Disability and Rehabilitation

ISSN: 0963-8288 (Print) 1464-5165 (Online) Journal homepage: https://www.tandfonline.com/loi/idre20

Understanding the therapeutic alliance in stroke

rehabilitation

Megan Bishop, Nicola Kayes & Kathryn McPherson

To cite this article: Megan Bishop, Nicola Kayes & Kathryn McPherson (2019): Understanding
the therapeutic alliance in stroke rehabilitation, Disability and Rehabilitation, DOI:
10.1080/09638288.2019.1651909

To link to this article: https://doi.org/10.1080/09638288.2019.1651909

Published online: 21 Aug 2019.

Submit your article to this journal

View related articles

View Crossmark data

Full Terms & Conditions of access and use can be found at

https://www.tandfonline.com/action/journalInformation?journalCode=idre20
DISABILITY AND REHABILITATION
https://doi.org/10.1080/09638288.2019.1651909

ORIGINAL ARTICLE

Understanding the therapeutic alliance in stroke rehabilitation

Megan Bishopa,b , Nicola Kayesa and Kathryn McPhersona,c
a
Centre for Person Centred Research, Auckland University of Technology, Auckland, New Zealand; bSchool of Physiotherapy, University of
Otago, Wellington, New Zealand; cHealth Research Council of New Zealand, Auckland, New Zealand

ABSTRACT ARTICLE HISTORY

Purpose: The quality of the therapeutic alliance between a client and their clinician is thought to play an Received 12 November 2018
important role in healthcare but there is limited research about this concept in stroke rehabilitation. This Revised 30 July 2019
study explored the core components of a therapeutic alliance and the factors perceived to impact on its Accepted 31 July 2019
development in a stroke rehabilitation unit.
KEYWORDS
Methods: Interpretive description methodology was used to gather and synthesise participants’ experien- Therapeutic alliance;
ces of their therapeutic relationships. Ten individual client interviews and one clinician focus group were working alliance;
conducted. Data was were analysed using conventional content analysis. therapeutic relationship;
Results: A therapeutic alliance appeared to consist of three overlapping core components: a personal rehabilitation; collaboration
connection, a professional collaboration, and family collaboration. Clients valued these components to dif-
ferent degrees and priorities could change over time. Alliance breakdowns were perceived to stem from
a clinician’s incorrect assumptions about their client’s relationship preferences or lack of responsiveness
to their needs. Recovery of the alliance seemed to depend on the strength of the pre-existing relation-
ship and steps taken to repair it.
Conclusions: Establishing and maintaining a therapeutic alliance appears to be an individualised and
complex process. A clinician’s ability to use their personal attributes therapeutically, and professional skills
flexibly, appeared integral to relationship quality.

ä IMPLICATIONS FOR REHABILITATION

Developing therapeutic relationships requires a person-centred and sometimes family/wh
anau-
centred approach.

The judicious use of self-disclosure may achieve emotional proximity and yet maintain professional
boundaries.

Maintaining relationship health requires a proactive approach to detect and manage relationship
disruptions.

Introduction maintenance, and repair of this collaboration [21,22]. The

The quality of the therapeutic alliance has long been viewed as a
catalyst for treatment success, with most evidence embedded in
30 years of psychotherapy and mental health research [1–3]. More
recently, researchers have demonstrated the alliance’s potential to
augment outcomes in cardiac [4] and musculoskeletal rehabilita-
tion [4,5], diabetes management [6], and chronic pain [7,8]. Within
brain injury rehabilitation, the therapeutic alliance has been linked
to enhanced functional status [9,10], productivity [10–15], self-
awareness [16], and emotional regulation [17]. These findings add
weight to the argument that it is not simply what we do with our
clients, but the way we engage with them that may mediate
rehabilitation outcomes [18–20].
The origins of the “therapeutic alliance” or “working alliance”
lie in psychodynamic theory, where it has been defined as the
degree in which the client–therapist partnership is engaged in
collaborative, purposeful work [21]. Bordin [21] proposed that the
alliance is comprised of the interpersonal bond between clinician
and client, and their level of agreement around the goals and
tasks of therapy. As the alliance supports and reflects the work of
therapy, careful attention needs to be paid to the development,
“therapeutic alliance” is a technical term, but it expands upon and
recognises the value in, a concept that is familiar to all health
professionals: the relationship between clinician and client that
acts as a vehicle to facilitate (or impede) treatment success
and recovery.
Bordin’s [21] theoretical framework underpins the majority of
studies evaluating the alliance and its effectiveness in brain injury
rehabilitation [9,10,13,17,23,24]. However, there is a paucity of
research examining whether psychotherapy-based alliance con-
cepts and measures actually fit brain injury rehabilitation.
Rehabilitation explicitly recognises individuals and their families as
integral members of the team [25], and recommends that clini-
cians share decision-making, knowledge, and expertise with family
to promote continuity of care and ongoing adjustment to disabil-
ity [26–29]. Such studies have concentrated on the team–family
(rather than team–client–family) partnership, so there remains a
lack of clarity around how clinicians might involve family within
their therapeutic alliances with their clients. For clients, it is plaus-
ible that the nature of their neurological difficulties and the
rehabilitation environment may necessitate a different type of

CONTACT Megan Bishop megan.bishop@otago.ac.nz School of Physiotherapy, Wellington School of Medicine, University of Otago, P.O. Box 7343, Wellington
South 6140, New Zealand
ß 2019 Informa UK Limited, trading as Taylor & Francis Group
2 M. BISHOP ET AL.
alliance compared to those engaged in counselling-based therapy.
Accruing information around therapeutic alliance priorities in
stroke rehabilitation is important as some people with brain inju-
ries have indicated that the quality of the clinical partnership is
more important than the content of therapy or its outcomes [30]
and affects their perceptions of rehabilitation success [17,31]. For
rehabilitation clinicians, understanding what contributes to and
influences the alliance could assist in illuminating the competen-
cies and attributes required to create, maintain, and repair this
seemingly important and valued therapeutic process. Certain pro-
fessional skills may augment the alliance. Neuropsychologists
have been proposed as more frequently able to form “good” or
“excellent” alliances, based on self-ratings, compared to physio-
therapists [15]. This may be explained by their expertise in sup-
porting clients to circumnavigate deficits in memory, insight, and
flexibility of thought, which are perceived to be particularly prob-
lematic in forming therapeutic alliances [32]. Alliance breakdowns
may stem from neuro-behavioural difficulties and the clinician’s
competencies. For example, video-recorded interactions between
people with brain injuries and speech–language therapy students
highlighted that the students’ inconsistent and inexperienced
responses to displays of sexually disinhibited behaviour contrib-
uted to these breakdowns [33]. However, empirical evidence has
demonstrated only weak correlations between a range of scores
on cognitive tests and the alliance, suggesting that a good alli-
ance may be feasible even for clients with cognitive deficits [24].
This research aimed to identify and explore the core components
of a therapeutic alliance from the perspectives of people who
have experienced a stroke and rehabilitation clinicians, and the
factors that may help or hinder these alliances.
Methods
Design
Interpretive description is an inductive exploration of a phenom-
enon (the therapeutic alliance) with the aim of illuminating pat-
terns and themes while accounting for differences between
individual perspectives [34]. As a result, it has been shown to cap-
ture the complexities and nuances associated with healthcare
[34]. With philosophical roots in naturalism, interpretive descrip-
tion is underpinned by the assumption that human experience is
constructed, complex, and context dependent; the researcher and
participant interact and influence each other, and that pre-exist-
ing theory is unlikely to account for the variations of the phenom-
enon under study [34,35]. Interpretive description was originally
developed for nursing research as a non-categorical, applied
methodology with a primary focus on developing practice
insights and solving complex health problems [36]. To that end, it
draws on methods from other methodological traditions such as
grounded theory and phenomenology where necessary to main-
tain the pragmatic utility of findings and to critically explore
anticipated and unanticipated differences within clinically relevant
phenomenon [37]. While other qualitative methodologies fre-
quently draw on a specific philosophical lens to inform the
approach and support their interpretation of data, interpretive
description aims to explore phenomenon in its natural state
[36,38]. A unique feature of interpretive description is that it
builds a theoretical scaffold from existing substantive theory and
evidence, as well as disciplinary knowledge [39], with a focus on
producing findings that have relevance, meaning, and utility to
augment practice knowledge [36,37]. The study received ethical
approval from the Auckland University of Technology’s Ethics
Committee (reference 12/237).
Participants
Participants included both clients admitted to and clinicians work-
ing in a regional, inpatient rehabilitation unit in New Zealand.
Client participants were eligible if they: (a) had a stroke diagnosis,
(b) had received a minimum of at least two weeks of inpatient
rehabilitation to ensure that they had opportunity to work with
several staff, (c) were able to communicate in English without the
support of an interpreter, and (d) were over the age of 18 and
able to provide informed consent. They were excluded if they had
underlying cognitive issues such as dementia if that would pre-
clude their meaningful participation in an interview and/or if their
stroke was not their primary reason for admission. Clinician partic-
ipants were eligible if they had worked in rehabilitation for at
least one year with the intention that this would allow sufficient
time to have consolidated their technical skills and be able to
reflect on the nature of their relationship with clients (their thera-
peutic alliances).
Sampling and recruitment
Members of the rehabilitation team referred prospective client
participants who met the eligibility criteria to the research team.
All potential participants received study information tailored to
allow those with cognitive and communication needs to make an
informed decision to take part or not. From there, client partici-
pants were purposefully selected to ensure diversity in type and
severity of impairments, ethnicity, and rehabilitation duration. As
the research progressed, theoretical sampling, a feature of
grounded theory [40], enabled exploration of preliminary ideas
and helped to refine interpretation of data. For example, concur-
rent data collection and analysis highlighted the dearth of rela-
tionship experiences with social workers; therefore the remaining
two client participants were recruited who had received substan-
tive social work input to further explore these perceptions. We
aimed to recruit ten client participants as this was considered suf-
ficient in providing a meaningful portrayal of the alliance and
enabled adequate diversity within the sample [39]. Consistent
with interpretive description, we drew on the concept of
“information power” [41] to inform sample size decisions. As such,
sample size (and consequent information power) was regularly
reviewed by examining the specificity of our sample and the qual-
ity of the data collected in relation to our study’s aims and ana-
lysis procedures [41].
For clinician participants, the study was advertised on the unit
and interested clinicians contacted the primary researcher directly.
Purposeful sampling sought diversity in discipline, clinical experi-
ence, ethnicity, and gender.
All those who chose to take part provided written consent. In
accordance with ethical and legislative requirements, study partici-
pation was voluntary, and privacy and confidentiality maintained.
Data collection
Semi-structured individual and focus group interviews following a
topic guide provided the primary source of data (see Table 1 for
examples of interview questions).
Client participants were individually interviewed in order to
accommodate their stroke-related needs and preferred level of
family/wh
anau (refers to the wider family network in Te Reo
M
aori, New Zealand’s indigenous language) involvement and
ranged from 34 to 84 min. One focus group interview (co-facili-
tated by MB and NK) was held with clinician participants, lasting
82 min. The lead author (MB) was the primary facilitator with NK

Table 1. Examples of interview questions.
Participant
group Sample questions
Client Can you think of a therapist or nurse with whom you have a
participants particularly close/more challenging relationship? Tell me
about this relationship
What are the key things that make this a strong relationship?/
Why do you think it is more difficult to form a relationship
with this person?
How does this relationship impact on you?
Has the relationship stayed the same throughout your
rehabilitation stay? (What factors have caused it to change?)
What have you done to contribute to this relationship?
Clinician What do you think are the core components of a therapeutic
participants relationship?
What skills or qualities are required to form these
relationships?
Tell me about relationships with clients that are more
challenging to establish or maintain.
How do families influence your therapeutic relationship with a
client?
How do ward based routines and requirements affect the
relationships you have with your clients?
primarily taking a support role as observer/note-taker. The term
“therapeutic relationship” was used during the interviews, as
“therapeutic alliance” is a jargon with specific theoretical roots in
psychotherapy that may not have been easily understood by par-
ticipants. Further, a review of the literature highlighted that an in-
depth conceptual exploration of this topic had not been under-
taken in stroke rehabilitation before, so the research team wanted
to remain open to the possibility of uncovering important con-
cepts that may or may not align with “therapeutic alliance” as it is
currently understood. Consequently the broader term “therapeutic
relationship” will be used for the remainder of this article.
Interviews for both sets of participants were audio-recorded
and transcribed for analysis. Field notes accompanied each inter-
view. Post-interview reflections shaped subsequent interviews,
promoting an iterative approach to data collection and analysis
[36,38]. Demographic and relevant health information was
obtained from both the client participants directly and their med-
ical files. Functional independence measure (FIM) scores captured
participants’ motor, cognitive, and social skills across a range of
day-to-day behaviour [42]; this scale ranges from 18 (lowest) to
126 (highest). Within inpatient rehabilitation, the FIM has excellent
internal consistency (Cronbach’s alpha of 0.93 for admissions and
0.95 for discharges) [42]. Further, there is excellent concurrent val-
idity between the motor-FIM and the Barthel Index (Spearman
correlation ¼ 0.95) [43]. Clinician participants completed a brief
questionnaire to capture their demographic and professional
details. Years of clinical experience are divided into bands in the
results section to maintain participant confidentiality.
Data analysis
Data were analysed using conventional content analysis as
described by Hsieh and Shannon [44]. Whole interview transcripts
were repeatedly reviewed and pertinent phrases manually coded
across three categories: “key components”, “influencing factors”,
and “family views”. The latter category was included so that data
contributed by family members could be considered separately.
However, given our intention to give primacy to the clients’ voice,
family input was only included when they were augmenting
understanding of a point made by their loved one. Wherever pos-
sible, codes remained in the participants’ own words or were
derived from a list of common codes formed inductively [37].
THERAPEUTIC ALLIANCE IN STROKE REHABILITATION 3
Within each transcript, the categories were examined separately,
and codes checked against each other and the raw data. A
“memo” column was added alongside the categories to enable
concurrent documentation of thoughts triggered during the cod-
ing process, and to document data that did not appear to initially
correspond to the existing categories but was potentially valuable
[39]. The transcripts and an initial coding framework were then
exported into NVivo 10 Computer Software Package [45]. Data
were coded to the coding framework in NVivo, with new codes
and categories created inductively by analysing data across partic-
ipants [44]. Patterns were identified as well as themes and rela-
tionships between data sources, with initial ideas repeatedly
assessed and challenged [36]. This involved continuously moving
between raw and coded data, analytical recordings, and relevant
literature [39], alongside discussions with the research team.
Diagramming, memos, and written summaries supported the con-
ceptualisation of key messages [38,39].
Rigour
Widely accepted strategies to enhance rigour were applied
throughout data collection and analysis [46]. A transcribed pre-
supposition interview acknowledged the primary researcher’s
assumptions and understandings around the therapeutic alliance
concept and its potential to influence data collection and analysis
[46,47]. The interview guide was reviewed by research supervisors
and rehabilitation clinicians, and a practise interview was carried
out with a skilled qualitative interviewer who provided valuable,
independent feedback on terminology and interview structure.
Purposeful and theoretical sampling techniques and a prolonged
period of data collection yielded rich, descriptive data and pro-
moted credibility and transferability of findings [46]. Auditability
was enhanced via the collation of interview field notes and tran-
scripts, demographical information, coding records, and written
reflections [46].The research team supported robust interpretation
of data by: (a) each coding two interview transcripts, (b) examin-
ing the NVivo-derived summary of categories, (c) reviewing the
initial findings which contained high levels of raw data, and (d)
engaging in regular critical discussions around emerging thematic
development. Credibility was further improved by sharing prelim-
inary findings with consenting client and clinician participants to
ensure resonance with the sample from which they were
derived [36,46].
Results
Ten out of 12 “client” participants contacted, and seven self-refer-
ring “clinician” participants took part in the study. Clients were
aged between 29 and 76 years, had experienced a range of
stroke-related difficulties, and spent between 2 and 13 weeks in
rehabilitation at the time of their interview. Clinicians included
registered nurses, social workers, a speech–language therapist,
and an occupational therapist; a range of ethnicities and years of
clinical experience were represented within this group.
Participants’ demographic details are outlined in Tables 2 and 3.
Pseudonyms are used when participants are quoted.
Everyone is different
The data emphasised the importance of considering how everyone
is different in building therapeutic relationships, as clients viewed
their therapeutic relationships in different ways. These diverse
perceptions could be synthesised into model consisting of the
4 M. BISHOP ET AL.

Table 2. Client participant characteristics.

Demographic variable Category n
Type of impairment(s) (n > 10 as some had Physical/sensory 9
multiple impairments) Cognitive 8
Communication 4
Severity of impairments based on FIM scores >107 4
(higher scores denote increased 72–107 4
independence) <72 2
Rehabilitation duration <1 month 3
1–2 months 5
2–3 months 1
>3 months 1
Ethnicity (n > 10 as some identified with NZ European 7
multiple ethnic groups) M
aori 3
Pacific Islander 2
Age <45 years 2
45–65 years 4
>65 years 4

Table 3. Clinician participant characteristics.

Demographic variable Category
Clinical experience <5 years
5–10 years
11–20 years
>20 years
Profession Registered nurses (RN)
Social workers (SW)
Speech language therapists (SLT)
Occupational therapists (OT)
Gender Female
Male
Ethnicity NZ European
M
aori
Pacific Islander
British/European
following core components: (a) a personal connection, (b) a pro-
fessional collaboration, and (c) family/wh
anau collaboration.
Participants prioritised each component differently depending
on their needs and preferences. These priorities could change
during their inpatient stay. Consequently, there appeared a need
for relationship plasticity to accommodate each client’s unique
and fluid priorities. Figure 1 attempts to illustrate the complexity
of these relationships through the overlapping circles, as well as
signifying the conceptual overlap between components. Inside
each circle are the skills and attributes that appeared important
for each component.
A personal connection
Connecting at a personal level appeared to be a central relation-
ship feature for a number of clients. Clinicians perceived that this
core component was operationalised by “being yourself” (Vailea,
RN, 11–20 years’ clinical experience). Clients often expressed simi-
lar views: “It’s being real, as a human and a therapist” (David, FIM:
118). This seemed to help participants understand and respond to
the person behind the stroke or uniform, creating an emotional
co-investment in each other. While many clients valued this core
component, others held more of a detached outlook. Focusing on
the therapeutic work and the “escape” home seemed to be more
of a priority for such participants:
Because the goal of everyone should be to get out as quickly as
possible. Hopefully they’ve got their own friends. Yeah. And don’t need
to make friends on the inside, ’cos that’s what it’s like, it’s like a prison
[laughter] (Toby, client, FIM: 71)
For those who valued a personal connection, it seemed to be
augmented by a clinician’s empathy and responsiveness, and
n
1
2
1 Figure 1. Core components of a therapeutic relationship.
3
3
2
1 reciprocity. Empathy and responsiveness often manifested
1 through a clinician’s sensitivity to their client’s emotional turmoil
7 and vulnerabilities associated with their stroke and inpatient stay.
0
1 Andy (client, FIM: 95) felt “totally at ease revealing myself” as a
2 result of his physiotherapist’s understanding, non-judgmental
2 manner when he became upset. This contrasted with his reserve
2 around his social worker – part of whose role involved counselling
– as he perceived her to be a hospital “official”. Andy appeared to
prioritise a clinician’s personal qualities above their professional
skills in this situation.
Being personally vested in each other appeared to involve a
degree of reciprocity. Reciprocity was sometimes reflected at an
emotional level. One client’s bond flourished with a nurse who
shared his distress during an adverse event, illustrating some cli-
ent’s heightened awareness of their clinician’s views and emo-
tions, and the value placed on complementary behavioural
responses. Shared humour was also appreciated by many clients.
For some, it played an important role in lifting their mood: “I
needed to be able to laugh, otherwise I would have done a lot of
crying” (Andy, client, FIM: 95). As such, humour seemed to pro-
vide a valuable rehabilitation survival tool promoting hope for
those who valued it.
Huia (client, FIM: 114) perceived the reciprocal sharing of one-
self to be an important indicator of relationship quality:
It’s in the korero [story/discussion] we share, you give a bit of yourself,
somebody gives a bit of themselves, you know? And I think that’s the
gauge to the relationship and the rapport and I think it’s very, very
special in the healing.
A clinician’s self-disclosure of personal information, such as
sharing family photos and reports of weekend activities, was often
perceived to dismantle expert–patient barriers and provide
“confirmation that life goes on” (Andy, client, FIM: 95). Some clini-
cians purposefully disclosed personal information to reduce the
power imbalance in particular situations: “But there’s definitely a
shower conversation, you have to share, the person’s very vulner-
able, it’s not appropriate, you-them. There’s definitely a sharing.
Trying to put them at ease” (Kerry, OT, 5–10 years’ clinical
experience).

Despite the important role that self-disclosure seemed to play
in fostering a close personal connection, the need for boundaries
was acknowledged by many participants. Clinicians identified a
range of factors that influenced their level of personal sharing,
including the client, the situation, and their own experience, skill,
and intuition. Clients recognised the need for boundaries too: “If
they’re going to be good nurses, if they’re going to be caring
nurses and … efficient, they’ve got to draw, there’s got to be a
line. They can’t spend all their time chatting to people” (Andy, cli-
ent, FIM: 95).
A professional collaboration
A strong professional collaboration appeared to be based on cre-
ating a mutual understanding of the client’s personal and clinical
contexts, and striving towards a shared focus. Establishing the
right “type” of professional collaboration seemed important as cli-
ents often held specific views around their preferred degree of
power and involvement in these partnerships. Some preferred the
“experts” to assume responsibility, while others favoured leading
their programme with relevant guidance from clinicians: “The
physio fits in and directs and guides and suggests multiple paths
and more effective paths, but it’s ultimately down to the patients
to … be able to indicate” (Toby, client, FIM: 71). Participants
noted the influence of neurological factors on relationship partici-
pation. Clinicians discussed utilising their skills or recruiting input
from family members, to ensure that their relationships with cli-
ents with severe impairments were both productive and personal-
ised. In such cases, collaborating with family/wh
anau (see below)
sometimes became a key focus. Some clients noted their relation-
ship capabilities or preferences changed over their rehabilitation
period. For example, as David’s (FIM: 118) confidence in his com-
munication skills improved, so too did his perceived ability to par-
ticipate in therapeutic relationships. Responding to these varying
degrees of collaborative predilections for each individual and over
time were perceived to help maintain a therapeutic relationship.
A professional collaboration appeared to develop from a
mutual understanding of the client’s personal and clinical con-
texts. A mutual understanding could then enable technical skills
and expertise to dovetail with a client’s needs and preferences.
For several clients, this led to the development of a shared focus
and optimal relationship:
It meant that they were listening. That’s that connection. It wasn’t just
getting pulled out of the sky and saying “this is the best for you,
because this has what’s happened to you.” No way. When you’re
included in the solution and are able to participate in the solution, I
think that’s a great thing (Huia, client, FIM: 114)
Clinicians were aware that facilitating goal attainment, such as
helping clients to walk again, could enhance the therapeutic con-
nection. Yet for many clients, the creation of a mutually agreed,
transparent rehabilitation blueprint – rather than goal achievement
itself – was enough to develop a constructive partnership: “I’ve a
close working relationship with both … the physio and … the
OT … um [PT’s name]’s is certainly more I would say productive,
purely because it’s easier to see the plan” (Toby, client, FIM: 71).
A common aim appeared to provide clients with windows of
hope and tangible evidence that their clinicians were working
with them to achieve what mattered most.
Family/wha
nau collaboration
Building optimal family/wh
anau collaborations seemed to require
consideration of the various roles, dynamics, and views of family
members. These factors had the potential to impact on the clini-
cian–client therapeutic relationship. Clinicians often prioritised a
THERAPEUTIC ALLIANCE IN STROKE REHABILITATION 5
family-centric approach due to its potential rehabilitation and
therapeutic relationship benefits. In contrast, a client’s wish or
need to involve family was revealed in a variety of responses. Six
of the 10 client participants valued clinicians who collaborated
with their family/wh
anau. Heke (client, FIM: 98) communicated
that the involvement of his wife was the most important constitu-
ent of his therapeutic relationships, as “everything goes back to
my wife”. However, forming an exclusive therapeutic relationship
with their clinician appeared a priority for other clients: “I am the
main conduit and everything goes through me … I do know what
works for me … so it’s better that they [my family] are left well
alone … they are only a support team for me” (Toby, client,
FIM: 71).
The roles, dynamics, and views of family seemed to influence
participants’ preferences around this core component. A number
of participants highlighted the central role loved ones may play
in a client’s life, including as principal decisions makers, rehabilita-
tion partners, and advocates. Identifying and involving relevant
family within these roles was often perceived to be an essential
component of the therapeutic relationship: “Ask or clarify early in
the piece, especially when working with M
aori: ‘who are the key
decision makers? Who should we speak with?’ Because in my
world … they’re wh
anau and they are very key to everything”
(Huia, client, FIM: 114).
Some family members were considered important rehabilita-
tion partners who could assist with skill development and provide
holistic and enduring support for the client: “that interaction of
that family with that patient, is going to be in an ongoing way
crucial, so our relationship with the family is also important to
build” (Susan, SW, >20 years’ clinical experience). Through their
advocacy role and sharing of client-specific information, it
appeared that family could bring out the human factor for their
loved ones. This seemed particularly salient for clients with
marked cognitive and communication difficulties. Family advo-
cates sometimes helped clinicians to reprioritise and build stron-
ger personal connections with clients, rather than being consumed
by a utilitarian approach in managing the day-to-day operations
and processes. Strong relationships between the team and family
had potential to improve a client’s engagement in their relation-
ships with staff, through the delivery of consistent messages:
“Wh
anau also supported me to stay … so it was everybody talking
to me really … so I wouldn’t rebuff against it, you know, keeping
that engagement alive and trusting” (Huia, client, FIM: 114). These
findings suggest that some family members may also enhance
the clinician–client professional collaboration by helping to align
views and expectations.
Conversely, data indicated that adverse family dynamics and
divergent views could affect relationships within the clinician–-
client–family triad. Clinicians were aware that intra-family conflict
sometimes caused clients to behave differently when loved ones
were around. This could influence how well a client engaged in
their therapeutic relationships and rehabilitation. Prior to collabo-
rating with families, David (client, FIM: 118) suggested that clini-
cians explore any “uncomfortable interplay” between members, to
ensure optimal collaboration with the most relevant loved ones.
Relationship disruptions
More than half of the clients interviewed conveyed specific or
cumulative experiences that led to a marked deterioration in at
least one therapeutic relationship. These relationship disruptions
were characterised in different ways. Two clients recalled an overt
confrontation whereby the clinician was presumably aware that a
6 M. BISHOP ET AL.
relationship disruption had occurred. In contrast, the remaining
client participants described their weakening relations by their
adverse but potentially suppressed emotional reactions: “It’s hurt
my feelings … I didn’t say anything, but … I’m nearly saying some-
thing to her” (Masina, client, FIM: 78). Some clinicians appeared to
be tuned into these subtle expressions of relationship dissatisfac-
tion whereas others seemed less observant or rationalised such
behaviour in different ways. Huia (FIM: 114) reported that her
clinicians assumed that she was depressed after she disengaged
from her rehabilitation and relationships. In reality, Huia’s with-
drawal stemmed from her M
aori values and beliefs being compro-
mised. Her experience outlines the potential for clinicians to
attribute “resistant” or “disengaged” behaviour to non-relationship
factors when it may be a response to how clinical staff act or
speak about things that greatly matter to clients.
Two main factors appeared to contribute to relationship dis-
ruptions: erroneous assumptions and/or a perceived lack of
responsiveness to one’s needs and preferences. These factors had
the potential to impact on the integrity of the core components
of a therapeutic relationship. One client recalled a clinician who
shared a lengthy story about her dog, potentially at the expense
of delivering care: “You could see with her that she was more
looking in the mirror at herself … and nothing with my require-
ments were met there” (Heke, client, FIM: 98). Similar sentiments
were expressed about clinicians who shared superfluous personal
details, such as their age. These clinicians may have compromised
the therapeutic relationship by misinterpreting what was needed
to establish the right level of personal connection with a person.
Alternatively, the type and level of information disclosed may
have detracted from, rather than enhanced, the per-
sonal connection.
Professional collaborations sometimes suffered when a client’s
preferred degree of collaboration or rehabilitation focus was
not respected:
I didn’t blow up for no good reason and perhaps [PT’s name] was
labouring under one set of goals, one set of results, but you’ve got to
understand I was in the process of saying “no, we are shifting the goal
posts” (Toby, client, FIM: 71).
Some clients reported discrepancies in desired outcomes for a
task, such as a washing and dressing assessment, where clinicians
appeared to adhere to their own ideals or pre-set performance
criteria, rather than stopping and considering their individual situ-
ation. Clinicians agreed that various factors may contribute a task-
focused and assumption-laden approach. These included the
need for a coping strategy for managing difficult clientele and
adherence to professional rules and regulations.
The data indicated that assumptions around the role of family
members endangered some therapeutic relationships. In one cli-
ent’s case, clinicians did not appear to be collaborating with the
most relevant members of her wh
anau:
… what’s really important when working with wh
anau – I can’t help
but stress – is get them to find out who are the key decision makers
’cos that’s how it can get messy, you know? I think, what I’ve been
experienced is they’ve been involved but I think some assumptions
were made in my case, you know? When there’s a number of wh
anau,
ask wh
anau, you know? It mightn’t be the only ones that are turning
up that are the key, they are only the soldiers – foot soldiers … (Huia,
client, FIM: 114)
Following a relationship disruption, some clients experienced a
relationship breakdown whereby they terminated their relation-
ship, actively distanced themselves physically or emotionally from
their clinician, or declined therapeutic input. All relationship
breakdowns appeared to adversely affect the client’s mood or
emotions. Andy (client, FIM: 95) felt “terrified, you know I am still
frightened of that woman” after an interaction with a nurse
whom he had not met before. He indicated that she, or anyone
else, had failed to discuss or repair their damaged relations.
Consequently, he ended his relationship with her by seeking her
removal from his care. Another client experienced a similar situ-
ation, again when the therapeutic relationship was in its infancy.
This suggests that some new relationships may have insufficient
foundations or lack core components required to withstand such
transgressions. In these examples the rationale for, and subse-
quent management of, each clinician’s behaviour appeared
absent, which may also have affected outcomes.
A surprise finding was that not all relationship disruptions
appeared detrimental. Connections could be preserved or even
strengthened if a relationship disruption was therapeutically man-
aged and co-existed with a robust relationship, as Toby (client,
FIM: 71) experienced: “I didn’t blow up for no good reason … but,
I mean, yeah in the end, we got there and it’s [the relationship]
better”. Toby reported an instant connection with his physiother-
apist and had worked with her for several months before the
“blow up” happened. His respect for her knowledge and skills was
evident, all of which may have created a safe platform for him to
vent his frustrations and pave the way for a more collaborative
partnership:
I think in the process of doing that … I would say we’d go, OK well I
can say what I want and it’s going to be listened to … all I wanted to
do was make sure that we are on the same page and we’re working
together in the same way and the next day I just felt a little bit more
empowered (Toby, client, FIM: 71).
Discussion
This study provided a detailed exploration of the core compo-
nents of a therapeutic relationship in stroke rehabilitation. Two
themes surfaced from the interviews of clients with stroke and
members of the clinical team. The first theme, everyone is differ-
ent, proposed three core relationship components that client par-
ticipants valued to varying degrees: a personal connection, a
professional collaboration, and family/wh
anau collaboration. The
second theme, relationship disruptions, suggested that erroneous
assumptions or a perceived lack of clinical responsiveness contrib-
uted to relationship disruptions. The strength of the pre-existing
relationship and active efforts in a repair process appeared to
impact on whether these relationships deteriorated to the point
of breakdown, or recovered. This research challenges existing alli-
ance theory and assumptions, and raises a number of implications
for rehabilitation practice.
Client participants’ therapeutic relationship preferences were
individual and nuanced. This inferred that an optimal relationship
did not necessarily require all core components to be strong, but
rather, each component needed to be emphasised, or balanced,
according to the person’s priorities. Priorities could change over
time, indicating a need for relationship plasticity within and
between clients. These findings are in contrast to Bordin’s theory
of working alliance [21], where each dimension (interpersonal
bond, tasks, and goals) was argued to be equally important to
the alliance. Further, existing alliance theory suggests that the
type of alliance required will depend on the genre of psychother-
apy [21]. In this sense, the therapeutic approach taken appears
most formative to the alliance (a more clinician-directed, tech-
nical, disciplinary-based approach). Interpretation of the current
study’s findings suggest the unique and specific needs and prefer-
ences of the client to be most formative to the development of

the therapeutic alliance. Rather than making assumptions, a key
responsibility for clinicians was to consider each client’s relation-
ship needs and preferences first and adapt how they worked with
each individual and potentially their family/wh
anau. This is con-
sistent with Leplege et al.’s [48] view of person-centredness as an
individualised and holistic approach that acknowledges variability
within and between individuals. Contextualising the “right” thera-
peutic relationship appeared to require an accurate appraisal of
each person and each situation, so that clinicians could tailor their
personal resources and clinical skills accordingly. Such skills
extended beyond technical abilities to include inter-personal and
emotional competencies similar to that described by McCormack
et al. [49].
This research proposed that family/wha
nau collaborations were
sometimes an essential element of therapeutic relationships. This
core component is notably absent from Bordin’s [21] theory of
working alliance. Other psychotherapy researchers categorise the
family system as an extra-therapeutic factor [50], where this
research identified them as potentially a core and indeed, signifi-
cant factor within therapeutic relationships. These different con-
ceptualisations of the alliance may be explained by the unique
needs of the population and context. For example, engaging with
families may be important in stroke rehabilitation [25,28,51] but
less pertinent in certain counselling-based interactions. Given
these different alliance conceptualisations and approaches, stroke
rehabilitation would benefit from a theoretical framework and
measures that are more contextually relevant.
A clinician’s disclosure of personal information seemed to play
an important role in augmenting therapeutic relationships pro-
vided they were responsive to one’s relationship and care needs.
These disclosures had the potential to reduce expert-patient bar-
riers and promote comfort, cultural responsiveness, and an emo-
tional investment in each other. In contrast, debate exists in the
psychotherapy literature around the benefits, ethics, and risks
associated with clinician-led disclosures as these have the poten-
tial to breach intimacy/sexual boundaries [52,53], create a role
reversal where the client feels obliged to “manage” the relation-
ship [52,53], contaminate the therapy process [54,55], or cause an
alliance/relationship rupture [53,56]. Many of these issues were
not identified in our study, which may reflect the different health
conditions, therapy contexts and subsequent therapeutic relation-
ship needs. All client participants appeared to require a degree of
human connection with their clinicians to promote comfort and
trust in their relationships. Appropriate self-disclosures could facili-
tate this process and provide a welcome distraction from their dif-
ficulties – the same distraction that may be considered a
hindrance in psychotherapy [55]. Evidence in this study and
others suggests that non-disclosure of personal information may
interfere with the formation of a therapeutic connection [53, 57].
It could be posited that failure to disclose any personal informa-
tion may be ethically unsound if the client requires this to actively
engage in their recovery.
Professional boundaries are considered to be important in
guiding a clinician’s disclosure of personal information [52,53,58],
yet clinician participants in our study often relied on intuition and
experience to establish appropriate boundaries. Community
speech–language therapists similarly identified that negotiating
boundaries are done implicitly and are difficult to explain [59].
Prior to applying self-disclosures, clinicians may wish to contem-
plate in advance what information they feel personally and pro-
fessionally comfortable sharing with their clients, and in what
situations. This may pre-empt instances where clinicians suddenly
find themselves at a decision-making juncture, such as a client
THERAPEUTIC ALLIANCE IN STROKE REHABILITATION 7
asking them about their personal lives. Researchers suggest that
clinicians with clear beliefs and rationale for disclosing are less
likely to impede the relationship or therapy when they share
some personal information [54]. Newly qualified staff may benefit
from training highlighting the potential risks and benefits of dis-
closures through facilitated discussions around common clinical
scenarios. Such training may improve clinicians’ reasoning, com-
fort levels, and professional safety when expressing their personal
identities. Reconsidering how professional boundaries are concep-
tualised may also be useful. Researchers have argued that the
term “boundary” indicates a rigid, pre-determined demarcation
separating out ethical from unethical behaviour, which does not
reflect the context-dependent nature of therapeutic relationship
boundaries in practice [60]. These researchers proposed
“therapeutic territory” as an alternative metaphor as it denotes a
shared space that clients and clinicians may enter and monitor.
Adapting therapeutic relationship “territories” to the individual’s
needs and preferences is congruent with the present
study’s findings.
A clinician’s behaviour following a relationship disruption
impacted on how clients viewed these relationships. Regaining
relationship health appeared to require reparation of whichever
core relationship component(s) had been compromised. For one
participant, this involved re-negotiating power levels and goals in
order to repair the compromised professional collaboration com-
ponent. These findings are supported by research that encourages
therapists to identify and address whichever relationship ele-
ment(s) has ruptured [61,62]. Addressing relationship discontent
may be particularly pertinent in the early phases of relationship
formation. A couple of participants experienced relationship
breakdowns with clinicians they had only recently met, with one
requesting that the staff member be removed from his care.
Other studies illustrate that some clients choose to end their clin-
ical partnership based on their initial impressions of their clinician
[63,64]. Elkin et al. [63] demonstrated that a clinician’s responsive-
ness and promotion of a positive therapeutic atmosphere within
the first two sessions is predictive of the clients’ perceptions of
the relationship and engagement in therapy. Relationship devel-
opment, review, and maintenance may need to be integrated
from the very first interaction(s) in order to facilitate continued
relationship and rehabilitation engagement.
This study has provided original insights around the challenges
that stroke clinicians may experience in identifying that a relation-
ship has been compromised – a pre-requisite for the reparation
process [65]. Several of our study’s client participants suppressed
their concerns and withdrew from the clinician or elements of
their rehabilitation, and some even declined care. Some perceived
that their dissatisfaction was erroneously attributed to a stroke-
related problem, such as fatigue or depression. This suggests a
risk that clinicians may sometimes misjudge or minimise their cli-
ent’s relationship experiences. Existing brain injury research may
implicitly encourage clinicians to conceptualise a rehabilitation or
relationship engagement issue as a neurological one. Qualitative
researchers report that a client’s neuropsychological, communica-
tion, or motivational difficulties are the main barriers to forming
or maintaining a strong alliance/relationship [32,33,66,67].
Quantitative studies have predominantly focused their investiga-
tions on client-related factors that may influence rehabilitation
compliance and/or alliance quality, such as injury variables [16],
metacognitive or cognitive skills [10,16,24], or emotional distress
[10]. Such aims and methods perhaps assume that the client is
the “common factor” impacting on, if not impeding, relationship
strength, with little consideration given to the role of the clinician.
8 M. BISHOP ET AL.
The present study’s design differed from previous studies examin-
ing therapeutic relationship quality or disharmony in that it also
explored client perspectives of the challenges they experienced
when engaging with clinicians across different professions. Our
findings support the view that clinicians play a vital role in client
engagement via the therapeutic relationship [19,68,69]. For clini-
cians in this field, avoiding making assumptions when clients
decline or disengage from their therapy or relationships may be
an important first step. Mindfully listening and observing client
behaviour and undertaking intermittent “relationship health
checks” to identify any issues or changing relationship priorities,
may also be important in ensuring that relationships are both co-
constructed and constructive.
Study limitations
There are limitations to this research that require acknowledg-
ment. Participants were recruited from one rehabilitation unit, lim-
iting the transferability of these findings to other rehabilitation
units with different client and clinician characteristics and service
idiosyncrasies. Interviews occurred at one time point for each cli-
ent, so the data are unlikely to have detected all the relationship
nuances over time, particularly for those with memory deficits.
Findings suggested that several participants had difficulty recall-
ing their initial interactions with staff when they entered rehabili-
tation many weeks earlier, whereas others had only recently
started working with their rehabilitation clinicians. As a result, it is
difficult to determine whether there were common relationship
patterns over an entire rehabilitation journey and if/how these
were affected by pivotal events such as discharge planning or key
staff turnover. Interviews were not separated into client–clinician
dyads due to the likely methodological challenges associated with
identifying and recruiting the most relevant clinician(s) for each
client. This is likely to have oriented the data and subsequent
findings to primarily reflect the clients’ relationship perceptions.
For example, the “culpable” clinician’s perspective of a relation-
ship disruption reported by a client was not elicited and may
have provided further contextual information. There were no eli-
gible male clinicians working on the rehabilitation unit at the
time of the focus group interview or member checking discussion.
However, none of the participants discussed the role of gender in
relation to their therapeutic relationships and it did not appear to
be an influencing factor in the studies reviewed during this
research. Finally, many participants had difficulties separating their
therapeutic relationships perceptions from their general rehabilita-
tion experiences. Other researchers have similarly observed
that these perceptions are often difficult to separate [30].
Consequently, there is potential that the present findings also
incorporate perspectives on what makes rehabilitation successful
for some participants and less so for others.
Future research
Future studies could explore whether it is necessary for all mem-
bers of the team to have a strong or “good enough” therapeutic
relationship with the client to impact on outcomes, or whether
some relationships may be more important than others. This is of
interest given the high numbers and different characteristics of
therapeutic relationships that clients are likely to encounter over
their rehabilitation stay. Eliciting family/wh
anau perspectives of
these relationships may also address some of this study’s limita-
tions and gaps in research. There are some promising practice
guidelines and models proposed for engaging with families
following an acquired brain injury, for example, Foster et al. [26]
and Sohlberg et al. [28], but there appear to be few studies pro-
viding clarity on how to manage the complexities of this collabor-
ation. For example, future research could explore whether the
therapeutic relationship with the client is of prime importance
and needs preserving at all costs, or perhaps, whether a systems
approach may be more appropriate given the enduring nature of
stroke-related deficits and influencing role of some family in a cli-
ent’s life. Due to the qualitative nature of this inquiry, there were
no formalised assessments undertaken of relationship quality or
its impact on rehabilitation outcomes. As such, while there were
anecdotal reports regarding the benefits of a strong therapeutic
relationship, further research is needed to explore the association
with outcome. Finally, this research may provide the basis for the
development of a conceptually sound, valid and reliable relational
measure, which will further support research aiming to explore
the impact of the therapeutic relationship on outcome.
Conclusion
This research contributes new knowledge around how people
with stroke, and rehabilitation clinicians, experience therapeutic
relationships in a stroke rehabilitation service. It is novel in its
explicit exploration of the therapeutic alliance in the inpatient
stroke rehabilitation setting, and of the views of clinicians from a
range of disciplinary backgrounds and clients experiencing a
range of difficulties post stroke. The findings suggest that the per-
sonal connection, professional collaboration and family/wh
anau
collaboration appear the most salient core components of these
relationships. Clients may preferentially “weight” these compo-
nents differently, and each component may move to the relation-
ship foreground or background over the course of their inpatient
stay. Developing and maintaining therapeutic relationships
appears to require a person-centred approach and relationship
plasticity to accommodate each client’s unique and fluid priorities.
This may involve a level of emotional and intellectual awareness,
as well as an ability to use one’s personal attributes and profes-
sional competencies therapeutically. Given the value many partici-
pants placed on their relationships, determining and addressing
or even prioritising each client’s therapeutic relationship needs
and preferences may augment rehabilitation processes
and outcomes.
Acknowledgements
Thank you to the participants in this study for sharing their thera-
peutic relationship experiences.
Disclosure statement
No potential conflict of interest was reported by the authors.
ORCID
Megan Bishop http://orcid.org/0000-0001-9666-8299
Nicola Kayes http://orcid.org/0000-0002-2747-667X
Kathryn McPherson http://orcid.org/0000-0003-1240-8882
References
[1] Krupnick JL, Sotsky SM, Simmens S. The role of the thera-
peutic alliance in psychotherapy and pharmacotherapy out-
come: findings in the National Institute of Mental Health

Treatment of Depression Collaborative Research
Programme. J Consult Clin Psychol. 1996;64:532–539.
[2] Martin DJ, Garske JP, Davis MK. Relation of the therapeutic
alliance with outcome and other variables: a meta-analytic
review. J Consult Clin Psychol. 2000;68:438–450.
[3] Horvath AO, Symonds BD. Relation between working alli-
ance and outcome in psychotherapy: a meta-analysis. J
Couns Psychol. 1991;38:139–149.
[4] Hall AM, Ferreira PH, Maher CG, et al. The influence of the
therapist–patient relationship on treatment outcome in
physical rehabilitation: a systematic review. Phys Ther.
2010;90:1099–1110.
[5] Lewis M, Morley S, Windt DA, et al. Measuring practitioner/
therapist effects in randomised trials of low back pain and
neck pain interventions in primary care settings. Eur J Pain.
2010;14:1033–1039.
[6] Lee Y, Lin J. The effects of trust in physician on self-effi-
cacy, adherence and diabetes outcomes. Soc Sci Med.
2009;68:1060–1068.
[7] Fuentes J, Armijo-Olivo S, Funabashi M, et al. Enhanced
therapeutic alliance modulates pain intensity and muscle
pain sensitivity in patients with chronic low back pain: an
experimental controlled study. Phys Ther. 2014;94:477–489.
[8] Ferreira PH, Ferreira ML, Maher CG, et al. The therapeutic
alliance between clinicians and patients predicts outcome
in chronic low back pain. Phys Ther. 2013;93:470.
[9] Evans CC, Sherer M, Nakase-Richardson R, et al. Evaluation
of an interdisciplinary team intervention to improve thera-
peutic alliance in post-acute brain injury rehabilitation. J
Head Trauma Rehabil. 2008;23:329–338.
[10] Sherer M, Evans CC, Leverenz J, et al. Therapeutic alliance
in post-acute brain injury rehabilitation: predictors of
strength of alliance and impact of alliance on outcome.
Brain Inj. 2007;21:663–672.
[11] Klonoff PS, Lamb DG, Henderson SW, et al. Outcome
assessment after milieu-oriented rehabilitation: new consid-
erations. Arch Phys Med Rehabil. 1998;79:684–690.
[12] Klonoff PS, Lamb DG, Henderson SW. Outcomes from
milieu-based neurorehabilitation at up to 11 years post-dis-
charge. Brain Inj. 2001;15:413–428.
[13] Klonoff PS, Talley MC, Dawson LK, et al. The relationship of
cognitive retraining to neurological patients’ work and
school status. Brain Inj. 2007;21:1097–1107.
[14] Prigatano GP, Klonoff PS, Oʼ brien KP, et al. Productivity
after neuropsychologically oriented milieu rehabilitation. J
Head Trauma Rehabil. 1994;9:91–102.
[15] Scho €nberger M, Humle F, Zeeman P, et al. Working alliance
and patient compliance in brain injury rehabilitation and
their relation to psychosocial outcome. Neuropsychol
Rehabil. 2006;16:298–314.
[16] Schonberger M, Humle F, Teasdale TW. The development
of the therapeutic working alliance, patients’ awareness
and their compliance during the process of brain injury
rehabilitation. Brain Inj. 2006;20:445–454.
[17] Scho €nberger M, Humle F, Teasdale TW. Subjective outcome
of brain injury rehabilitation in relation to the therapeutic
working alliance, client compliance and awareness. Brain
Inj. 2006;20:1271–1282.
[18] Kayes NM, McPherson KM. Human technologies in rehabili-
tation: ‘Who’ and ‘How’ we are with our clients. Disabil
Rehabil. 2012;34:1907–1911.
THERAPEUTIC ALLIANCE IN STROKE REHABILITATION 9
[19] Bright FA, Kayes NM, Worrall L, et al. A conceptual review
of engagement in healthcare and rehabilitation. Disabil
Rehabil. 2015;37:643–654.
[20] Lawton M, Haddock G, Conroy P, et al. People with apha-
sia’s perception of the therapeutic alliance in aphasia
rehabilitation post stroke: a thematic analysis. Aphasiology.
2018;32:1397–1417.
[21] Bordin ES. The generalizability of the psychoanalytic con-
cept of the working alliance. Psychother: Theory, Res Prac.
1979;16:252–260.
[22] Hatcher RL, Barends AW. How a return to theory could
help alliance research. Psychother. 2006;43:292–299.
[23] Klonoff PS, Olson KC, Talley MC, et al. The relationship of
cognitive retraining to neurological patients’ driving status:
the role of process variables and compensation training.
Brain Inj. 2010;24:63–73.
[24] Scho €nberger M, Humle F, Teasdale TW. The relationship
between clients’ cognitive functioning and the therapeutic
working alliance in post-acute brain injury rehabilitation.
Brain Inj. 2007;21:825–836.
[25] Stroke Foundation of New Zealand, New Zealand
Guidelines Group. New Zealand clinical guidelines for
stroke management 2010. Wellington (NZ): Stroke
Foundation of New Zealand; 2011.
[26] Foster AM, Armstrong J, Buckley A, et al. Encouraging fam-
ily engagement in the rehabilitation process: a rehabilita-
tion provider’s development of support strategies for
family members of people with traumatic brain injury.
Disabil Rehabil. 2012;34:1855–1862.
[27] McLaughlin AM, Carey JL. The adversarial alliance: develop-
ing therapeutic relationships between families and the
team in brain injury rehabilitation. Brain Inj. 1993;7:45–51.
[28] Sohlberg MM, McLaughlin KA, Todis B, et al. What does it
take to collaborate with families affected by brain injury? A
preliminary model. J Head Trauma Rehabil. 2001;16:
498–511.
[29] Visser-Meily A, Post M, Gorter JW, et al. Rehabilitation of
stroke patients needs a family-centred approach. Disabil
Rehabil. 2006;28:1557–1561.
[30] Peiris CL, Taylor NF, Shields N. Patients value patient–thera-
pist interactions more than the amount or content of ther-
apy during inpatient rehabilitation: a qualitative study. J
Physiother. 2012;58:261–268.
[31] Darragh AR, Sample PL, Krieger SR. Tears in my eyes “cause
somebody finally understood”: client perceptions of practi-
tioners following brain injury. Am J Occup Ther. 2001;55:
191–199.
[32] Judd D, Wilson SL. Psychotherapy with brain injury survi-
vors: an investigation of the challenges encountered by
clinicians and their modifications to therapeutic practice.
Brain Inj. 2005;19:437–449.
[33] Kovarsky D, Schiemer C, Murray A. Humor, rapport, and
uncomfortable moments in interactions with adults with
traumatic brain injury. Top Lang Disord. 2011;31:325–335.
[34] Thorne S, Con A, McGuinness L, et al. Health care commu-
nication issues in multiple sclerosis: an interpretive descrip-
tion. Qual Health Res. 2004;14:5–22.
[35] Guba EG, Lincoln YS. Epistemological and methodological
bases of naturalistic inquiry. ECTJ. 1982;30:233–252.
[36] Thorne S, Kirkham SR, MacDonald-Emes J. Interpretive
description: a noncategorical qualitative alternative for
developing nursing knowledge. Res Nurs Health. 1997;20:
169–177.
10 M. BISHOP ET AL.
[37] Thorne S, Reimer Kirkham S, O’Flynn-Magee K. The analytic
challenge in interpretive description. Int J Qual Meth. 2004;
3:1–11.
[38] Hunt MR. Strengths and challenges in the use of interpret-
ive description: reflections arising from a study of the
moral experience of health professionals in humanitarian
work. Qual Health Res. 2009;19:1284–1292.
[39] Thorne S. Interpretive description. Walnut Creek (CA): Left
Coast Press; 2008.
[40] Charmaz K. The grounded theory method: an explication
and interpretation in contemporary field research. A collec-
tion of readings. In: Robert M. Emerson, editor.
Contemporary field research. Boston: Little Brown and
Company; 1983. p. 109–226.
[41] Malterud K, Siersma VD, Guassora AD. Sample size in quali-
tative interview studies: guided by information power. Qual
Health Res. 2016;26:1753–1760.
[42] Dodds TA, Martin DP, Stolov WC, et al. A validation of the
functional independence measurement and its perform-
ance among rehabilitation inpatients. Arch Phys Med
Rehabil. 1993;74:531–536.
[43] Kwon S, Hartzema AG, Duncan PW, et al. Disability meas-
ures in stroke. Stroke. 2004;35:918–923.
[44] Hsieh H-F, Shannon SE. Three approaches to qualitative
content analysis. Qual Health Res. 2005;15:1277–1288.
[45] QSR International Pty Ltd. Qualitative research analysis soft-
ware: NVivo10 for Windows 2015. Available from: http://
www.qsrinternational.com/products nvivo.aspx
[46] Guba E. Criteria for assessing the trustworthiness of natur-
alistic inquiries. ECTJ. 1981;29:75–91.
[47] Mays N, Pope C. Qualitative research in health care: assess-
ing quality in qualitative research. Br Med J. 2000;320:
50–52.
[48] Leplege A, Gzil F, Cammelli M, et al. Person-centredness:
conceptual and historical perspectives. Disabil Rehabil.
2007;29:1555–1565.
[49] McCormack B, Karlsson B, Dewing J, et al. Exploring per-
son-centredness: a qualitative meta-synthesis of four stud-
ies. Scand J Caring Sci. 2010;24:620–634.
[50] Leibert TW, Smith JB, Agaskar VR. Relationship between
the working alliance and social support on counseling out-
come. J Clin Psychol. 2011;67:709–719.
[51] Clark MS, Smith DS. Changes in family functioning for
stroke rehabilitation patients and their families. Int J
Rehabil Res. 1999;22:171–180.
[52] Audet CT, Everall RD. Therapist self-disclosure and the
therapeutic relationship: a phenomenological study from
the client perspective. Brit J Guid Couns. 2010;38:327–342.
[53] Hanson J. Should your lips be zipped? How therapist self-
disclosure and non-disclosure affects clients. Couns
Psychother Res. 2005;5:96–104.
[54] Peterson ZD. More than a mirror: the ethics of therapist
self-disclosure. Psychother: Theory Res Prac Train. 2002;39:
21–31.
[55] Wachtel PL. Therapeutic communication: knowing what to
say when [EBL Ebook library]. New York: Guilford Press;
2011.
[56] Gutheil TG, Gabbard GO. The concept of boundaries in clin-
ical practice: theoretical and risk-management dimensions.
Am J Psych. 1993;150:188–196.
[57] McGilton KS, Boscart VM. Close care provider–resident rela-
tionships in long-term care environments. J Clin Nurs.
2007;16:2149–2157.
[58] Gardner A. Therapeutic friendliness and the development
of therapeutic leverage by mental health nurses in commu-
nity rehabilitation settings. Contemp Nurse. 2010;34:
140–148.
[59] Hersh D. Aphasia therapists’ stories of ending the thera-
peutic relationship. Top Stroke Rehabil. 2010;17:30–38.
[60] Austin W, Bergum V, Nuttgens S, et al. A re-visioning of
boundaries in professional helping relationships: exploring
other metaphors. Ethics Behav. 2006;16:77–94.
[61] Safran JD, Muran JC. Resolving therapeutic alliance rup-
tures: diversity and integration. J Clin Psychol. 2000;56:
233–243.
[62] Richards C. Alliance ruptures: etiology and resolution.
Couns Psychol Rev. 2011;26:56–62.
[63] Elkin I, Falconnier L, Smith Y, et al. Therapist responsive-
ness and patient engagement in therapy. Psychother Res.
2014;24:52–66.
[64] Knobloch-Fedders LM, Pinsof WM, Mann BJ. The formation
of the therapeutic alliance in couple therapy. Fam Process.
2004;43:425–442.
[65] Bennett D, Parry G, Ryle A. Resolving threats to the thera-
peutic alliance in cognitive analytic therapy of borderline
personality disorder: a task analysis. Psychol Psychother-T.
2006;79:395–418.
[66] Jones M, O’Neill P, Waterman H, et al. Building a relation-
ship: communications and relationships between staff and
stroke patients on a rehabilitation ward. J Adv Nurs. 1997;
26:101–110.
[67] Lefebvre H, Pelchat D, Swaine B, et al. The experiences of
individuals with a traumatic brain injury, families, physi-
cians and health professionals regarding care provided
throughout the continuum. Brain Inj. 2005;19:585–597.
[68] Bright FA, Kayes NM, Cummins C, et al. Co-constructing
engagement in stroke rehabilitation: a qualitative study
exploring how practitioner engagement can influence
patient engagement. Clin Rehabil. 2017;31:1396–1405.
[69] Bright FA, Kayes NM, McPherson KM, et al. Engaging peo-
ple experiencing communication disability in stroke
rehabilitation: a qualitative study. Int J Lang Commun
Disord. 2018;53:981–994.