Speech and language therapists' perspectives of therapeutic alliance construction and maintenance in aphasia rehabilitation post-stroke.

BACKGROUND: Therapeutic alliance refers to the interactional and relational processes operating during therapeutic interventions. It has been shown to be a strong determinant of treatment efficacy in psychotherapy, and evidence is emerging from a range of healthcare and medical disciplines to suggest that the construct of therapeutic alliance may in fact be a variable component of treatment outcome, engagement and satisfaction. Although this construct appears to be highly relevant to aphasia rehabilitation, no research to date has attempted to explore this phenomenon and thus consider its potential utility as a mechanism for change. AIMS: To explore speech and language therapists' perceptions and experiences of developing and maintaining therapeutic alliances in aphasia rehabilitation post-stroke. METHODS & PROCEDURES: Twenty-two, in-depth, semi-structured interviews were conducted with speech and language therapists working with people with aphasia post-stroke. Qualitative data were analysed using inductive thematic analysis. OUTCOMES &
RESULTS: Analysis resulted in the emergence of three overarching themes: laying the groundwork; augmenting cohesion; and contextual shapers. Recognizing personhood, developing shared expectations of therapy and establishing therapeutic ownership were central to laying the groundwork for therapeutic delivery. Augmenting cohesion was perceived to be dependent on the therapists' responsiveness and ability to resolve both conflict and resistance, as part of an ongoing active process. These processes were further moulded by contextual shapers such as the patient's family, relational continuity and organizational drivers. CONCLUSIONS & IMPLICATIONS: The findings suggest that therapists used multiple, complex, relational strategies to establish and manage alliances with people with aphasia, which were reliant on a fluid interplay of verbal and non-verbal skills. The data highlight the need for further training to support therapists to forge purposive alliances. Training should develop: therapeutic reflexivity; inclusivity in goal setting, relational strategies; and motivational enhancement techniques. The conceptualization of therapeutic alliance, however, is only provisional. Further research is essential to elucidate the experiences and perceptions of alliance development for people with aphasia undergoing rehabilitation.

What this paper adds

What is already known on the subject

Evidence is emerging in the field of rehabilitation to suggest that the therapeutic alliance is an active ingredient in therapy delivery, which may be associated with both treatment adherence and engagement. However, we have yet to explore how this phenomenon interacts with treatment interventions in aphasia rehabilitation.

What this paper adds to existing knowledge

The findings suggest that the therapeutic alliance should include as a minimum getting to know the person, developing shared expectations of therapy, establishing a sense of role and goal ownership in therapy, and effective responsiveness.

What are the potential or actual clinical implications of this work?

Therapists should consider actively challenging their own cognitive biases, which may indirectly affect alliance construction, by engaging in a process of reflexivity. Further training to enhance alliance‐building could consider developing relational strategies, goal collaboration and motivation enhancement techniques specific to working with people with aphasia.

Introduction

In the fields of psychotherapy and counselling, the concept that treatment efficacy can be attributed to factors that are common to all therapies is well established (Martin et al. 2000, Horvath et al. 2011). The therapeutic alliance (hereafter referred to as ‘alliance’), sometimes referred to as the working or helping alliance, describes the therapist–client interactional and relational processes operating during therapy delivery (Green 2006). The alliance has been found to account for a modest but robust effect size in psychotherapeutic treatment outcome (Martin et al. 2000, Horvath et al. 2011).

In 1979, Bordin, a renowned American professor of psychology, developed his pan‐theoretical model of the working alliance which has largely dominated alliance conceptualizations in the latter part of the 20th century. Bordin's tripartite conceptualization of alliance incorporates: (1) the client–therapist agreement on the goals of therapy; (2) the mutual agreement and collaboration on explicit tasks required to meet those goals; and (3) the development of an affective bond (Bordin 1979). The conceptualization of the alliance, however, is not static, and as the healthcare fields in which the concept is applied expand, so does the construct, which has now been extended to include dimensions such as communication, empowerment and family systems (Pinsof et al. 1994, Kim et al. 2001, Elvins and Green 2008, Kayes and McPherson 2012).

The notion that the alliance may in fact be a variable component of treatment outcome has only recently begun to gain recognition in the arenas of healthcare and rehabilitation (Hall et al. 2010, Lawton et al. 2016, Morrison and Smith 2013). Preliminary findings suggest that the development of a positive alliance may affect treatment adherence, depressive symptoms, satisfaction, engagement and treatment efficacy (Hall et al. 2010, Bright 2015, Kayes et al. 2015).

Emergent data suggest that the development of a positive alliance is perceived by both clinicians and people with aphasia to be central to purposive aphasia rehabilitation (Fourie 2009, Hersh 2010, Worrall et al. 2010, Tomkins et al. 2013). Indeed, for people with aphasia, the therapeutic relationship was perceived to be as important as the targeted therapeutic activities (Worrall et al. 2010). The existential consequences of aphasia have been found to be extensive, leading to feelings of humiliation, alienation and loneliness (Nyström 2006). Nyström (2006) maintains that these consequences can be reduced if healthcare professionals engage in positive communication and interaction with the person with aphasia, highlighting the need for a strong patient–therapist alliance within the context of aphasia rehabilitation.

The potential relevance of the construct of alliance to the field of stroke rehabilitation has been highlighted in a recent meta‐ethnography (Lawton et al. 2016). However, emerging evidence suggests that psychotherapeutic constructs do not reliably capture the core components of the alliance relevant to a range of rehabilitation settings (Hall et al. 2010, Besley et al. 2011, Lawton et al. 2016). Further evidence is therefore required to ameliorate our understanding of this construct in relation to aphasia rehabilitation. The main objective of the study was to explore speech and language therapists’ perceptions and experiences of alliances, as one part of the dyad, with a view to conceptualizing how therapists develop and maintain therapeutic alliances in the context of aphasia rehabilitation.

Methods

Study design

This explorative, qualitative study used semi‐structured, in‐depth interviews to explore speech and language therapists’ perceptions and experiences of alliance construction and maintenance in aphasia rehabilitation (Braun and Clarke 2006). Qualitative enquiry was selected as it is ideally suited to encapsulating multiple, diverse realities and has the potential to offer the researcher rich, idiographic insights into the participant's lived experience (Braun and Clarke 2014).

This study was framed within a constructivist paradigm. Constructivism acknowledges that the findings are a result of a co‐created process, in which the researcher is perceived to have an active role in constructing and interpreting the data. Thus, the researcher and participant are inextricably linked through the interactive investigation process (Guba and Lincoln 1994). The authors sought to understand how individuals construe the world, from an emic (insider) perspective, with the aim of revealing the underlying processes inherent in alliance construction.

Ethical approval was obtained from the University of Manchester, UK Research Ethics Committee (reference number 15215) in addition to permissions from research and governance departments at individual National Health Service (NHS) sites, prior to the commencement of the study.

Participant recruitment and setting

Participants were eligible for inclusion if they were: (1) Health and Care Professionals Council (HCPC)‐registered speech and language therapists; (2) working in the field of aphasia rehabilitation in the UK; and (3) working with patients following a stroke. Participants were excluded who worked exclusively in the private sector because the development and construction of the alliance is likely to be inherently different due to divergent system constraints (Lawton et al. 2016). Purposive sampling based on the first author's professional networks was used to select sites, within a feasible 90‐mile radius of the field researcher (M. L.). Sites selected represented a range of different aphasia rehabilitation practices, encompassing acute hospitals, outpatient, early supported discharge and community services. Invitations to participate and information sheets were disseminated via e‐mail to lead speech and language therapists in 14 different NHS trusts in the North of England and to members who were affiliated to aphasia specialist interest groups in the North West of England. Lead speech and language therapists in the NHS sites were asked to distribute this information to speech and language therapists working in aphasia rehabilitation. Participants were invited to contact the researcher via e‐mail or telephone if they wished to participate in the study. Participants were given comprehensive written and verbal information, detailing the objectives of the study. All participants provided written informed consent prior to taking part in the study.

Data collection

The first author collected data using a semi‐structured, open‐ended interview frame from June to August 2015. Interviews were conducted on a face‐to‐face basis at the participant's place of work. Each participant took part in one interview. The majority of interviews ranged in duration from approximately 60 to 90 min. An interview schedule, focusing on broad concepts relevant to alliance construction, helped to direct the interviews. Alliance concepts included: relationship, collaboration, expectations and roles, motivation, experience, and familial involvement (Bordin 1979, Kim, Boren and Solem 2001, Besley et al. 2011, Bishop 2015, Lawton et al. 2016). Questions included:

Can you describe a time when you had a particularly good/difficult relationship with a patient?

How do you set goals for patients?

How do you encourage patients?

In what ways do your expectations align with the patients’ expectations during goal setting?

However, a flexible approach was used, allowing for new foci of interest to be discussed in depth as they emerged. The interview schedule was modified after each interview, as part of an iterative process, as new areas of theoretical enquiry became apparent, allowing for greater exploration of the emergent constructs. Interviews were audio‐recorded (Olympus digital voice recorder WS‐550M) and transcribed verbatim. All identifiable information was removed at transcription and pseudonyms were assigned.

Data analysis

Data were analysed in NVivo Version 10 (QSR Pty Ltd, 2012) following data collection, employing inductive thematic analysis, as described by Braun and Clarke (2006). This method was chosen because it provides a robust, systematic framework for analysis, leading to a rich narrative of the data, which is epistemologically flexible (Boyatzis 1998, Braun and Clarke 2006).

The first author listened to the audio recording and reread the transcripts on three occasions, noting emergent patterns and meanings across the data corpus that captured participants’ key ideas and perceptions (table 1, phase 1). In phase 2, data were coded systematically across the entire data set. Units of text that conveyed meaningful information relevant to the research question were coded by the first author. Data were coded inductively given the exploratory nature of the study. In phase 3, codes with brief descriptors on pieces of paper were juxtaposed and reorganized manually to identify categories, subthemes and tentative overarching themes. Themes were identified at a latent level, focusing on the implicit meaning underlying therapists’ assumptions and ideas (Braun and Clarke 2006), consistent with a constructivist epistemology. Negative cases that did not appear to fit with the core themes were examined and themes were revised to ensure these cases were reported. In phase 4, thematic maps were created to form a graphical representation of the tentative themes and their relationships, which were continually refined as part of an iterative process. Tentative themes were merged and modified, until consensus within the research team was reached regarding the validity and coherence of the themes, in relation to their underlying subthemes and codes. Theme development was further ‘refined and defined’ (Braun and Clarke 2006: 92) to capture more accurately the content of the given themes via discussion and consensus within the research team (phase 5).

Table 1

Phases of thematic analysis, as described by Braun and Clarke (2006)

1.	Familiarizing yourself with the data
2.	Generating initial codes
3.	Searching for themes
4.	Reviewing themes
5.	Defining and naming themes
6.	Producing the report

No further data collection was indicated following analysis as no new themes were emerging from the data following interview 18; therefore, themes were deemed to have reached conceptual saturation. However, data analysis continued until interview 22 to determine whether any further themes could be identified. All data analysed were included in the findings.

Trustworthiness and credibility were address by: (1) regular peer‐research team debrief meetings which ensured that the methodological decision‐making process was grounded within the data and subsequently defensible; (2) a detailed audit trail, which reflected on the development of key themes and provided associated rationales; and (3) negative case analysis which challenged and enhanced theme development. Alongside a detailed audit trail, the first author met regularly with the research team, some of whom had different clinical backgrounds, which served to challenge the first author's interpretations. Given that the use of member validation remains questionable in qualitative research (see Sandelowski 1993, 1998 for further discussion), the authors decided not to employ member checking to enhance the validity of the findings.

According to a constructivist stance, the prior knowledge and theoretical presuppositions of the researcher should be acknowledged and explored rigorously (Charmaz 2008). The researcher cannot be viewed as a simple bystander but must acknowledge their role in the interview and analytical process in order to challenge their own influences reflexively. This was particularly pertinent, as the first author is an experienced speech and language therapist who has worked extensively in the field of stroke rehabilitation. All decisions were therefore documented, which not only enabled the first author to reflect on those interpretations but also allowed decisions to be challenged and justified within regular research team debrief meetings.

Participants

Twenty‐two therapists (18 female, four male) took part in in‐depth interviews. Therapists’ average age was 37 years (range 24–58 years). Therapists worked in a variety of settings in either community, early supported discharge, inpatient or outpatient services (see table 2 for further information).

Table 2

Participant demographics

Participant number	Gender	Age (years)	Years post‐qualification	Years working in stroke rehabilitation	Setting
P01	Male	29	1	0.3	Inpatient/home
P02	Male	25	1	0.4	Home/outpatient
P03	Female	24	1	0.6	Inpatient
P04	Female	27	4	0.8	Inpatient
P05	Female	26	2	1.5	Inpatient/home
P06	Female	29	2	1.7	Inpatient/home
P07	Female	29	3	2.0	Home/outpatient
P08	Male	36	5	3.0	Home
P09	Female	28	6	5.5	Inpatient
P10	Female	36	10	9	Inpatient
P11	Female	58	11	10	Home
P12	Female	39	12	12	Inpatient
P13	Female	41	19	12	Home
P14	Female	36	15	15	Inpatient
P15	Female	40	18	15	Inpatient
P16	Female	39	16	16	Home
P17	Male	42	16	16	Home
P18	Female	44	21	20	Home
P19	Female	48	22	21	Home/inpatient/outpatient
P20	Female	48	25	25	Home/inpatient
P21	Female	50	28	29	Home/outpatient
P22	Female	50	29	29	Home

Results

Findings

Three themes were identified from the process of analysis: laying the groundwork; augmenting cohesion; and contextual shapers. The first two themes describe how therapists engaged in specific strategies, with the aim of fostering an effective alliance. The latter theme explores how external variables interacted to both impede and facilitate alliance development.

Laying the groundwork

The theme laying the groundwork describes the processes employed by therapists, which were perceived to underpin genuine mutuality, establishing the context for patient engagement. The three sub‐themes: recognizing personhood, sharing expectations and activating ownership elucidate the interpersonal processes used by therapists which set the scene for targeted therapeutic interventions (table 3).

Table 3

Laying the groundwork and augmenting cohesion: theme development and examples

Theme	Sub‐theme	Processes	Examples
Laying the groundwork	Recognizing personhood	Getting to know the person	just providing them with as much opportunity for them to express themselves, to get what they're all about and who they are as a person really. (P02)
		Showing empathy	you acknowledge that yeah you've had a really bad day, and you know be very reassuring and all of that, and acknowledging all of that. (P04)
		Using self‐disclosure	You know people want to kind of know a little bit about you as well because it makes it more balanced. But as long as you're not banging on about yourself. (P03)
		Making value judgements	It's difficult to treat someone when you … don't like I suppose a lot of their views or erm, the way they express them, er, or their, yeah, or their politics or their, you know, those relationships are more difficult. (P16)
	Sharing expectations	Attending to realism	we try to set a very realistic picture based on their initial presentation really. So it's trying to be, make sure that the patient is informed throughout the whole journey. (P15)
		Preserving hope	I don't think that you can dismiss someone's hopes, and say no I know you're not going to improve right from the outset because you've lost them really haven't you. (P18)
	Activating ownership	Delineating roles	So that's what I explain I need from them, and for them to continue I need them to engage, so I'm kind of setting out what I want from them as well really. (P06)
		Encouraging goal ownership	I still will make it collaborative, in the sense that I'll say ‘This is what I think, but what about this, you know you touched on about how it was difficult to do X and X and X, so what about if we worked on something like that.’ And they might say ‘Yeah yeah.’ So I try and make it collaborative as much as I can. (P08)
Augmenting cohesion	Being responsive	Being attuned	We've got a chap at the moment, highly, highly anxious, and I have to be very, very careful about the, you know, that the tasks that we do, or activities, are … don't stretch him too much because he really, really, really gets very anxious then and he can't perform or can't produce what he'd like to and then, and then he, then it gets worse, so it's working within that. (P22)
		Being encouraging	I feel, he required of me a kind of erm, ‘I need you to believe that I can manage this.’ (P20)
		Seeing the point	Just to explain what you're doing, so they kind of know what's happening, and they don't really feel like everything's happening to them, they're kind of like, ok, I'm engaged in this process, I need to do this for this reason. (P12)
		Using humour judiciously	‘I take the lead from the patient, if they're happy to laugh at themselves. I just think sometimes it diffuses a lot of frustration sometimes. (P14)
	Resolving conflict	Negotiating participation	Sometimes I just end, well end up asking them about other things, and if they're struggling a bit with the therapies, maybe deviate away from what you had planned but have a conversation and see whether you can work out why they're lacking in motivation. (P12)
		Managing boundaries	She'd got herself referred back, nothing had changed neurologically, but her husband had died, and she was quite lonely and she managed to express to me that ‘Nobody chats to me the way you chat to me.’ (P14)

Recognizing personhood

The sub‐theme recognizing personhood describes the processes used by therapists to acknowledge the patient as a person. For some therapists, this extended to indicating their own personhood, through the use of self‐disclosure. Recognizing personhood was perceived to provide the synergistic context for openness and trust and laid the foundations for developing mutuality.

Acknowledging personhood was dependent, for many therapists, on getting to know the person with aphasia. The majority of therapists spoke of prioritizing getting to know the person over formalized assessment at the beginning of therapy, in part to understand individual priorities, but also in an attempt to personalize the connection. Both eliciting and acknowledging personhood became more challenging for therapists in the context of global impairment and therapists described interactions as more superficial. However, therapists used multiple supportive communication strategies, such as gesture, drawing and pictures, to provide a scaffold for conversations to circumvent these barriers, techniques which were minimally adopted by other professionals, contributing to the perception that the connection in itself was unique to the therapist–patient dyad. In the following excerpt, one therapist described how the implementation of supported communication strategies and the allocation of time resources were used to establish a sense of personhood, even in the context of more severe linguistic deficits:

It's weirdly still quite easy to form a relationship with someone even though they can't really talk to me, I think I go a lot off their cues and gestures, and erm, obviously it's more tricky if someone can't talk and tell you the answer, but I'm always getting my pen and paper out and trying to find out a bit about them, and I think quite often with someone more severe, because we're there and we're going to spend more time with them … I think that person is quite often pleased we're sat there with them, trying to get a bit of what they say. (P09)

Getting to know the person was not always prioritized by therapists and some therapists felt that allocating time to this process was not justified, particularly if it did not directly inform therapeutic interventions. This perception was largely ascribed to novice therapists, but several more experienced therapists also felt that spending time getting to know the person took valuable time away from therapeutic interventions and was perceived to be incongruent with patients’ expectations of therapy: ‘often I don't feel the patients do want to chat that much about other stuff. Because that's not what you're there for’ (P11).

Acknowledgment of personhood was dependent on therapists being cognizant of the emotional sequelae of both the stroke and aphasia. For many therapists, demonstrating an understanding of the devastating impact of stroke and aphasia was central to imparting a sense of empathetic awareness and was reportedly evidenced through listening, acknowledging and normalizing the patient's post‐stroke experience, allowing the person with aphasia to recount their personal narrative. As one therapist suggested, ‘it's a bit like childbirth, because everybody has to tell you their story’ (P11). Several therapists described empathy as evolving, in line with their own life experience. In the following excerpt, one very experienced therapist described the impact of her own experience on her clinical practice:

I've had more experience of different sorts of relationships in your own life, so dealing with difficulties or seeing what happens in your family or friends when they have difficulties to deal with. I can be more appreciative of how hard it can be to reach that starting point of being able to understand. (P21)

Therapists described using self‐disclosure to varying degrees in an attempt to reveal their own personhood. At one end of the continuum, therapists felt that revealing personal information was central to the construction of the alliance, allowing the patient to perceive the therapist as a person, thus cementing equipoise between the therapeutic dyad: ‘It's about laying yourself on the line as well and expecting them to … you're a human being and they want to see that you're a human being’ (P01). Whereas at the other end of the continuum, therapists perceived that the focus should be on the individual and that professional boundaries should be maintained, sharing personal details infrequently. Therapists appeared to have very fixed ideas about how much information they should share with patients about themselves, irrespective of experience or the setting in which they worked.

Therapists felt it was easier to connect with someone who exhibited attributes of openness, genuineness and a sense of humour, which in turn influenced their own reciprocal behaviour. Indeed, they described feeling more relaxed and open with patients who engaged actively in reciprocal interaction, in both a verbal and a non‐verbal context. Similarly, when patients were judged to be closed off, reserved and were uncomfortable disclosing personal information or held significantly divergent opinions to them, therapists assumed a greater degree of formality. Therapists’ value judgements were inextricably linked to their own behaviours which were more likely to mirror those of the patient, reinforcing or impeding potential disconnect. In the following example, one therapist describes emotionally distancing herself from a patient who held disparate political views: ‘it's a professionalism, that kind of takes over, I suppose it is less of a bond because it's more of an autopilot’ (P16).

Sharing expectations

Sharing expectations was part of the process of laying the groundwork and was deemed to be central to reaching a consensual understanding regarding the direction of therapy. There was an implicit assumption, for many therapists, that the patients’ expectations should align with their own perceptions of what was achievable within the context of the therapeutic process. The majority of therapists perceived that many patients had unrealistic expectations about the likely outcomes of therapy and spoke of adopting a realistic approach in the early phases of rehabilitation, as they wanted to avoid any potential misunderstandings, make goals achievable and avoid fostering false hope. Several therapists felt that a failure to address what they perceived to be unrealistic expectations at this stage, not only had the potential to disengage patients, but could also lead to conflict in the relationship. In the following excerpt, one therapist felt that she would be culpable for a perceived lack of progress in the future if she did not manage patient expectations early on:

They have these high expectations that you're going to make them better and I think if you don't do that [manage expectations] from the outset then they think it's all about you and therefore if they don't progress then it's you that's failed them. (P18)

Therapists recognized that attendance to a realistic approach had the potential to damage hope, which in turn had the potential to affect motivation. The majority of therapists attempted to ‘walk the line’, carefully balancing what they perceived were realistic expectations with hope. The patient's ability to absorb these messages was not only perceived to be dependent on the degree of trust already established, but also was reliant on information sharing and cognitive ability. Several therapists prioritized the preservation of hope over a realistic approach in aphasia rehabilitation, as they believed that they were both protecting the individual's perceived fragile psychological status and safeguarding the therapeutic connection. These therapists believed that patients’ expectations would adjust over time and focused on working on all of the patients’ goals, irrespective of how unrealistic they may seem. The processes of self‐realization and experiential learning were perceived to negate the requirement to give explicit information about therapeutic recovery. As one therapist explained, ‘it might not even be healthy for them to be realistic at that point, because it might be that's all they can manage to keep themselves going. They need to have that hope’ (P13).

Activating ownership

Activating ownership was perceived to be integral to developing a therapeutic partnership based on mutuality. During the early stages post‐stroke, therapists attempted to encourage patients to take ownership of their rehabilitation, with the aim not only of promoting patient empowerment but also of shifting the responsibility from the clinician towards a collaborative venture. Therapists felt it was important to demarcate both their own and the patient's role early on in the therapeutic process. One therapist spoke of using his favourite ‘lines’ with patients:

I'm not a magician, I'm a guide, I can't do the work for you, I can show you what you need to do and I'll help you as much as I can along the road with it, but a lot of it is going to be down to how much you put in. (P14)

Activating ownership for many therapists involved encouraging patients to participate in the process of identifying amenable goals for intervention. Goal ownership and collaborative working was perceived to be only fully realized when patients engaged in identifying their own personally relevant goals for therapy. However, many patients were perceived to have a limited awareness about what they wanted to achieve, particularly in the early stages post‐stroke and many therapists spoke of adopting a benevolent paternalistic approach, deferring involvement in goal setting until a later stage. Several more experienced therapists spoke of co‐constructing goals, a process which was reliant on the therapist taking a more active role in guiding discussions about aspirations and areas of difficulty. The level of direction was often dictated by the cognitive, linguistic and psychological status of the patient and the patient's desire to become involved in identifying goals and driving therapy forward. Goal ownership, therefore, was a continuum in which patients assumed varying degrees of ownership, dependent on their ability and desire to participate, the therapist's resources and the passing of time.

Augmenting cohesion

Augmenting cohesion encapsulates the active process in which therapists sought to strengthen the partnership through the sub‐themes: being responsive and resolving conflict, with the aim of sustaining engagement (table 3).

Being responsive

Therapists were cognizant of individuals’ relational and informational needs and described adapting their behaviour to meet these needs. The therapist's ability to ‘read’ the patient's non‐verbal cues was perceived to dictate how effectively they could respond to patient apathy or decide when it is appropriate to push or not to push and, similarly, when to use praise and explicit feedback to motivate. These processes were thought to be a variant of their own clinical experience and innate ability. Therapists’ ability to respond effectively to the psychosocial impact of the impairment, rather than focusing on the communication deficit in isolation, had the potential to further solidify the alliance:

People have more trust in you and if you're, you know sometimes you go to do a session and actually emotionally they're not in the right place, and responding to that rather than saying well I've planned this and we're gonna do this … how you respond in those sessions, I think enables a better relationship. (P12)

Humour was used reflexively by several therapists to mitigate embarrassment and to provide light relief within the context of challenging therapy session. Its perceived success lay in its accessibility and potential to bypass verbal channels, allowing therapists to use humour with patients with varying degrees of aphasia. The following quotation illustrates how humour was used responsively both in relation to the individual and the context: ‘You might joke to a degree with the patient, about their actual difficulties, and that's only some people that would get benefit, and you can usually gauge that from them, because they tend to do a little bit themselves’ (P08).

Therapists also needed to be responsive to individuals’ informational needs. Some patients were reportedly happy to engage in therapeutic tasks without grasping an understanding of their relevance; however, for others, lack of understanding of why they were engaging in specific exercises was perceived to be problematic. Therapists believed that ‘seeing the point’ of participating in a given therapeutic activity or being able to relate that task to a goal was central to enhancing both the alliance and subsequent engagement. Developing a patient's understanding of how a given activity related to their goals could reportedly be challenging for those with more severe linguistic and cognitive deficits. Repeated explanations, both verbally and non‐verbally, were often described as futile because it was not always possible to embed these understandings. In these instances, several therapists spoke of the importance of making tasks more relevant and therefore meaningful. However, several therapists suggested that trust could override this need to make sense of therapy activities, as indicated in the following excerpt:

If you can build some trust with somebody that they've feel that I've got their best interests at heart. And, whatever I'm doing, so if I don't know, say it's auditory discrimination or something, that can seem a long way from the problems that they're having, that they gradually begin to see it does have a connection. (P21)

Resolving conflict

Therapists responded to disengagement by using a range of strategies including re‐negotiating involvement, identifying potential barriers and revisiting goals. Some therapists described using bargaining techniques in which they attempted to persuade the patient to engage collaboratively in the therapeutic sessions for short periods of time, with the premise that they could leave at any point or take a break, allowing the patient to assume greater control over the situation. In contrast, several therapists engaged in a more direct, frank approach, in which they outlined their expectations of the patient's role, in an attempt to shift the responsibility of change towards them. In the following example, one therapist, who spoke of using a more direct approach, explained how she confronted patients who were not engaging in the therapeutic process:

If they're not giving a good chunk, then nothing is going to change. Erm, and sometimes you do have to have that conversation with people. That if they're not doing anything in‐between sessions, or not altering the way they do things, then how can they expect it to change. (P21)

Therapists recognized that negotiation was not always viable and sometimes patients needed a break from the intensity of rehabilitation. However, therapists working in the acute sector and early discharge teams felt that they had a limited time window to conduct therapy which meant that many therapists wanted to re‐engage patients in the therapeutic process rather than waiting for them to feel ready. Many therapists described feeling ill‐equipped, lacking the necessary skills, to instigate behavioural change to influence motivational readiness. Indeed, even those who were very experienced felt a sense of helplessness when dealing with those who were perceived to be completely disengaged in the rehabilitation process, which impacted on their own perception of competence: ‘I don't really know how to go about trying to draw somebody out I guess and trying, trying to engage if they're very much, you know, not wanting to’. (P12)

Although the development of an affective bond was perceived to be fundamental to treatment efficacy and engagement, there was an awareness that if the bond became too close, it could create dependency. Therapists’ responses to attachment varied. Several therapists reinforced this attachment by confirming their indispensable status within the dyad: ‘she plateaued in terms of impairment based therapy. … But now, I'm probably going to have to stay involved a bit longer, just to kind of hold her hand for a bit’ (P01). In contrast, other therapists argued that creating any form of dependence was detrimental to the patient's rehabilitation. One therapist described how she was acutely aware that her personal boundaries had been crossed when her patient turned up at her wedding uninvited. She was, however, unclear as to how to demarcate these boundaries without damaging the close interpersonal connection they had established: ‘I'm not sure what I could have done differently to be honest, to keep that lady out of my life [laugh]’ (P14).

Contextual shapers

The theme contextual shapers captures the variant influences of organizational drivers, relational continuity and family in facilitating and impeding the development of the alliance (table 4).

Table 4

Contextual shapers: theme development and examples

Categories	Examples
Setting	I think once I'd moved out of the hospital I realised that I probably hadn't been seeing the real person. (P22)
Organization drivers	At senior management level anyway it's more looked upon that ‘you need to be meeting your targets and why have you missed this one’, as oppose to thinking about the person. (P09)
Relational continuity	And it's about patients feeling cared for I suppose, isn't it? If you're interchangeable with every other therapist, then how much do you really care about seeing them through sort of thing. (P11)
Family	if the family are there as well, and if they seem a bit closed off with me, that will probably affect how I am with the patient as well. (P09)

The relationship in the hospital environment was described by many therapists as clinical and impersonal, characterized by lack of depth, which was not only ascribed to the clinical environment, with its white washed walls and uniforms, but also to the time and resource constraints afforded to the inpatient setting. ‘I just feel because we don't feel they're going to be in that long, that we don't build up that relationship as much. So we don't spend the time doing so, it's much more clinical’ (P03). Similarly, one participant (P11) spoke of being unable to identify a patient she had formed a close bond with in the last few years whilst working in an acute setting.

Time constraints and a pressure to ‘do’ therapy to meet early discharge targets and adhere to a hospital policy that favoured SMART goals were perceived to influence both relational and collaborative elements of alliance development. Those patients who were given the opportunity to collaborate were described as being ‘steered’ (P09) towards these ‘realistic’ short‐term goals, in order to meet early discharge targets:

It's kind of going back to the patient, you know, maybe saying what they would like to achieve, and then jointly trying to tail back a little bit to make it into a SMART goal, you know what we can review on a regular basis. (P15)

In contrast, the home environment was perceived to cultivate a relaxed atmosphere, conducive to building rapport and collaborative working, altering the potential dynamics of the relationship. Therapists who were part of community teams or early supported discharge teams tended to have more flexibility in terms of time and were therefore likely to adhere to a more patient‐centric approach: ‘Being accepting of all their goals, not just our goals for this rehab period, and certainly not just the speech therapy goals, you know, holistic goals’ (P16).

Developing an alliance was not only isolated to the therapist–patient dyad but also extended to the family. The familial connection was of paramount importance to therapists and, for some, commensurate with the patient–therapist relationship. It was thought to influence indirectly the patient's perception of his/her therapeutic relationship; the therapists’ own behaviour with the patient; and the therapists’ perceived ability to be effective. In the following excerpt, one therapist describes how she had found it difficult to establish a positive connection with the patient in the presence of negative behaviours exhibited by family members:

I found it very difficult to build a rapport when his wife was in the room rolling her eyes at me [laugh]. That felt like, all the usual things I'd do like talk to him about his interests or, see what his sense of humour was like, that was quite difficult to do when someone seems to be scathing at everything that's coming out of your mouth. (P14)

Although therapists recognized that they could develop a form of connection within a few sessions, the depth of connection and trust was perceived to be a product of time and frequency of contact. One therapist spoke of a patient with whom she had developed a particularly close bond, ascribing this close connection to relational continuity. In the following excerpt, she described how the patient had disclosed personal family information to her. She attributes this level of trust to being present throughout the patient's rehabilitation journey: ‘she just felt connected … she actually was one that I'd seen all the way through, I'd seen her in acute, she was, I was one of the first people that they saw, very early on, on day one’ (P19).

Discussion

The findings suggest that therapists used multiple strategies to forge alliances with people with aphasia with the aim of laying the groundwork for therapeutic working. Lack of homogeneity in alliance development, evidenced by differing approaches to establishing professional boundaries, managing expectations and activating ownership, suggests the construction of the alliance was highly complex and subject to individual variation, a process in which therapeutic skills, organizational drivers and value systems interacted to dictate the success of the alliance.

Therapists in the current study employed a range of strategies to establish and maintain relational connectivity in aphasia rehabilitation, a finding mirrored by studies in speech and language therapy (Fourie 2009, Worrall et al. 2010), physiotherapy (Besley et al. 2011) mental health (Shattell et al. 2007) and occupational therapy (Palmadottir 2006). The prevalence of these strategies across a range of disciplines suggests that a number of generic approaches to alliance construction appear to be applied regardless of the therapist's vocation or training. Indeed, Bordin's (1979) construct of alliance, applicable to psychotherapeutic contexts, appears to be highly relevant to the current findings. Within the first theme, the importance of bond is exemplified in the sub‐theme of personhood. Similarly, the collaborative element of tasks and goals resonates through the sub‐theme activating ownership. However, the strategies employed by therapists to establish personhood and develop a sense of ownership were subtly nuanced in the context of language impairment. This is evidenced in therapists’ skilled integration of both verbal and non‐verbal skills—more specifically in the use of humour, gesture, graphics and non‐verbal expression—to build and maintain alliances.

In contrast to other professions working in rehabilitation (Rosewilliam et al. 2011, Lawton et al. 2016) communication impairment was not seen as a major barrier to establishing a connection because of therapists’ perceived understanding of aphasia and enhanced communication skills. Similarly, speech and language therapists in Hersh's study (2010: 289) perceived that they were in a unique position as they were often the only person to be able to forge a ‘real relationship’, precisely because of their communication skills. Communication impairment was, however, thought to impact on the degree of collaboration and subsequent ability to reach consensus, supporting findings from earlier systematic reviews (Lawton et al. 2016, Rosewilliam et al. 2011). Conversely, Rohde et al. (2012) were able to elicit goals from participants (people with aphasia) engaging in aphasia rehabilitation, where speech and language therapists had reportedly been unsuccessful. Therefore, the authors contended that it was not, in fact, the communication impairment that was constraining collaboration, but instead ascribed this lack of collaboration to; the therapists’ priorities and value systems; a service delivery model prioritizing an impairment based approach and constraining time; lack of supportive communication skills and materials; and the client's reluctance to participate. Indeed, a number of studies have shown that even people with moderate to severe linguistic impairments and cognitive impairment can be included in goal setting with the use of pictorial resources, such as talking mats (Bornman and Murphy 2006), life interest and values cards (Haley et al. 2013). Goal collaboration, it would seem, is certainly possible for many people with aphasia but is reliant on the therapist's skill, time resources and the patient's desire to participate.

The importance of aligning expectations and maintaining hope in the construction of the alliance accords with findings from previous studies (Besley et al. 2011, Bright et al. 2011, Soundy et al. 2014, Lawton et al. 2016). Soundy et al. (2014) identified five strategies used by health providers that promoted hope which align closely to those strategies employed by speech and language therapists in the current study. These strategies included getting to know the person; imparting information in a sensitive and empathetic way; promoting autonomy; demonstrating understanding; and creating a light‐hearted atmosphere conducive to fostering hope (Soundy et al. 2014). Given the similarity in the strategies employed by healthcare providers, it could be postulated that the generation of hope is closely associated with alliance formation.

Therapists in the current study largely assumed a realistic approach to managing expectations, however patient qualitative data have suggested that therapists’ adherence to a realistic approach, focusing on explicit instruction may not be as effective as experiential learning (Kubina et al. 2013). In contrast, patients in Shattell et al.’s study wanted direct, honest and frank feedback in order to identify solutions to their problems (Shattell et al. 2007). Further research from the patient's perspective is needed to improve our understanding of how to approach and manage patients’ expectations of what is achievable in therapy if patients’ expectations are in fact unrealistic (Alaszewski et al. 2004).

Activating patient ownership of rehabilitation was deemed to be core to establishing an effective alliance, but was often perceived to be problematic. People with aphasia may be more reluctant to assume ownership of their therapy because of their perceived role expectations. Simmons‐Mackie and Damico (2011) maintain that patients’ role expectations may well be grounded in the contextual frame of the aphasia rehabilitation setting which individuals may align to similar, more familiar contextual frames such as those of the classroom and medical interactions. Therefore, people with aphasia are more likely to adapt their interactions to accommodate this new context within their existing institutional frames of reference and be less likely to take ownership of their therapy. The contextual framing of the patients’ expectations allows us to understand the complexities not only of arriving at shared expectations but also of developing collaborative goals and the patient's positioning in this process.

Although the level of aphasia was not perceived to be a major obstacle to establishing personhood, the patient's perceived personal attributes did appear to influence the ease with which the dyad established a connection. Therapists tended to attribute positive and negative behaviours to specific attributes which, in turn, influenced their own affective reactions and subsequent behaviour. This finding has not been clearly documented in previous stroke rehabilitation studies; however, therapists’ behaviours may be understood in within an attribution framework. Attribution theory proposes that an observer is more likely to attribute a person's given behaviour directly to his/her stable qualities, failing to recognize the impact of external social influences (Ross 1977). Attribution theory thus helps us to understand why therapists were more likely to attribute salient behaviours to internal factors, overlooking the potential situational explanation for a given behaviour.

Alternatively, it could be posited that patients who are perceived to be closed off or reserved are a tangible threat to the speech and language therapist's role. A failure to comply with the role of the ideal patient (i.e. a reluctance to engage in conversation) can deny the legitimacy of the provider's role, and thus threaten competence and control (Kelly and May 1982, Koekkoek et al. 2006). This challenge to individual competence and control has been found to lead to distancing and labelling of patients as ‘difficult’, with subsequent ramifications for patient care (Shattell 2004). The findings highlight the need for therapists to engage in a process of reflexivity to actively challenge these value judgements. In order to challenge these biases, Varcarolis et al. (2006) suggest that professionals should: acknowledge the presence of value judgements; explore how and where these judgements were acquired; and develop new ways of interpreting patient behaviours.

The alliance, however, does not exist in a vacuum and cannot be simply viewed as a product of two dyadic agents. Time and physical boundaries provided the impetus to both impede and drive alliance development. The stability of the alliance was forged by the frequency of contact, a finding replicated in the mental health literature (Littauer et al. 2005), which was shaped by the resource capacity of the health authority and the environment. The question remains as to whether a strong therapeutic relationship, requiring relational continuity, as opposed to a good enough relationship, as suggested in psychotherapy (Horvath et al. 2011), is an essential ingredient for purposive aphasia rehabilitation.

The current findings suggest that familial involvement was perceived to be central to the therapeutic intervention and therapists spoke of developing different but essential relationships with key family members. Pinsof et al. (1994) posit that the alliance should allow for both the patient's system (including family members, carers etc.) and the therapist's system (including all other persons involved in the therapeutic process, assistants, receptionists etc.). Therapists in this study believed that the alliance with the patient's key communicative partner was not only essential for rehabilitation but also may indirectly affect the patient–therapist alliance, highlighting the requirement for future research to explore whether a potential third domain is relevant to alliance construction and maintenance in aphasia rehabilitation.

Although this study advances our current understandings of the construct of alliance in aphasia rehabilitation, limitations are acknowledged in the interpretation of the findings. Firstly, this study was limited in terms of its focus on the therapist's perspective of the alliance, omitting the central perspective of the patient and potentially a third dynamic: the family or carers. Furthermore, the current findings were limited to the setting in which the data were collected. It is viable that the development and maintenance of alliance may be subject to cultural and organizational variation, although it appears that many of the strategies reportedly employed by therapists were generically applied by a range of professionals cross‐culturally. The sample size was, however, meaningful (n = 22) and diverse in terms of the cultural diversity within the speech and language therapy profession, the therapists’ gender, clinical experience, age and healthcare sites (n = 10), which served to enhance the internal validity of the findings.

Clinical implications

The data suggest that therapists employed a complex interplay of skills to foster and maintain alliances; however, even experienced therapists reported they did not always know how to motivate patients or demarcate boundaries and were unaware of how cognitive biases may affect their own behaviour. Emergent studies suggest that training can be implemented both to foster alliance development (Crits‐Christoph et al. 2006) and to manage alliance ruptures effectively (Castonguay et al. 2004), suggesting that further training, at both under‐ and postgraduate levels, may improve alliance development and maintenance in aphasia rehabilitation. The current findings suggest that training should focus on developing: (1) professional reflexivity within peer‐supervision sessions to reflect on potential value judgements and/or boundary issues (Hersh 2013); (2) motivational enhancement techniques to effect behavioural change to influence motivational readiness and manage disengagement (Crits‐Christoph et al. 2006); and (3) strategies to support collaborative goal setting using total communication (Bornman and Murphy 2006, Haley et al. 2013). Furthermore, consideration should be given to the system in which the alliance operates. Relational discontinuity and a medicalized model promoting SMART goals were perceived to inhibited alliance development. Assignment of key therapists and adherence to SMARTER goals (Hersh et al. 2012a) would facilitate further alliance growth with minimal financial implications. Perhaps more costly is the allocation of time resources; however, this is central to getting to know the patient which is inextricably bound to the process of patient‐centred goal setting (Hersh et al. 2012b).

Alliance development was not only reliant on the therapist's skills but also required time. It is questionable whether clinicians will be able to prioritize the development of the alliance, particularly in a target‐driven healthcare culture, which paradoxically deprioritizes a patient‐centred agenda (Lawton et al. 2016). Further empirical evidence is needed to determine whether the alliance is a variable component of treatment efficacy in aphasia rehabilitation by using a robust alliance measure.

Conclusions

The findings provide novel conceptual understandings of alliance development and construction in aphasia rehabilitation from the perspective of the therapist, which can be used to develop a conceptually robust measure of alliance relevant to aphasia rehabilitation and provide practice guidelines for rehabilitation specialists. Our analysis provides new insights into the way in which therapists develop and manage alliances in order to maintain a context conducive to therapeutic engagement.

This research will be not only pertinent to speech and language therapists working in stroke rehabilitation but also relevant to all healthcare professions working in rehabilitation, particularly with patients with aphasia. The findings do not, however, provide insights into the patient's perspective of the alliance or which elements of the alliance are necessary, as opposed to optional. Further qualitative enquiry should focus on exploring the perspective of people with aphasia regarding alliance construction and determining which elements of the alliance are integral to the therapeutic process. Training and practice guidelines should focus on the content of therapy and also explore how therapy is constructed. We would argue that the alliance is an essential ingredient in purposive rehabilitation which has yet to be exploited to its full potential.