furthermore, in a model containing therapeutic relationships, quality of peer relationships and the extent to which staff-related negative events have been experienced, the difference between crisis house and hospital in satisfaction is no longer significant, suggesting that differences in these variables can account for a considerable degree for the difference in satisfaction between the two settings.
Crisis houses have recently been strongly advocated, especially by service users, as meeting needs for residential crisis care in a way that is more acceptable than hospital care. 9 Our findings further affirm the distinctive nature of service user experiences in these services, resulting in very much more positive reports than acute hospital admission. Differences in clinical populations admitted, routes to care and scope for selecting residents who are likely to benefit from the service (see below) must be borne in mind in interpreting these findings. Nonetheless, given the current policy emphasis in the UK on improving service user experiences,54 our findings provide some support for crisis houses as a valuable component in local acute care pathways.
An emerging model of service user satisfaction with residential crisis care
As well as providing a basis for understanding differences between service users experiences of crisis houses and inpatient wards, our findings potentially contribute towards a more general model for understanding the determinants of service user satisfaction with acute care. Service user satisfaction with mental health services is shaped by individual expectations and preferences as well as the specific circumstances surrounding an assessed area of care. Duggins and Shaw55 have usefully conceptualised these as internal factors (expectations and preferences) and external factors (the actual experience), suggesting that the relationship between experiences and satisfaction is a complex one. In our study, demographic and diagnostic factors and admission history, even Mental Health Act status, appeared to have relatively little impact on service user satisfaction. This suggests that it is likely to be fruitful to seek explanations for varying levels of satisfaction in the specifics of service users’ experiences – that is, the external factors influencing satisfaction ratings – more than in service users’ individual characteristics.
The external influences on satisfaction identified in our study were the quality of relationships with staff and peers, and the experience of negative events relating to staff. These have been summarised in Table 16 , along with the factors and experiences that our qualitative and quantitative components suggest have the greatest mediating influence. These mediating factors are discussed at greater length in the following sections, and include issues relating to the personal qualities and behaviours of staff; the contextual and organisational influences on relationships; and service users’ experiences of and desire for peer support. Table 16 also offers suggestions for potential strategies to enhance service user satisfaction based on our findings (these are described in more depth in Implications). They include recruitment strategies that give credence to personal qualities, interpersonal skills and experiences; creating positive working practices and cultures; ensuring service users’ have the greatest levels of freedom and autonomy, with treatment based on consent; and promoting opportunities for peer support. However, these potential strategies are untested and their value should be the subject of further research (see Future research priorities).
| Determinants of service user satisfaction | Mediating factors and experiences | Potential strategies to enhance service user satisfaction |
|---|---|---|
| Therapeutic alliance | Staff personal qualities and behaviours:
|
|
Contextual and organisational factors:
|
|
|
| Peer support |
|
|
| Staff-related negative events | Most commonly:
|
As for strategies to promote therapeutic alliance, also consider:
|
An interesting finding relating to our model of service user satisfaction with acute care is the lack of association between recovery and satisfaction. The main association we found with recovery was in fact a diagnostic association with less positive recovery attitudes among people diagnosed with depression and personality disorder. Thus differences in the extent to which service users feel they have experienced recovery are not likely to account for differences in satisfaction between ward and crisis houses: a parallel is in the prior TAS where comparisons on a more conventionally clinical definition of recovery, i.e. improvement in symptoms and disability, gave no indication that people recover better in crisis houses.
It is important to note that satisfaction research notoriously elicits high levels of service user satisfaction. 56,57 Indeed, levels of service user satisfaction with mental health services have remained relatively constant, despite progressive mental health policies. 58 It has therefore been suggested that satisfaction is not a good measure of service quality or service user experience. Although some service users might critically evaluate their care, the majority of studies suggest that most service users are very uncritical of it, allowing care to be of extremely poor quality before expressing dissatisfaction. 59 Rather than undermining the emerging model of service user satisfaction, this limitation of satisfaction research suggests that our finding of neutral satisfaction on acute wards should not necessarily be taken at face value.
Staff personal qualities and behaviours and therapeutic relationships
Our qualitative results provide a basis for considering strong therapeutic alliances and how these marked differences between hospital and crisis houses may arise. Several potential determinants of good therapeutic relationships were identified, of which the most prominent were the personal qualities of staff and the way they deployed these in their everyday interactions with service users.
Personal qualities
Findings chime with a number of other studies regarding aspects of ward care that service users value highly20,21,60 in the prominence given to personal qualities such as kindness, warmth, interest, humour, respect and empathy. The detailed accounts that we elicited of valued contacts between staff and service users tended to involve demonstrations of straightforward warmth and interest, with quite simple interactions often valued remarkably highly when they conveyed to service users a sense of staff compassion for them and humanity. Both in ward and crisis house, these priorities were in fact congruent with staff views: they also saw basic human qualities and the ability to convey these as at the core of strong therapeutic relationships. There were not many staff with disclosed personal experience of using mental health services in any of the services we investigated, so we were not able to investigate relationships with them in any depth; however, there were some indications that therapeutic alliances with them might be characterised by a deepened trust and sense of empathy and connection.
The qualities that are valued above all are aspects of personality, which might be seen as inherent in staff members rather than influenced by training or professional experience. However, accounts both from service users and staff also clearly indicated that particular skills are required for staff to deploy personal resources of flexibility, kindness, warmth and empathy effectively, while also maintaining more specifically professional skills (e.g. meticulous attention to the confidentiality of all, an ability to respond calmly to high levels of disturbance, to take control and maintain a safe environment when needed, and to guard against confusion by maintaining boundaries where this is appropriate). The optimal therapeutic attitude might be described as ‘professionalism plus’: the combination of professional skills and attributes with strong interpersonal skills that convey a sense of valuing the person as an individual, being warm and caring, and taking the time to listen and understand a person’s experiences. This sometimes brings challenges for staff in terms of maintaining an appropriate balance between two ways of being in their relationships with service users.
Thus our findings are very much in keeping with recent discussions of the need to place compassion at the centre of mental health care,61 and with the concerns underlying the recently launched Compassion in Nursing drive,62 advocating for the centrality of courage and compassion in NHS health care. Central to both our findings and recent initiatives on compassion is the idea of health care as a vocation, not just a job: the key requirement is for staff to be able and willing to give of themselves rather than, as was the perception of some service users regarding ward staff, to go through the motions of working tasks in order to collect a salary.
Negative behaviours and events
As well as styles of interaction that were highly valued, our findings – especially regarding the ward – identified some staff behaviours that were actively aversive and negative. In crisis houses, some staff were negatively described as ‘aloof’ or ‘standoffish’, whereas participants in ward contexts went further than this and described some staff members as ‘actively ignoring’ service users with, in some cases, subsequent escalation in tensions. Lack of interest and engagement seemed to be the most prevalent and important negative behaviour. The toxic effects of ignoring service users were described not only by service users but also by some staff who were unhappy with some of their colleagues’ styles of interaction, so that this seems a very salient and important negative behaviour, confirmed by the fact that in the quantitative component of the study it was the most frequent negative staff-related event.
Going beyond this, considerable numbers of negative events were reported, especially on the wards where the median number of service user-related negative experiences was two and of staff-related negative experiences one. Interestingly, the service user-related negative experiences did not appear associated with overall satisfaction once adjustment was made for service user and service characteristics, whereas staff-related negative experiences were strongly associated with satisfaction, although the untested psychometric properties of our measurement instrument need to be noted. The most frequently identified staff-related negative events after being ignored or dismissed were physical restraint and forcible medication: the dilemmas and difficulties in establishing effective therapeutic alliances where freedom is restricted and coercion is prevalent are discussed further below.
Therapeutically oriented interactions
Personal qualities dominated discussions of therapeutic relationships, but some participants also described valuing more goal-directed forms of interaction, in which there was a clear focus on talking through the difficulties that had led to the crisis and developing strategies for overcoming these difficulties and achieving recovery in a planned way. The relative lack of prominence of this may in part have been because the full-scale deployment of distinct psychotherapeutic techniques appeared to be at a relatively low level in all settings. However, in the crisis house, in particular, service user participants seemed to value more ‘therapeutic’ conversations addressing clear goals and tasks and the causes of crisis and breakdown.
Contextual and organisational influences on therapeutic relationships
The factors in therapeutic relationships discussed so far are largely at an individual staff level – their prominence is a reflection of the observation that in both crisis house and ward contexts service users’ experiences of therapeutic alliances appeared to vary greatly between staff members, shaped by the personalities and approaches of particular professionals (and presumably also by the personal attributes of service users, although we did not explore in detail how these influenced interactions). Thus although there was a general tendency for therapeutic relationships to be found more satisfactory in crisis houses than hospital, there was also substantial evidence of strong relationships with particular professionals being highly valued in a hospital context. However, we also found evidence of factors at a service level, related to service environment, organisation and processes that appeared influential.
Freedom and autonomy
Particularly striking in the narratives of service users was the high impact on them of restrictions on their freedom and of lack of autonomy on the wards. It was clear that ward staff face major obstacles to establishing effective therapeutic relationships where they are also required to restrict a variety of freedoms, where much of service users’ autonomy has been removed and where compulsory treatment results in coercive actions such as forcible medication and physical restraint. A major threat to the authenticity of relationships on wards appears to be service users’ need to try to convince staff that they are ready for more freedom, for example in the form of leave, even if this involves concealing significant distress. The effects of deprivation of liberty were less prominent in staff than in service user accounts, suggesting that they may not always understand how important their effect is on relationships.
The effects of lack of freedom is one respect in which the odds are stacked against hospital staff in being able to form therapeutic relationships as good as in crisis houses: acute wards necessarily have to accept all local referrals within their criteria whether or not service users are willingly admitted, whereas both crisis house service users and staff have some choice when decisions are made about who is admitted to these services. Moreover, the current dominant model of psychiatric hospital provision and treatment tends to focus on risk management and aversion, and typically sees the deprivation of liberty and use of coercive measures as essential components of care and treatment. Thus the obligation of ward staff to impose restrictions on service users’ liberty is an inevitable impediment to the formation of therapeutic relationships.
Despite the prominence of freedom in accounts of why better therapeutic relationships were formed in hospital, service users who had been detained under the Mental Health Act did not appear to be less satisfied than other ward service users in the multiply adjusted quantitative model of service user satisfaction. Thus it may well be, as some participants suggested, that restrictions of freedom have an important impact on relationships for voluntary as well as compulsorily detained service users. Indeed, it is possible that those admitted voluntarily might feel greater resentment regarding restrictions on their freedom when these do not emanate from a clear legal process of detention.
Time to care
A key restriction reported by hospital staff in particular on their capacity to form relationships was the many demands on their time, a difficulty often reported in previous investigations of care on acute wards. 20 Service users were also often quite aware of limits on what staff could offer that resulted from low staffing levels and high demands on wards, whereas crisis house staff were perceived as able to spend more time with service users. Lack of time and competing demands are certainly potentially important impediments to establishing good therapeutic relationships; however, this should be seen in the context of the finding of the preceding TAS that, despite a perception that crisis house staff spent more time with service users, face-to-face contact time was in fact very similar between inpatient wards and alternatives. 36
Environment and atmosphere
Our findings also cohere with many reports on the difficult conditions experienced on acute wards. One of the most recent of many resounding indictments comes from the Schizophrenia Commission:10
In our evidence we heard of many acute units which were stressful, chaotic and scary places. No one seemed to be in charge. Violence, theft and sexual harassment against staff and patients, boredom, poor environments, lack of activity or staff-patient engagement were highlighted as criticisms. Un-therapeutic services, characterised by a sense of hopelessness, staff who do not engage with patients, together with bleak décor and furnishings, can lead to people reacting badly to their hospitalisation. Indeed, when people later relapse, the memory of being in hospital leads to people refusing admission and being ‘sectioned’.
This has many echoes both in our qualitative findings and in the reports of negative events, where theft of property was a particularly frequent type of incident on the wards and reports of physical and sexual assault by staff alarmingly common. Although many initiatives have been put in place to improve the inpatient environment in the past decade, a pressure that is likely to have worked against these is the falling number of inpatient beds and increasing tendency for those who are less distressed and who want the support of services to be managed in the community. Although this has obvious benefits for those offered a community alternative, an unintended effect has been the concentration in hospital of those who are most distressed and who do not want service interventions or treatment; this can pose a challenge both to staff and to fellow service users. 63 The scope for negative cycles to develop where overwhelmed and burnt out staff shut themselves off from service users, provoking all the more disturbance, was all too apparent to several of the study participants.
Shared activities
A further area in which service user and staff reports agreed was on the importance of shared activities, and indeed also shared spaces. Eating together emerged as one simple and significant example of the shared activities which created opportunities for the sort of warm and informal interactions that were highly valued, perhaps not surprisingly given the global universality of shared meals as a way of creating bonds within a wide variety of types of relationships. In crisis houses, both physical environment, which was characterised by more shared spaces such as kitchens, and daily routines seemed to create more opportunities for staff and service users to be together informally than in hospital.
Peer support
A significant secondary finding from both quantitative and qualitative components of the study concerned the importance for service user experiences of good peer relationships. Expectations of such relationships seemed to vary, with some service users hoping only for tranquillity and the absence of conflict, whereas others valued highly the mutual understanding possible between people in the same situation. This again appeared to be fostered more effectively in the crisis house than in hospital, although it should again be noted that crisis houses are able to exclude particular service users, for example because they have histories of violent behaviour.
Methodological strengths and limitations
Strengths
Strengths of the study include recruitment of suitable numbers for the study questions to be addressed, the inclusion of several services of each type, and the clear grounding of study questions in previous research on this topic. The use of mixed methods is a further strength. The qualitative data allowed for a richer interpretation of the quantitative findings; the quantitative findings ensured that qualitative exploration is founded on a clear empirical understanding, for example of the inter-relationship of therapeutic relationships and service user satisfaction.
In the quantitative study, a good response rate was achieved with a sample likely to be representative of service users in the relevant services. There is little missing data, and we used well-validated instruments to examine all relevant concepts except for negative events, where a checklist needed to be constructed specifically for this and the sister study, the PET study, due to lack of an appropriate instrument.
In the qualitative study, the goals of purposive sampling were largely achieved, resulting in good representativeness, and our impression was that thematic saturation had been reached. For the conduct and analysis of the qualitative study, a further strength was the inclusion in the team of researchers who engaged with the topic from multiple standpoints. The two lead authors of this report are mixed-methods service user researchers and the senior author is a mixed-methods health services researcher with a clinical background in psychiatry. Others in the team included a further service user researcher, two experts in qualitative methodologies, and researchers with clinical backgrounds in psychiatry, psychology, nursing and social work. Conducting health services research from within multidisciplinary teams is increasingly being seen as essential in order that multiple standpoints can influence the direction and design of the research and the interpretation of data. 44 A key strength of our research is that we utilised these multiple perspectives when analysing and interpreting the qualitative data set. This can be seen as a form of multiple coding and we believe it enhances the validity of our findings. Furthermore, the inclusion of service user researchers as the key study researchers, engaged not only in data collection, as is frequently the case, but also in study design and instrument development, data analysis and interpretation and writing up, is a particular strength. This is likely to have enhanced both the validity of the data collected (as described above) and the interpretation of that data, and particularly of service user accounts of their experiences.
Limitations
Although the study goes beyond single site research, all participating services were in inner London, an area well known for its demographic distinctiveness, and only two mental health trusts were involved. The crisis houses included several models and funding sources, but none was service user-led and none had a very high degree of service user involvement in service management and delivery: such services may well be different from those we investigated and deserve investigation. Comparisons have been made throughout between crisis houses and wards, but these must all be interpreted with an awareness that the populations admitted to these settings will differ in many respects, especially as the hospital service users in our study were all people who had been admitted to an acute ward despite the local availability of crisis houses. Some potential confounding differences, such as demographic and diagnostic differences were measured, but others remained unmeasured.
It is important to acknowledge that the crisis house and acute ward groups were very different in a number of respects, particularly with regards length of stay, detention under the Mental Health Act,47 ethnicity and diagnosis. Although our multivariable approach to the analysis attempts to adjust for these differences between the groups, it is unlikely to be able to fully account for them. Therefore, it is important to recognise the differences in the profile of people using each service. Nonetheless, service user demographics did not appear to have a strong influence on satisfaction, and it appears unlikely that they are the main factor explaining the marked difference in satisfaction. Although it may not be feasible to reproduce in a hospital setting all features of crisis houses that are associated with greater satisfaction, it remains very likely that some of them are potentially important in considering how service user experiences in acute care settings may be improved.
Our sampling methods – that is, inviting all eligible service users to participate in the quantitative study component and purposively sampling staff and service users for the qualitative component – meant that participants were similar in demographic characteristics and service history to those who routinely use these particular services. However, the sample is not fully representative of all service users, or even of service users of these services, because we only spoke to those who met the inclusion criteria and who were willing to be interviewed. In addition, for the qualitative component of the study staff acted as gatekeepers, having a decisive role in selecting participants who met the inclusion criteria and purposive sampling framework. It is possible that this introduced selection bias, with staff identifying participants they knew to be satisfied with their care or unlikely to report critically.
Regarding objectivity in data collection, there is evidence that service users are more willing to give critical accounts of services when interviewed in a neutral setting and when a period of time has elapsed since the service was experienced,64 rather than, as in our study, while still resident at the service. This suggests that although retrospective interviewing can be criticised for recall bias, it may also be the case that having some distance – both temporal and spatial – between one’s experiences and one’s evaluation of those experiences is necessary to enable critical reflection. This is particularly true when service users are giving accounts of staff who have great power over them, who they are dependent on for daily care and who they may feel, to a certain extent, at the mercy of. Additionally, it may also be the case that people are simply trying to survive their crises, and that retrospective interviewing would mean that the crisis had passed and so people could engage fully with the interview process. The former effect may have been somewhat offset by predominantly using service user researchers as interviewers, as there is evidence that service users give more candid responses to people they regard as peers. 65 However, in a small number of interviews, participants gave the service user researcher interviewer the impression that they were simply going through the motions of the interview and felt unable to give a critical account of relationships. It is hoped that having a substantial data set goes some way towards addressing this as many participants did talk openly and at length about their experiences. However, readers should be mindful of the limitations of interviewing participants while resident within the service they are appraising.
Regarding the analysis of the quantitative study, we have conceptualised better therapeutic relationships as a potential explanatory mechanism for the greater satisfaction experienced in crisis houses. However, it could also be regarded as a component of therapeutic relationships. The potential presence of unmeasured confounders is particularly important for the interpretation of our quantitative findings.
Regarding the qualitative analysis, we have presented here a broad brush overview of the main themes relevant to our research questions that have been extracted on a first team analysis of the data. The qualitative data set is rich and there is scope for much more fine-grained analysis yielding more complex theoretically driven interpretations; such further analyses will be presented in subsequent research papers.
Implications
Implications for service users
The small body of research evidence already available on residential alternatives to hospital suggests that they offer a real alternative to admission at least for some people seeking help in a mental health crisis. Our study has confirmed that service users tend to be more satisfied with crisis houses. We have also found that they enjoy better relationships with staff, experience fewer negative events, are granted greater freedoms and autonomy, and find their environment more homely and therapeutic. A crisis house environment also seems to foster support among peers more than in hospital. This support is very important to some – though not all – service users, enabling people to feel understood and listened to within a mutual helping relationship (often of more significance to service users than the support they receive from staff). Previous research has also found that the quality of contact between staff and service users is prioritised in alternative residential crisis care, with less emphasis placed on medical and forced interventions and more emphasis on support, caring and a destigmatising approach to mental distress. 21 It is perhaps unsurprising then that crisis houses tend to be favoured by service users. However, it should be noted that crisis houses typically exclude those who are detained under the Mental Health Act or people with a substantial history of violence. This means that they are currently only a viable alternative to psychiatric hospital for a limited group.
There has been even less research on service user-led approaches to crisis, including service user crisis houses, than on residential alternatives in general, even though such services are often advocated. We suggest that this knowledge deficit is addressed in order to develop an evidence base for a range of alternatives to psychiatric hospitalisation. This would give service users greater choice over the support they receive when in acute crisis. It may well be that different models of crisis house care fit the needs of different groups of service users.
Finally, although we did not explicitly explore the role of peer support workers, some crisis house service users reported that they felt more understood by peer workers with similar life experiences. This observation finds support in a growing literature that is promoting the value of peer support workers in contemporary mental health services. 66
Implications for practice
Our study has found that the personal qualities of staff – such as warmth, empathy, humour and the ability to listen to and show an interest in others – are crucial in determining the therapeutic relationships that staff and service users are able to form. These personal qualities may not be easily learnt if they are not already present. However, it may not be right to assume that all staff who do not display such qualities in their everyday work are entirely lacking in them, or that the hospital workforce is inherently lacking in the human qualities possessed by crisis house staff. Although some staff may indeed be wrong for the job, other potential reasons for a lack of human response might include a defensive retreat from an environment that is found very stressful and difficult into a rigid and excessively boundaried ‘professionalism’; the difficulties of coping with establishing therapeutic relationships where staff are also obliged to restrict freedoms; and burnout (one of the components of burnout, as generally measured,67 is depersonalisation, which is defined as a loss of the ability to treat recipients of care as fellow human beings rather than objects). Staff may also sometimes simply not understand the importance of the informal interactions with service users and the qualities that are valued in these: this is an area in which training has very often not been received. Addressing these issues should be a clear priority for future action in order to substantially enhance practice and the experiences of service users.
A further issue that we identified as critical in the early formation of therapeutic relationships is the extent to which service users are able to exercise freedom and autonomy. Of particular concern was the finding that service users who are not detained under the Mental Health Act are subject to the same restrictions by default, for example because there are no staff available to open a door or escort people on leave. It seems essential that services ensure that service users have the maximum possible levels of freedom and autonomy. Addressing this would go some way towards removing an early obstacle to forming therapeutic relationships.
Implications for management
Given the importance of qualities of personality in good therapeutic relationships, it is highly desirable to recruit staff whose personalities are right for the job: some individuals were perceived by service users as basically wrong for the job. Strategies such as including service users on interview panels or even formally testing personality and attitudes may help recruit appropriate people, although when mental health professionals are being recruited, the pool of qualified applicants is often rather limited. Of note, the crisis houses employed people from a wider range of backgrounds than the hospitals, raising the possibility that a recruitment strategy based more on personal qualities and experiences than on qualifications and professional seniority might work better for finding the right people [although the dangers of de-skilling hospital workforces that may already struggle to deliver a full range of National Institute for Health and Care Excellence (NICE)-recommended guidelines should be noted]. As some participants suggested, and as has been widely discussed in the NHS recently,66 peer support workers may be helpful in enhancing the capacity of the workforce to respond empathically to service users.
Our qualitative interviews with staff suggested that heavy workloads, limited staff numbers and funding cuts – particularly on acute wards – are limiting contact time between staff and service users, and that this is damaging therapeutic relationships. Staff also highlighted the critical role of team morale in enabling them to feel supported, confident, motivated and able to meet the many demands of the role. A small number of staff described the difficulties caused by witnessing poor practice by colleagues (with their descriptions of poor practice notably mirroring those of service users). Managers should consider addressing these issues, for example by creating a workplace culture and environment that enables staff to support one another, but also to challenge instances of poor practice.
Implications for service planning and policy
Although approximately a third of the total adult mental health-care budget is spent on acute services and mental health policy has in the past mandated crisis resolution teams as a component of local catchment area services, crisis residential care has featured surprisingly little in policy. Furthermore, very little is known about the constituents of effective acute care or the aims of hospitalisation beyond risk management. Although clinical practice guidelines and codes of ethics typically contain statements regarding the values of acute crisis care provision, clinical staff receive very little formal training or guidance to support values-based practices. This lack of policy and clear guidance occurs against the backdrop of consistent criticism of the practices and procedures of psychiatric hospitals, with findings of poor or neutral satisfaction and negative experiences among service users all too common. 8–10 At the same time, a research-derived evidence base for crisis houses is developing, which our study has contributed to. This provides us with some confidence in crisis house models, as well as an emerging explanation for why they are more acceptable to service users. Thus the expansion of community residential crisis house provision, especially in the many areas where none is available, might provide service users with greater choice about the support they receive in a crisis. It might also result in a more flexible and diverse local acute care system that may be more likely to deliver care that fits individual needs. Specific crisis houses tailored to the needs of particular demographic or diagnostic groups, such as women, members of particular ethnic minorities, or people with early psychosis, have been reported in some catchment areas,27 and may further enhance the capacity of a catchment area acute mental health care system to deliver care tailored to the needs of individual service users.
Following our discussion above, we further suggest that service planners consider the support and training available to staff, especially that which focuses on the ways in which they engage and interact with service users; the skill mix of staff, giving greater weight to personal qualities and experiences; ways to support the NHS-wide Compassion in Nursing drive and to extend it across professional groups; and the development of recruitment strategies that enable the right staff to be employed.
Future research priorities
Several further research questions are suggested by these findings.
-
What is the nature and purpose of social interaction in residential crisis facilities, and how does this differ between crisis houses and acute wards? Our findings suggest that although service users want to engage in therapeutic and goal-orientated talk with compassionate and understanding staff, such interaction is often lacking, particularly on acute wards. To inform the development of strategies to enhance interaction, we recommend that research is conducted to explore the nature, processes and purposes of social interaction between service users and staff in residential crisis settings (including hospital wards) and to critically examine the social and contextual factors mediating such relationships. This should yield important insights into how service users and staff seek out, engage in and avoid interaction, and what this reveals about the nature and purposes of contact, social interaction and therapeutic relationships in acute settings. Such research should be undertaken collaboratively between service user and clinical researchers. We recommend that a grounded theory approach is adopted in order to generate a model or theory to account for the processes of social interaction in acute residential settings that is grounded in service user and staff experiences.
-
Why are acute wards vulnerable to a lack of compassion and humanity? Our research joins a growing body of evidence that has identified a lack of compassion and humanity on acute wards and staff disengagement from service users. One of the most recent reports comes from the Schizophrenia Commission which found that the greatest criticisms of the values and practices of staff were directed at those who work in acute care settings. 10 We do not conclude that acute ward staff are inherently unable to demonstrate warmth and compassion, or that the therapeutic alliances that service users and staff form on wards are never characterised by these qualities. Rather, it appears that the culture and practices of acute wards mean that reports of poor relationships between service users and staff – and even of physical and sexual violence perpetrated by staff on service users – are common. Consequently, before research into strategies to enhance ward cultures and therapeutic alliances is conducted, we recommend that in-depth qualitative research is undertaken to extend understanding gained from the current study of the drivers for this, including the reasons that staff are not deploying their personal resources in interactions with service users. Such research should include the views of service users who may be exposed to a lack of compassion from staff, as well as the views of staff and managers, who are ideally placed to reflect on the factors that enable a lack of compassion and humanity to manifest.
-
Can we recruit, train and support staff so that they are more able to use their personal qualities effectively in relationships with service users? Although research into strategies to improve therapeutic relationships is in its infancy68 and we do not want to pre-empt the findings of future research, our study suggests that it is critical that staff with the right personal qualities are employed – that is qualities of warmth, caring, empathy and so on – and that the culture and context of the workplace enables staff to deploy these personal resources effectively in their relationships with service users. Therefore, we recommend that research into strategies to improve therapeutic relationships between staff and service users explicitly explores the role of training and support for staff, and workplace cultures and contexts. Close collaboration between service users and clinicians is likely to be needed to develop feasible and effective methods of improving relationships. Participatory action research may be an appropriate model to develop and test interventions involving enhanced support and training for staff in the ways in which they engage and interact with service users. A cluster randomised trial design with wards as the unit of randomisation may be a suitable design for a more definitive investigation of the outcomes of such interventions.
-
Do therapeutic relationships differ between different crisis house models, including service user-led crisis houses? The crisis houses studied in this research comprised a range of models, from statutory services staffed in the main by mental health clinicians to voluntary services employing few staff with clinical mental health qualifications. The operating principles of the crisis houses varied accordingly; for example, one crisis house adopted a recovery-based model of mental health which saw service users as facilitators of their own recovery with medical and social support, whereas another, women’s only crisis house adopted a more social and psychological model of mental health and distress which understood women’s mental health in the context of their life experiences, such as childhood sexual abuse. None of the crisis houses included in the study were run by service users, largely because such services remain rare. In order to understand the ways in which the organisation and operating principles of crisis houses affects therapeutic relationships, we recommend that an ethnographic study is conducted, focussing on a range of crisis houses that operate along clear values and principles, preferably including a crisis house led and run by service users. This would enable an in-depth exploration of the ways in which the values and principles that drive the culture of crisis houses influences therapeutic relationships.
-
Can a model be developed of the main determinants of service users’ experiences of and satisfaction with acute care? Our findings go some way towards the further development of a model of the important determinants of service users’ experiences of, and satisfaction with, acute care. Given the current emphasis in the NHS on improving service users’ experiences and satisfaction, future research should use mixed methods to develop and test a model that more fully pinpoints the aspects of acute experiences that are the most important influences on service user satisfaction. We recommend that such research is conducted with or by service user researchers.
Further to this, our work suggests a number of areas that are somewhat beyond the immediate scope of this report where developing and testing service improvement strategies is desirable: these include enhancing opportunities for peer support, reducing adverse events on the ward,69 and enabling staff to spend more time with service users, either through higher staffing levels or reorganisation of their working practices. Strategies based on any of these would have some support from our findings as potential ways of enhancing service user experiences in acute settings.
In addition, service users have a long-standing interest in developing their/our own responses to mental health crises, including establishing residential alternatives to psychiatric hospitalisation. These responses are typically focussed on enabling the person experiencing mental distress to retain choice and control and tend to reject the deprivation of liberty; our study suggests that removing freedoms may impede the development of therapeutic relationships. As relatively little is known about service user-led responses to crises, initial research should be undertaken to map the extent, diversity and values of service user-led responses to acute mental distress and to identify the ways in which these differ from standard acute psychiatric care. It will then be possible to conduct further research into service user-led alternatives to mental health crises, including mixed-methods research to compare therapeutic alliances and outcomes in service user-led and mainstream acute services.
Acknowledgements
We are very grateful to the service users and staff at all the participating centres who generously gave their time and shared their views and experiences to allow the study to take place. Some of the comparison data used in the present study is drawn from the PET study (Chief Investigator Fiona Nolan): we are grateful to the PET study team, especially the research workers Charlotte Kirton and Abdul Choudhury, for contributing data from their study. Joanne Taylor and Clementina Galli Zugaro provided invaluable help by contributing to the draft introduction and preparing the final data set for analysis. Mike Slade contributed to the conception and design of the study. Mental Health Research Network (MHRN) Clinical Studies Officers Naomi Bateman, Kirsty Collins and Raphael Underwood assisted with quantitative data collection.
Department of Health Disclaimer: The views and opinions expressed therein are those of the authors and do not necessarily reflect those of the Health Services and Delivery Research (HS&DR) Programme, NIHR, NHS or the Department of Health.
Contributions of authors
Angela Sweeney contributed to the study design once the study was underway, collected study data, carried out and coordinated the qualitative data analysis, drafted sections of the final report and coordinated its writing.
Sarah Fahmy contributed to the study design once the study was underway, collected study data, contributed to the qualitative data analysis and drafted sections of the final report.
Fiona Nolan designed and let the PET study, from which data was obtained to use in the current study.
Nicola Morant contributed to the conceptualisation and design of the study, oversaw it as part of the study management group, contributed to the qualitative data analysis, drafted sections of the final report and was consulted throughout for qualitative advice.
Zoe Fox was the study statistician and analysed the quantitative study data.
Brynmor Lloyd Evans contributed to the conceptualisation and design of the study, oversaw it as part of the study management group and drafted sections of the final report.
David Osborn contributed to the conceptualisation and design of the study, and oversaw it as part of the study management group.
Emma Burgess contributed to the qualitative data analysis and drafted sections of the final report.
Helen Gilburt contributed to the conceptualisation and design of the study, and oversaw it as part of the study management group.
Rosemarie McCabe contributed to the conceptualisation and design of the study, oversaw it as part of the study management group and was consulted throughout for qualitative advice.
Sonia Johnson contributed to the conceptualisation and design of the study, was responsible for overall study management and supervision, drafted sections of the final report and carried out the final editing process.
All authors contributed to interpreting and presenting data in the final report.
Sweeney A, Fahmy S, Nolan F, Morant N, Fox Z, Lloyd-Evans B, et al. The Relationship Between Therapeutic Alliance and Service User Satisfaction in Mental Health Inpatient Wards and Crisis House Alternatives: A Cross-Sectional Study. PLOS ONE 2014; July. doi:10.1371/journal.pone.0100153
Disclaimers
This report presents independent research funded by the National Institute for Health Research (NIHR). The views and opinions expressed by authors in this publication are those of the authors and do not necessarily reflect those of the NHS, the NIHR, NETSCC, the HS&DR programme or the Department of Health. If there are verbatim quotations included in this publication the views and opinions expressed by the interviewees are those of the interviewees and do not necessarily reflect those of the authors, those of the NHS, the NIHR, NETSCC, the HS&DR programme or the Department of Health.
References
- The Pathway to Recovery: A Review of NHS Acute Inpatient Mental Health Services. London: Healthcare Commission; 2008.
- Mental Health Policy Implementation Guide: Adult Acute Inpatient Care Provision. London: DH; 2005.
- Quirk A, Lelliott P. What do we know about life on acute psychiatric wards in the UK? A review of the research evidence. Soc Sci Med 2001;53:1565-74. http://dx.doi.org/10.1016/S0277-9536(00)00457-3.
- Ingham T. Mental Health Strategies for the Department of Health. London: Department of Health; 2012.
- Listening, Learning, Working Together: A National Survey of How Well Healthcare Organisations Engage Local People in Planning and Implementing their Services. London: Commission for Healthcare Audit and Inspection; 2009.
- Rose D. Users’ Voices, the Perspectives of Mental Health Service Users on Community and Hospital Care. London: The Sainsbury Centre; 2001.
- Baker S. Environmentally Friendly? Patients Views of Conditions on Psychiatric Wards. London: Mind; 2000.
- Ward Watch: Mind’s Campaign to Improve Hospital Conditions for Mental Health Patients. London: Mind; 2004.
- Listening to Experience: An Independent Inquiry Into Acute and Crisis Mental Healthcare. London: Mind; 2011.
- The Abandoned Illness: A Report by the Schizophrenia Commission. London: Schizophrenia Commission; 2012.
- Weich S, Griffith L, Commander M, Bradby H, Sashidharan SP, Pemberton S, et al. Experiences of acute mental health in an ethnically diverse inner city: qualitative interview study. Soc Psychiatry Psychiatr Epidemiol 2012;47:119-28. http://dx.doi.org/10.1007/s00127-010-0314-z.
- McGuire R, McCabe R, Priebe S. Theoretical frameworks for understanding and investigating the therapeutic relationship in psychiatry. Soc Psychiatry Psychiatr Epidemiol 2001;36:557-64. http://dx.doi.org/10.1007/s001270170007.
- McCabe R, Priebe S. The therapeutic relationship in psychiatric settings. Acta Psychiatr Scand 2006;113:69-72.
- Catty J. Attachment to the clinical team and its association with therapeutic relationships, social networks, and clinical well-being. Psychol Psychother 2012;85:17-35. http://dx.doi.org/10.1111/j.2044-8341.2010.02011.x.
- McGuire-Snieckus R, McCabe R, Catty J, Hansson L, Priebe S. A new scale to assess the therapeutic relationship in community mental health care: STAR. Psychol Med 2007;37:85-9. http://dx.doi.org/10.1017/S0033291706009299.
- Catty J, Winfield H, Clement S. The therapeutic relationship in secondary mental health care: a conceptual review of measures. Acta Psychiatr Scand 2007;116:238-52. http://dx.doi.org/10.1111/j.1600-0447.2007.01070.x.
- Johansson H, Eklund M. Patient’s opinion on what constitutes good psychiatric care. Scand J Caring Sci 2003;17:339-46. http://dx.doi.org/10.1046/j.0283-9318.2003.00233.x.
- Noble L, Douglas C. What users and relatives want from mental health services. Curr Opin Psychiatry 2004;17:289-96. http://dx.doi.org/10.1097/01.yco.0000133832.42167.76.
- Catty J. ‘The vehicle of success’: theoretical and empirical perspectives on the therapeutic alliance in psychotherapy and psychiatry. Psychol Psychother 2004;77:255-72. http://dx.doi.org/10.1348/147608304323112528.
- Stenhouse RC. ‘They all said you could come and speak to us’: patients’ expectations and experiences of help on an acute psychiatric inpatient ward. J Psychiatr Ment Health Nurs 2011;18:74-80. http://dx.doi.org/10.1111/j.1365-2850.2010.01645.x.
- Cleary M, Hunt GE, Horsfall J, Deacon M. Nurse–patient interaction in acute adult inpatient mental health units: a review and synthesis of qualitative studies. Issues Ment Health Nur 2012;33:66-79. http://dx.doi.org/10.3109/01612840.2011.622428.
- Hummelvoll JK, Severinsson EI. Imperative ideals and the strenuous reality: focusing on acute psychiatry. J Psychiatr Ment Health Nurs 2001;8:17-24. http://dx.doi.org/10.1046/j.1365-2850.2001.00346.x.
- Björkdahl A, Palmstierna T, Hansebo G. The bulldozer and the ballet dancer: aspects of nurses’ caring approaches in acute psychiatric intensive care. J Psychiatr Ment Health Nurs 2010;17:510-18. http://dx.doi.org/10.1111/j.1365-2850.2010.01548.x.
- Hopkins JE, Loeb SJ, Fick DM. Beyond satisfaction, what service users expect of inpatient mental health care: a literature review. J Psychiatr Ment Nurs 2009;16:927-37. http://dx.doi.org/10.1111/j.1365-2850.2009.01501.x.
- Woods A, Springham N. On learning from being the in-patient. Int J Art Ther 2011;16:60-8. http://dx.doi.org/10.1080/17454832.2011.603697.
- Totman J, Hundt GL, Wearn E, Paul M, Johnson S. Factors affecting staff morale on inpatient wards: a qualitative investigation. BMC Psychiatry 2011;11. http://dx.doi.org/10.1186/1471-244X-11-68.
- Johnson S, Gilburt H, Lloyd-Evans B, Osborn DPJ, Boardman J, Leese M, et al. In-patient and residential alternatives to standard acute psychiatric wards in England. Br J Psychiatry 2009;194:456-63. http://dx.doi.org/10.1192/bjp.bp.108.051698.
- Johnson S, Lloyd-Evans B, Morant N, Gilburt H, Shepherd G, Slade M, et al. Alternatives to standard acute inpatient care in England: roles and populations served. Br J Psychiatry 2010;197:S6-13.
- Lloyd-Evans B, Slade M, Jagielska D, Johnson S. Residential alternatives to acute psychiatric hospital admission: systematic review. Br J Psychiatry 2009;195:109-17. http://dx.doi.org/10.1192/bjp.bp.108.058347.
- Johnson S, Bingham C, Billings J, Pilling S, Morant, N, Bebbington P, et al. Women’s experiences of admission to a crisis house and to acute hospital wards: a qualitative study. J Ment Health 2004;13:247-62. http://dx.doi.org/10.1080/09638230410001700880.
- Lichtenberg P. The residential care alternative for the acutely psychotic patient. Psychiatr Q 2011;82:329-41. http://dx.doi.org/10.1007/s11126-011-9176-0.
- Lloyd-Evans B, Johnson S, Slade M, Barrett B, Byford S, Gilburt H, et al. In-Patient Alternatives to Traditional Mental Health In-Patient Care (The Alternatives Study) 2010. www.nets.nihr.ac.uk/__data/assets/pdf_file/0006/64464/FR-08-1304-075.pdf (accessed 2 March 2014).
- Slade M, Byford S, Barrett B, Lloyd-Evans B, Gilburt H, Osborn D, et al. Alternatives to standard acute care in England: short-term clinical outcomes and cost-effectiveness. Br J Psychiatry 2010;197:S14-19. http://dx.doi.org/10.1192/bjp.bp.110.081059.
- Gilburt H, Slade M, Rose D, Lloyd-Evans B, Johnson S, Osborn D. Service users’ experiences of residential alternatives to standard acute wards: a qualitative study of similarities and differences. Br J Psychiatry 2010;197:S26-31. http://dx.doi.org/10.1192/bjp.bp.110.081075.
- Osborn D, Lloyd-Evans B, Johnson S, Gilburt H, Byford S, Leese M, et al. Alternatives to standard acute care in England: a comparison of satisfaction, ward atmosphere and service user experiences. Br J Psychiatry 2010;197:S41-5.
- Lloyd-Evans B, Johnson S, Morant N, Gilburt H, Osborn D, Jagielska D, et al. Alternatives to standard acute care in England: differences in content of care and staff–patient contact. Br J Psychiatry 2010;197:S46-51. http://dx.doi.org/10.1192/bjp.bp.110.081117.
- Byford S, Sharac J, Lloyd-Evans B, Gilburt H, Osborn D, Leese M, et al. Alternatives to standard acute care in England: readmissions, service use and cost after discharge. Br J Psychiatry 2010;197:S20-5. http://dx.doi.org/10.1192/bjp.bp.110.081067.
- Johnson S, Lloyd-Evans B, Howard LM, Osborn DPJ, Slade M. Where next with residential alternatives to admission?. Br J Psychiatry 2010;197:S52-4. http://dx.doi.org/10.1192/bjp.bp.110.081125.
- McCabe R, Priebe S. The therapeutic relationship in the treatment of severe mental illness: a review of methods and findings. Int J Soc Psychiatry 2004;50:115-28. http://dx.doi.org/10.1177/0020764004040959.
- Hansson L, Berglund M. Stability of therapeutic alliance and its relationship to outcome in short-term inpatient psychiatric care. Scand J Soc Med 1992;20:45-50.
- Barker P. The Tidal Model: a person-centred approach to psychiatric and mental health nursing. Perspect Psychiatr Care 2001;37:79-87.
- Tansella M, Thornicroft G. Implementation science: understanding the translation of evidence into practice. Br J Psychiatry 2009;195:283-5. http://dx.doi.org/10.1192/bjp.bp.109.065565.
- Galster GC, Hesser GW. Residential satisfaction: compositional and contextual correlates. Environ Behav 1981;13:735-58. http://dx.doi.org/10.1177/0013916581136006.
- Sweeney A, Greenwood KE, Williams S, Wykes T, Rose DS. Hearing the voices of service user researchers in collaborative qualitative data analysis: the case for multiple coding. Health Expect 2013;16:e89-e99. http://dx.doi.org/10.1111/j.1369-7625.2012.00810.x.
- Kent M. My mental health: protected therapeutic time. Ment Health Prac 2005;8:22-3.
- Simpson A, Brooker C, Repper J. Mental Health: From Policy to Practice. China: Churchill Livingstone Elsevier; 2009.
- Code of Practice, Mental Health Act 1983. London: The Stationary Office; 2008.
- Attkisson C, Zwick R. The Client Satisfaction Questionnaire: psychiatric properties and correlations with service utilisation and psychotherapy outcome. Eval Program Plan 1982;5:233-7.
- Tilden C, Nelson C, May B. The IPR Inventory: development and psychometric characteristics. Nurs Res 1990;39:337-43. http://dx.doi.org/10.1097/00006199-199011000-00004.
- Corrigan P, Salzer M, Ralph R, Sangster Y, Keck L. Examining the factor structure of the Recovery Assessment Scale. Schizophrenia Bull 2004;30:1035-41. http://dx.doi.org/10.1093/oxfordjournals.schbul.a007118.
- Schrank B, Slade M. Recovery in psychiatry. Psychiatr Bull 2007;31:321-5. http://dx.doi.org/10.1192/pb.bp.106.013425.
- Catty J, Cowan N, Poole Z, Ellis G, Geyer C, Lissouba P, et al. Attachment to the clinical team and its association with therapeutic relationships, social networks, and clinical well-being. Psychol Psychother 2012;85:17-35. http://dx.doi.org/10.1111/j.2044-8341.2010.02011.x.
- Kim SC, Kim S, Boren D. The quality of therapeutic alliance between patient and provider predicts general satisfaction. Military Med 2008;173:85-90.
- Kendall T, Crawford MJ, Taylor C, Whittington C, Rose D. Improving the experience of care for adults using NHS mental health services: summary of NICE guidance. BMJ 2012;344. http://dx.doi.org/10.1136/bmj.e1089.
- Duggins R, Shaw I. Examining the concept of patient satisfaction in patients with a diagnosis of schizophrenia: a qualitative study. Psychiatr Bull 2006;30:142-5. http://dx.doi.org/10.1192/pb.30.4.142.
- Blenkiron P, Hammill C. What determines patients’ satisfaction with their mental health care and quality of life?. Postgrad Med J 2003;79:337-40. http://dx.doi.org/10.1136/pmj.79.932.337.
- Clark C, Scott E, Boydell K, Goering P. Effects of client interviewers on client-reported satisfaction with mental health services. Psychiatr Serv 1999;50:961-3.
- Webb Y, Leiper R, Field V. Counting for Something in Mental Health Services: Effective User Feedback. Avebury: Aldershot and Brookfield; 1993.
- William B. Patient satisfaction: a valid concept?. Soc Sci Med 1994;38:509-16.
- Thomas SP, Shattell M, Martin T. What’s therapeutic about the therapeutic milieu?. Arch Psychiatr Nurs 2002;16:99-107.
- Spandler H, Stickley T. No hope without compassion: the importance of compassion in recovery-focused mental health services. J Ment Health 2011;20:555-66. http://dx.doi.org/10.3109/09638237.2011.583949.
- Schofield B. Concept Analysis – Compassion in Nursing n.d. www.rcn.org.uk/__data/assets/pdf_file/0011/445817/Research2012Mo16.pdf (accessed 2 January 2013).
- Tyrer P. Has the closure of psychiatric beds gone too far?. BMJ 2011;343. http://dx.doi.org/10.1136/bmj.d7457.
- Rose D, Fleischmann P, Wykes T, Leese M, Bindman J. Patients’ perspectives on electroconvulsive therapy: systematic review. BMJ 2003;326. http://dx.doi.org/10.1136/bmj.326.7403.1363.
- Polowycz D, Brutas M, Orvietto BS, Vidal J, Cipriana D. Comparison of patient and staff surveys of consumer satisfaction. Hosp Community Psychiatry 1993;44:589-91.
- Faulkner A, Bassett T. A helping hand: taking peer support into the 21st century. Ment Health Soc Incl 2012;16:41-7. http://dx.doi.org/10.1108/20428301211205892.
- Maslach C, Jackson S. The Maslach Burnout Inventory. California, CA: Consulting Psychologists Press; 1981.
- Jungham UM, Leese M, Priebe S, Slade M. Staff and patient perspectives on unmet need and therapeutic alliance in community mental health services. Br J Psychiatry 2007;191:543-7. http://dx.doi.org/10.1192/bjp.bp.107.037978.
- Bowers L, Allan T, Simpson A, Nijman H, Warren H. Adverse incidents, patient flow and nursing variables on acute psychiatric wards: the Tompkins Acute ward study. Int J Soc Psychiatry 2007;53:75-84. http://dx.doi.org/10.1177/0020764007075011.
Appendix 1 Service user demographics form
Appendix 2 Negative events schedule
Appendix 3 Form for collecting additional information from case notes
Appendix 4 Service user demographics form for qualitative study
Appendix 5 Service user interview schedule
Appendix 6 Staff demographics form for qualitative study
Appendix 7 Staff interview schedule
Glossary
- Acute hospital or inpatient care
- Treatment provided in hospital following admission in a mental health crisis.
- Acute wards/standard acute care
- The hospital wards to which most crisis admissions of adults of working age take place/the care routinely provided on them in the absence of a specific programme intended to change or improve this.
- Acute residential care/residential crisis care/acute residential settings
- Terms for any mental health crisis care where service users have at least an overnight stay. It includes standard care (psychiatric hospital) and alternative care (such as crisis houses).
- Cluster sandwich estimate
- A statistical technique for taking into account the lack of independence between individuals within the same subgroup in a sample.
- Coding frame
- A conceptual framework for qualitative analysis that comprises individual and grouped codes (concepts, categories and themes).
- Community mental health teams
- These are teams of mental health professionals based in the community. They have, over the past decades, been the main providers for most people with significant mental health problems who are living in the community and receiving care from NHS mental health services.
- Crisis house
- The most frequent community residential alternative to hospitalisation for people experiencing mental crises. Crisis houses tend to be smaller, more homely and less medical than hospital wards. Unlike hospitals, they do not accept all referrals.
- Crisis residential alternatives/residential alternatives
- A collective term for alternatives to standard acute care, including crisis houses.
- Crisis resolution teams
- Community-based teams who work with people experiencing mental distress or crisis in their own homes who otherwise would be admitted to psychiatric hospital. They usually provide both crisis assessment and short-term intensive treatment at home.
- Mental distress
- A term often favoured by service users to describe deeply distressing mental and emotional experiences.
- Multivariable linear regression
- A technique for modelling the influence that a number of variables have on an outcome.
- Peer support
- The mutual support between service users based on respect and empathy.
- Protected Engagement Time
- An initiative that aims to increase the quantity and quality of interactions between staff and service users on acute wards.
- Psychometric testing
- A test that establishes the properties of a measure, such as reliability (reproducibility of results) and validity (whether or not the instrument really measures what it is supposed to measure).
- Purposive sampling
- A sampling method that selects participants on the basis of their characteristics, such as age, sex and history of service use. Purposive sampling usually aims to ensure that participants reflect the diversity and breadth of the population about whom the investigators hope to reach conclusions.
- p-value
- The probability that a result has occurred by chance or error. If p is > 0.05 it is assumed that any difference found is probably the result of chance only rather than a true statistical finding.
- Qualitative
- Research that generates an in-depth understanding of phenomenon through analysing text and talk (e.g. interviews).
- Quantitative
- Research that uses measurement and numbers to test hypotheses.
- Recovery
- A term that often refers to the journey a person takes to regain control over and meaning in their life. Service users rarely use recovery to mean the absence of symptoms, though this definition is traditionally adopted by clinicians.
- Response rate
- The proportion of people who were eligible to participate in the study and who agreed to take part.
- Service user
- A person who uses mental health services.
- Service user satisfaction
- The extent to which service users are satisfied with the care that they receive, sometimes conceptualised as being influenced by internal factors (expectations and preferences) and external factors (the actual experience).
- Sociodemographics/demographics
- The characteristics of the people who participated in the study, such as their age, sex and ethnicity.
- Staff
- People employed within the mental health system as clinicians or support workers who have direct therapeutic contact with service users. We did not include administrative or domestic staff in the current sample, though they too may have important relationships with service users.
- The Alternatives Study
- The study that preceded the current study with some of the same investigators involved. This investigated residential alternatives to standard acute care. Findings included that service users in crisis houses were more satisfied than service users on acute wards.
- Thematic analysis
- A form of qualitative analysis that seeks to understand data by categorising it into emerging themes or categories.
- Therapeutic alliance
- The affective bonds between service users and clinicians in which the service user feels valued and supported as an individual.
- Therapeutic relationships
- The collaborative bonds between service users and clinicians based on shared goals and positive personal qualities.
List of abbreviations
- CSQ
- Client Satisfaction Questionnaire
- IPRI
- Interpersonal Relationship Inventory
- NIHR
- National Institute for Health Research
- PET
- Protected Engagement Time
- RAS
- Recovery Assessment Scale
- SDO
- Service Delivery and Organisation
- STAR-P
- Scale To Assess the Therapeutic Relationship-Patient
- TAS
- The Alternatives Study
- TAS-2
- The Alternatives Study 2