Exploring the therapeutic relationship through the reflective practice of nurses in acute mental health units: A qualitative study

Abstract Aims and objectives To explore the therapeutic relationship through the reflective practice of nurses in acute mental health units. Background In mental health units, the therapeutic relationship is especially relevant for increasing the effectiveness of nursing interventions. Reflective practice is considered an essential aspect for improving nursing care. Design Action and observation stages of a participatory action research project. Methods Data were collected through reflective diaries designed for the guided description and reflection of practice interactions related to the therapeutic relationship and content analysis was applied. A total of 152 nurses from 18 acute mental health units participated. The COREQ guidelines were used. Results The results were classified into three categories as follows: (i) Nursing attitude as a core of the therapeutic relationship. For the nurses, the attitudinal component was key in the therapeutic relationship. (ii) Nursing practices that are essential to the therapeutic relationship. Nurses identified practices such as creating a conducive environment, using an appropriate verbal approach, offering help and working together with the patient as essential for establishing a therapeutic relationship in practice. (iii) Contextual factors affecting the therapeutic relationship. The nurses considered the patient's condition, the care dynamics of the unit and its regulations, as well as the structure and environment of the unit, as contextual factors involved the establishment of an adequate therapeutic relationship in daily clinical practice. Conclusions This study has provided knowledge of the importance and role of the nurses' attitude in the context of the nurse–patient therapeutic relationship based on the reflections of nurses in mental health units regarding their own practice. Relevance to clinical practice These findings help nurses to increase awareness and develop improvement strategies based on their own knowledge and day‐to‐day difficulties. Moreover, managers can evaluate strategies that promote motivation and facilitate the involvement of nurses to improve the therapeutic relationship with patients.


| INTRODUC TI ON
The nursing discipline is defined as a significant, therapeutic and interpersonal process that acts in conjunction with other human processes that make health possible for individuals (Peplau, 1988).
The relationship established between nurse and patient is therapeutic, regardless of the setting in which care is provided (Stevenson & Taylor, 2020). However, in the mental health unit setting, the therapeutic relationship is especially relevant to increase the effectiveness of any nursing intervention (McAndrew et al., 2014). Reflective practice is considered an essential aspect of improving nursing care and generating knowledge (Vaughan, 2017). This paper aims to deepen the knowledge of the therapeutic relationship based on the reflections of nurses regarding their practice, in the context of current challenges within the mental health acute care setting.

| Background
Based on Peplau's model of interpersonal relationships by (1988), which is the most widely held theory in the mental health nursing community, many authors have based their models on personcentred mental health nursing (Barker & Buchanan-Barker, 2010;O'Brien, 2001;Scanlon, 2006). All of them identify the therapeutic relationship as the foundation of nursing practice and the pillar upon which mental health nursing has been built (McAllister et al., 2019;Moreno-Poyato et al., 2016). The proper establishment of the nurse-patient therapeutic relationship is especially relevant to increase the effectiveness of any nursing intervention in acute psychiatric units (McAndrew et al., 2014).
The therapeutic relationship could be defined as a human exchange (Peplau, 1988) that is based on effective communication that favours the possibility for a person to help another person to improve their health condition, with the objective that, through such communication, the person will be able to develop interpersonal and problem-solving skills (Forchuk et al., 1998). To this end, concepts such as understanding, interest, availability, individuality, authenticity, warmth, respect and self-knowledge are basic pillars for the nurse (Moreno-Poyato et al., 2016). The literature points out that mental health nurses seem to be knowledgeable of the importance of the therapeutic relationship in inpatient units; however, the reality of clinical practice leads us to believe that theoretical knowledge is not enough to create a good bond with patients (Moreno-Poyato et al., 2016). In addition, the literature points out that for nurses, the implementation of the therapeutic relationship in the current context of mental health units has suffered a strong impact related to neoliberal policies, with increased management and a risk-centred approach (Kingston & Greenwood, 2020). Thus, today's environments are chaotic, and nurses are committed to therapeutic work, yet they struggle to balance it with the new demands of management (Kingston & Greenwood, 2020). In addition, barriers such as lack of time, communication problems (Harris & Panozzo, 2019a), the physical structures of the units, the ratios or the cultures of care are external factors that limit the therapeutic relationship (Tolosa-Merlos et al., 2021). If nurses are unable to become aware of how they respond to time pressure, frustration or unclear care policies, there is a risk that these barriers will become entrenched, new ones will be created and the patient will perceive their actions as lacking care, presence or involvement (Harris & Panozzo, 2019b). Thus, although nurses recognise the importance of self-awareness and knowing how to recognise how their actions can impact the therapeutic relationship and the care provided to patients, they are also aware of the need for self-awareness (Thomson et al., 2019), institutions and, in general, care policies should encourage nurses to be Relevance to clinical practice: These findings help nurses to increase awareness and develop improvement strategies based on their own knowledge and day-to-day difficulties. Moreover, managers can evaluate strategies that promote motivation and facilitate the involvement of nurses to improve the therapeutic relationship with patients.

K E Y W O R D S
action research, narrative, nurse-patient relationships, psychiatric nursing, reflective practice What does this paper contribute to the wider global clinical community?
• An in-depth analysis of nurses' reflections regarding the aspects that underlie the therapeutic relationship in their clinical practice enables the nurses themselves to become aware and to develop strategies for improvement based on their own knowledge.
• Understanding and confirming how the attitudinal component is a key element for nurses in the practice of the therapeutic relationship allows managers to evaluate strategies that promote motivation and facilitate the involvement of nurses to improve their practice with patients.
• The results point to the need for further studies aimed at identifying and implementing strategies that facilitate mental health nurses to incorporate and improve attitudinal skills related to establishing the nurse-patient therapeutic relationship in clinical practice.
aware of interpersonal influences, as well as the desirability of providing a safe and supportive clinical environment for these relationships (Stevenson & Taylor, 2020).
From the patients' point of view, in the complex environment of inpatient units, their interactions with staff are central components to their satisfaction regarding their experience with admission (Molin et al., 2021). When staff spend time, engage in daily activities, and recognise patients as individuals, patients seem to find it easier to be physically and emotionally closer to each other and to themselves (Eldal et al., 2019;. However, this therapeutic commitment is not always met in practice, and interventions to improve participation are few and far between and ineffective (McAllister et al., 2021).
Thanks to the therapeutic relationship, nurses are in a key position to lead the development of customised interventions (Molin et al., 2021). However, there is a significant gap in the literature regarding improving the quality of the therapeutic relationship in acute mental health units (Hartley et al., 2020). The nursing profession is characterised by its ability to reflect on practice to improve care and provide more person-centred care, which is why there is a need to increase the use of evidence-based practice (Vaughan, 2017). In fact, reflective practice allows practitioners to learn from their experiences (Bulman & Schutz, 2013;Schön, 1987). When nurses are given time to reflect through guided reflection questions they are able to gain valuable insight into practice (Bolg et al., 2020); therefore, reflective practice helps nurses integrate their emotional response and practical experience into a better understanding of the care they provide, incorporating knowledge and applying theory (Vaughan, 2017). Thus, although the nurse-patient therapeutic relationship has been extensively studied, no studies to date provide knowledge on the establishment of the therapeutic relationship and its implications based on the reflection on the nurses' own practice.
Consequently, knowing the meaning of the therapeutic relationship together with the elements that facilitate and hinder its implementation in the complex practice of current acute mental health units can be a starting point for both nurses and managers to become aware of the needs and for the design of strategies for improvement, suited to the reality of clinical practice.
In this regard, the aim of this study was to explore the phenomenon of the therapeutic relationship through the reflective practice of nurses in acute mental health units.

| Design
This study is part of a multicentre mixed methods study involving 18 acute mental health units in Catalonia (Spain) (MiRTCIME. CAT). The principal aim of the project is to improve the nursepatient therapeutic relationship through the implementation of evidence. The project was carried out following a sequential and transformational design. Quantitative methods were used based on a single-group quasi-experimental design with baseline and follow-up measurements in phases I and III of the project. In the second phase, qualitative methodology was used. In its qualitative component, participatory action research (PAR) was proposed, framed within the constructivist paradigm and following the model by Kemmis and Mctaggart (2008). A two-cycle process consisting of four stages each was designed to carry out the PAR. Specifically, this work corresponds to the action and observation stages of the first cycle. These stages are basic in the PAR process of change and make it possible to generate relevant knowledge regarding habitual practice (Cusack et al., 2018). In fact, it allows nurses to understand their practices as the product of particular circumstances and thus to identify the crucial aspects on which it may be possible to transform the practices they are carrying out (Kemmis & Mctaggart, 2008). The study is reported in line with the Consolidated criteria for reporting qualitative research guidelines (COREQ: Tong et al., 2007) (File S1).

| Participants
All the acute mental health units that were part of the Catalan Mental Health Network (n = 21) were informed of the study. The principal investigator presented the research project and its objectives to the management of each centre through informative sessions. Finally, 18 units agreed to participate. A nurse from each unit joined the research team and this researcher was in charge of coordinating the study at their centre and recruiting the nurses from each unit. All nurses employed in the participating units (n = 235) were invited to participate in the study. The inclusion criteria for the participating nurses were belonging to the permanent or interim staff and being assigned to the acute unit at the time the intervention began. The following nurses were excluded from the study: nurses who were training to obtain 'the official qualification of mental health nurse', staff nurses who were scheduled to be on leave or maternity leave during the intervention. Thus, a convenience sample of 195 nurses agreed to participate in PAR, of which, ultimately 152 nurses completed the action and observation stages of the first part of this study.

| Data collection
During a previous meeting among the entire research team, a guide was agreed upon so that the nurses could self-observe their clinical practice in relation to the establishment of the therapeutic relationship. The research team sent the self-observation guide by email to each nurse, along with a reflective diary in which the nurses were asked to record the self-observation data (File S2).
The diary was to include the description and reflection of three types of common interactions in their usual clinical practice: (a) a standard situation of welcoming a patient for admission, (b) an interaction in which there was a pre-agitational state that required verbal de-escalation and (c) an interaction whereby the patient is approached individually, promoted by the nurse and in the absence of any demand on behalf of the patient. The structure of the diary, together with the instructions for completion, pursued two purposes. First, to enable nurses to reflect on their starting assumptions, to understand their practice, to understand themselves and their patients, and, finally, to understand their profession (Price, 2017). Second, to monitor the process of change planned for the PAR, according to the proposals of Kemmis and Mctaggart (2008). In this sense, for each interaction, the nurses had to record the description of the situation, the type of verbal and nonverbal language they had used, their reflected intervention, their emotions during the interaction and, finally, a reflection on the influence of the environment on the interaction. Once the nurses had completed the diary, they sent it to the research team by e-mail. The data were collected between April and June 2018.

| Ethical considerations
This study was approved by the Research Ethics Committees of all the participating hospitals. The nurses participated on a voluntary basis, and all participants signed an informed consent form. Nurses did not receive any compensation or incentive for participating in the study. To maintain the confidentiality and anonymity of the data obtained, each nurse received an alphanumeric code that was incorporated into their diary. The diaries were sent to a generic e-mail of the project that was only accessible to the principal investigator of the project, subsequently, the data were stored on a computer used exclusively for this study.

| Data analysis
The content analysis method was used to analyse the data (Crowe et al., 2015). The diaries reached the first author and were coded to preserve the anonymity and confidentiality of the participants.
Under their responsibility, the entire coding and categorisation process was carried out in a consensual manner by a collaborative team that formed the backbone of the process of developing a rigorous coding system (Merriam, 2016). In the first stage of analysis, the text was fragmented into descriptive codes assigned exclusively according to their semantic content. In a second stage, these initial codes were grouped into more analytical subcategories, which classified the codes according to the meaning of the linguistic units and their combinations. This led to a third hierarchical stage in which, considering the semantic analysis of the previous subcategories, the codes were ranked inductively. The first and second steps were taken iteratively until a more specific understanding of the subcategories was achieved. These steps were carried out primarily by the first author and discussed and reflected upon continuously and critically within the research team. Throughout the process, the QRS NVivo 12 program was used as computer support.

| Rigour
Reflexivity was continuous throughout the process. Most of the researchers were experts in mental health, with training in qualitative methodology and experience in previous similar studies. As this was a multicentre study and a very large research team, neutrality was ensured as team members adopted an open attitude towards sharing, reasoning and discussing the findings as they emerged. In addition, the team became aware of its initial onto-epistemological positioning, which was reflected in the design of the self-observation guide for this stage of the process. As the research progressed, team members repeatedly contrasted the experiences identified in the participants' diaries with their own opinions. They asked follow-up questions for the generation of new knowledge without guiding the participants' responses, so that this initial positioning could not influence the subsequent analysis. Similarly, the credibility and confirmability of the data should be emphasised, given the triangulation of the researchers in the analysis process and the constant auditing of the results by the participants in subsequent groups. In relation to the transferability of the results, in the case of this study, where participation is so high and from so many centres, it ensures that the results are valid for all units.

| FINDING S
The diaries of 152 nurses working at 18 centres were collected and analysed. The nurses ranged in age from 22 to 62 years, with a mean age of 33.6 years (SD = 9.4). Over 70% of the nurses were female. Their experience in mental health was a mean of 7.6 years (SD = 7.5). Almost a quarter of them had the official title of mental health nurse specialist and over 25% of the nurses had a doctoral or master's degree. All facility shifts were equally represented in the sample, although 40% of the nurses had rotating shifts or served on an as-needed basis ( Table 1) Patients are confused when they are first admitted and need the staff to listen to them and spend time with them. I always try to use an empathetic approach and be honest from the very beginning. I think it is very important for the patient to know that they can count on me, I try to convey that I am available if they need me.
(01DR101) However, they also identified that, in order to maintain this attitude, they had to be aware of barriers such as the presence of prejudice, the unavailability of other team members, the belief that the therapeutic relationship is useless, or lying to the patient.
In addition, they considered it extremely important that, as caregivers, they should adapt to the other person, that is individualise the care they provide in the context of the therapeutic relationship. This implies considering the patient's psychopathological and emotional state at any given moment, as well as the patient's age, language or culture. This often meant postponing interviews, adapting language, using sign language to communicate, agreeing on a special type of diet, or even relaxing the rules and letting the patient make a call outside the usual hours.

| Essential nursing actions for the therapeutic relationship
This category refers to the nurses' reflections on their actions in the context of the therapeutic relationship with patients. In their diaries, the nurses were describing and reflecting on different interventions and activities that were carried out in their usual practice and they detected certain actions that were common to all of them.
First, the nurses pointed out the importance of generating an appropriate environment to build a bond and facilitate the relationship with the patients. A calm, intimate, comfortable, unhurried environment without external stimuli or interruptions.
The room is quiet with the door closed and without any interaction from the environment…A pleasant and silent environment favors the therapeutic relationship between the professional and the patient. (04DR110) In relation to the establishment of a good therapeutic bond, the nurses agreed that the welcome provided on admission was a fundamental intervention. This was viewed as one of the situations in which the therapeutic relationship took on a greater relevance, since this first contact was considered the key to the success of the subsequent relationship with the patient.
Without welcoming the patient when he or she enters the unit, a better quality of the patient/professional relationship cannot be achieved.
Secondly, the nurses felt that the verbal approach was also a relevant aspect of their practice in the context of the therapeutic relationship. For them, it was an essential step in order to be able to carry out any intervention, such as when welcoming a patient when they are admitted to the unit, the use of verbal de-escalation techniques to ease the tension with very demanding and uncooperative patients or, on the contrary, to approach isolated patients who hardly interact with the environment, although the use of words is not always as effective as they would like it to be.

F I G U R E 1 Nurses' reflections on the practice of the therapeutic relationship in acute mental health units
Verbal containment is one of the most relevant parts of our work. In a pre-agitation situation, we may be able to transition a patient from pre-agitation to calmness or from pre-agitation to psychomotor agitation. (09DR108) In this sense, the nurses described that the act of offering the patient their assistance was at the heart of the therapeutic relationship.
They stated that this action was carried out in the context of being present, listening or through agreement with the patient by proposing alternatives to the demands and needs that they cannot meet.
As he speaks I give him my support with non-verbal language. I take his hand and he hugs me. I offer my help. We agree that he will make an effort to eat some solid food at dinner and that I will give him a supplement (he has it prescribed if he needs it). (01DR101) The nurses also acknowledged that interventions such as mechanical restraint were sometimes the only measure to reduce stimuli or were implemented because of patient aggressiveness, risk of escape or even medical indication. However, the nurses reflected that, although this intervention was performed relatively often, it could be seen as a failure and a deterioration in the therapeutic relationship.

| Contextual factors affecting the therapeutic relationship
The nurses identified contextual factors that facilitated or, on the contrary, acted as barriers to the therapeutic relationship. Indeed, they described that the type of admission could already condition the therapeutic bond, with voluntary admission being a facilitator.
The same is true of other factors such as knowing the patient from previous admissions, and whether the patient remembers having a good experience in those previous admissions. However, the nurses also considered elements that are intrinsic to the patient, such as language, culture or bad experiences of previous admissions, as factors limiting the establishment of the therapeutic relationship.
He is open to help and agrees to the admission (03DR110). I must admit that the fact that I know the user from previous admissions has helped the situation to unfold smoothly.

| DISCUSS ION
This study aimed to explore the phenomenon of the therapeutic relationship from the reflective practice of nurses in acute mental health units. The nurses highlighted that attitude was the core aspect of the therapeutic relationship after reflecting on their practice.
Similarly, they also reflected on the actions that were customary in the habitual interventions carried out in the context of the therapeutic relationship, identifying the most common barriers encountered in practice. Finally, the nurses reflected on those aspects of the context of care that conditioned the therapeutic relationship in the clinical practice of acute mental health units.
These findings offer knowledge about relational competence, a competency of professional nursing that is highly relevant in mental health (D'Antonio et al., 2014). This competence is directly linked to participation in practice and incorporates not only knowledge and skills, but also attitudes and professionalism that involve applying evidence and learning to practice (Casey et al., 2017;Moreno-Poyato, Casanova-Garrigos, et al., 2021). Specifically, the attitudinal component highlighted in the results and its importance in the context of the nurse-patient therapeutic relationship has been described from a theoretical perspective by authors such as Peplau or Orlando (Forchuk, 1991), Travelbee (1971) and Watson (Turkel et al., 2018). Similarly, the empirical literature has collected multiple studies that study the importance of nurses' attitudes towards more general aspects of mental health, such as stigma (Young & Calloway, 2021), recovery (Gyamfi et al., 2020), coercion (Doedens et al., 2020;Laukkanen et al., 2019) or severe mental disorder (Economou et al., 2019). However, there is hardly any empirical evidence that explicitly shows the relevance and identifies the specific attitudinal skills of nurses in the context of the practice of the therapeutic relationship. Thus, it is likely that the fact that the nurses were able to reflect on their practice made them more aware of the importance of attitude in the context of the therapeutic relationship (Harris & Panozzo, 2019a), as they were able to respond to the real challenge of establishing an adequate therapeutic relationship in their day-today work in the acute mental health units (Choperena et al., 2019).
Moreover, the attitudinal capacity identified by the nurses encompassed aspects already empirically recognised in the context of the therapeutic relationship, such as availability, communication and individualisation (Delaney & Johnson, 2014;Harris & Panozzo, 2019b;McAllister et al., 2019;Moreno-Poyato et al., 2016). However, the nurses also highlighted other aspects that have been less empirically studied, such as the importance of self-confidence and selfassurance, both in a positive way in order to be able to establish an appropriate therapeutic relationship, (Roche et al., 2011;Van Sant and Patterson, 2013) as well as negatively, in the form of limitation (O'Connor & Glover, 2017;Van Sant and Patterson, ). These results confirm the relevance of Peplau and Orlando's theoretical approaches and the use of the nurse's awareness as a fundamental part of the nursing relationship (Forchuk, 1991;Thomson et al., 2019).
The results indicate that by reflecting on their practice, the nurses were able to identify those skills (practices) that are essential for the development of the therapeutic relationship and which were transversal to any intervention. The nurses emphasised the importance of generating an adequate environment for the relationship, considering the environment not only as an element of context typical of many acute care units, but also as an element that is essential for the development of the therapeutic relationship (Kingston & Greenwood, 2020), also considering that it was their responsibility to be able to build the space where the relationship could take place (McAllister et al., 2021;Raphael et al., 2021). As in other studies, nurses also identified skills such as verbal engagement, offering help or working with the patient as basic practices for the development of effective interventions in the context of the relationship with their patients (Harris & Panozzo, 2019a;McAllister et al., 2019;Molin et al., 2018). Furthermore, in relation to specific interventions, reflection on practice allowed nurses to identify and become aware of nursing admission assessment and mechanical restraint as two common interventions in mental health units that were particularly influential in the therapeutic relationship with the patients.
In this sense, for the nurses, welcoming the patient on admission was considered an essential intervention determining a large part of the success in building the therapeutic relationship with the patients (Forchuk et al., 1998;Peplau, 1997). However, the use of mechanical restraint compromised the therapeutic relationship and the patient's trust (Kinner et al., 2017), although they understood that, even if this measure was undesirable, at times it was necessary (Doedens et al., 2020).
In addition, the nurses reflected on the contextual factors that directly affected the therapeutic relationship with the patients. In

| CON CLUS IONS
The present study contributes to the understanding of the phenom-

| RELE VAN CE TO CLINI C AL PR AC TI CE
These findings have important implications. The study findings demonstrate that participatory methods stimulate nurses' reflection, motivation and critical thinking. By learning from the reflection of the nurses themselves about the aspects that underlie the therapeutic relationship in their clinical practice, this enables the nurses themselves to become aware and to develop strategies for improvement based on their own knowledge. Moreover, the individual reflection involved in these first stages of a participatory process provides the nurses with an intrinsic knowledge of how they approach the therapeutic relationship and shows that the attitudinal component is basic for them. In this sense, understanding and confirming how the attitudinal component is a key element for nurses in the practice of the therapeutic relationship allows managers to evaluate strategies that promote motivation and facilitate the involvement of nurses in improving their practice with patients. Moreover, these results point to the need to conduct mixed or qualitative studies aimed at exploring the aspects that facilitate the motivation, empowerment and attitudinal training of nurses in greater depth, rather than studies that only seek improvements in the theoretical knowledge of the therapeutic relationship.

ACK N OWLED G EM ENTS
We would like to acknowledge all the participants of MiRTCIME.CAT project.

CO N FLI C T O F I NTE R E S T
No conflict of interest has been declared by the authors.

AUTH O R CO NTR I B UTI O N S
Study design: ARMP and PDH; Data collection: APT, FGP and GCG; Data analysis team: DTM; Final report draft: DTM, ARMP and PDH; Supervision the process of data collection and analysis and provide support and feedback during all study phases: ARMP; Contribution of the manuscript, and read and approved the final manuscript: All authors.

DATA AVA I L A B I L I T Y S TAT E M E N T
The data that support the findings of this study are available from the corresponding author upon reasonable request.