It’s time to talk about our relationships: Exploring the role of therapeutic alliance in speech and language therapy,with stuttering intervention as an exemplar

1 Introduction

Therapy is both science and an art, and research should reflect this by investigating the operational step-by-step techniques employed by therapists, as well as the relational processes that lead to positive outcomes (Larner, 2004). Research in the discipline of SLT has primarily focused on the experimental validation of selected treatment methods and techniques (Hansen et al., 2021). The content of clinical practice in SLT is primarily technique focused; less attention has been paid to other aspects of therapy that are known to influence outcomes, such as the therapeutic alliance and person-related variables. This narrow focus is problematic because it ultimately limits the ability of SLTs to help their clients.

To be an effective therapist, two types of competence must be mastered: clinical competence and relational competence (Sylvestre & Gobeil, 2020; Wampold, 2001). Clinical competence is the mastery of knowledge relating to a specific area of expertise (e.g., the ability to administer an assessment, or to execute a specific intervention). Relational competence incorporates the attitudes and skills that are necessary to establish a therapeutic alliance with a client (Wampold, 2001). A therapist’s adherence to a specific treatment protocol, whilst important, is not the sole contributor to treatment outcomes. Although the focus of evidence-based practice has primarily centred on efficacy evidence, it must also include client-based evidence, such as client opinions, values and treatment preferences (Greenhalgh et al., 2014; McCurtin et al., 2019). To facilitate the collection of client evidence, therapists must develop and use their relational competence and build a therapeutic alliance with their clients. The discipline of SLT must therefore investigate the role of variables beyond specific treatment techniques. This will also help to align SLT with other disciplines, such as psychotherapy, that have been demonstrating the value of these variables for several decades.

2 What is therapeutic alliance?

Various terms have been used, often interchangeably, to describe the interactional and relational processes that operate during therapy. These include: therapeutic alliance, working alliance, rapport, therapeutic bond and therapist-client relationship (Hansen et al., 2021; Kayes & McPherson, 2012; Lawton et al., 2018a). These terms carry different meanings, with some emphasising the affective element of the relationship (e.g., therapeutic bond), and others placing more weight on the interactional elements of the relationship (e.g., working alliance) (Kayes & McPherson, 2012). The term therapeutic alliance is used throughout this paper as it encompasses both the interactional and relational processes that are co-constructed by the client and therapist during therapy (Green, 2006; Walsh & Felson Duchan, 2011). Importantly, the therapeutic alliance is not a static state; rather, it develops and evolves continuously over time between the individuals involved (Walsh & Felson Duchan, 2011). These individuals include the client and the therapist, but can also include other parties such as caregivers who may play an essential role in therapy (Freckmann et al., 2017).

The origins of the therapeutic alliance are grounded in psychotherapy, namely Freud’s (1913) and Rogers’ work (1957). Bordin presented a tripartite conceptualisation of what he termed the “working alliance”, which was composed of the client’s and therapist’s agreement on therapy goals, the mutual agreement and collaboration on tasks to meet the goals, and the interpersonal bond between the client and therapist (Bordin, 1979, p. 252). This conceptualisation parallels more contemporary foci in healthcare such as shared decision-making and person-centred care, which emphasise the merits of integrating client treatment preferences and desired outcomes (Bomhof-Roordink et al., 2019; WHO, 2015). Bordin’s conceptualization has more recently been expanded due to its increased application to a wider range of healthcare disciplines, and additional constructs such as communication, client empowerment (including power sharing and client self-efficacy), family systems and social context have been added (Elvins & Green, 2008; Göldner et al., 2017; Kim et al., 2001, Lawton et al., 2018a). Based on this prior literature, we will use the term therapeutic alliance to refer to the interactional and relational processes that are cocreated by the client and therapist throughout therapy and influenced by variables within and beyond the client-therapist dyad.

Sylvestre and Gobeil (2020) differentiated between therapeutic relationship and therapeutic alliance. They asserted that the therapeutic relationship is itself a component of the broader therapeutic alliance, and they highlighted that this relationship develops from the very first interaction between the client and therapist. The therapeutic relationship is the trusting foundational affective bond on which the therapeutic alliance is built; it consists of the attitudes and feelings that the client and therapist experience and express with one another (Fourie et al., 2011; Sylvestre & Gobeil, 2020). In order to establish and maintain this initial therapeutic relationship and the subsequent therapeutic alliance, therapists are required to adopt a person-centred care approach and implement the process of shared decision-making (Sylvestre & Gobeil, 2020). In a person-centred care approach, the clinician acknowledges each individual client in their personal context, and respects, listens to, informs and involves them in all aspects of their care (Epstein & Street, 2011). It is a form of appreciative healthcare, in which the therapist is aware of the positive potential in their client (Roberts & Machon, 2015). Person-centred care is endorsed as an essential approach to health service delivery, with frameworks such as the International Classification of Functioning, Disability and Health (ICF) providing a useful roadmap for therapists to support its implementation (WHO, 2013, 2015).

Using stuttering intervention as an example, a person-centred care approach involves the therapist acknowledging the individual nature and experience of stuttering, incorporating collaborative decision-making into the intervention (such as agreement on goals and treatment choice), and using client-derived outcome measures that are specific to the client’s personal goals. In their exploration of the association between therapeutic alliance and person-centred care, Hamovitch et al., (2018) concluded that the two are inextricably linked, with each influencing the other. In other words, the therapeutic alliance is both a facilitator and an outcome of person-centred care. Shared decision-making, a core component of person-centred care, has been proposed to facilitate the establishment of the therapeutic alliance (Sylvestre & Gobeil, 2020). It involves the client and therapist working together to agree on the tasks required to meet the client’s personal goals through consideration of the best evidence and the client’s values and preferences (Ferro-López et al., 2021; Haesebaert et al., 2019; Sylvestre & Gobeil, 2020). Elements of shared decision-making include describing treatment options to clients, including client preferences, and increasing clients’ choice awareness (Bomhof-Roordink et al., 2019). Stewart (2022) advocates for the creation of a context of choice to support the therapeutic alliance when working with individuals who stutter. In line with research identifying the bidirectional relationship between the therapeutic alliance and person-centred care, it can be assumed that shared decision-making both facilitates and is enhanced by the therapeutic alliance (Hamovitch et al., 2018).

3 What variables influence the therapeutic alliance?

A range of person-related and external variables have been identified in the healthcare literature as influencing the establishment, maintenance and quality of the therapeutic alliance.

4 What can we learn about the therapeutic alliance from the psychotherapy literature?

The range of variables outside of the specific intervention technique that contribute to therapeutic change have been acknowledged in the psychotherapy literature for several decades (Cartwright & Hardie, 2012; Lambert, 2013; Wampold, 2015). While SLT cannot be considered a psychotherapy, working with clients with communication disorders such as stuttering certainly requires elements of psychological knowledge and skills, if the totality of the disorder and its implications are to be successively managed. Wampold’s (2015) Contextual Model detailed three core pathways that lead to psychotherapeutic outcomes: an authentic alliance between the client and therapist, client expectations of therapy, and the specific treatment technique. Wampold (2015) argued that an initial bond between the client and therapist needs to be established prior to the employment of these three pathways. Similarly, Sylvestre and Gobeil (2020) argued for the necessity of an initial trusting affective bond between an SLT and their client, on which the therapeutic alliance can then be constructed. Lambert (2013) estimated that 40% of psychotherapy outcome variance is due to extra-therapeutic factors, which include client qualities (e.g., motivation) and environmental variables (e.g., social support); 30% is due to common factors, such as the therapeutic alliance or clinician-related variables (e.g., empathy), and only 15% is due to the specific therapeutic techniques. The important message for SLTs, who tend to be highly focused on technique, is that their mastery of clinical competence alone is not sufficient for the delivery of effective interventions. Relational competence is also required to facilitate positive treatment outcomes. Further, given the primary focus on the experimental validation of treatment techniques that currently characterises SLT research, increased attention to the measurement of other variables, such as therapeutic alliance and person-related factors, needs to be prioritised.

Numerous systematic reviews and metanalyses in the discipline of psychotherapy (and more recently physical rehabilitation) have concluded that the strength of the therapeutic alliance is related to treatment adherence, treatment outcomes and client satisfaction with treatment (Baldwin et al., 2007; Flückiger et al., 2018; Hall et al., 2010; Horvath et al., 2011; Lakke & Meerman, 2016; Martin et al., 2000). A client’s perceived quality of the therapeutic alliance early in therapy (by the third session) is a significant predictor of outcome, with favourable judgements generally resulting in positive outcomes, and unfavourable judgements associated with increased dropout from therapy (Castonguay et al. 2006; Leibert, 2011). The first few sessions of therapy, therefore, represent a “window of opportunity” during which the therapist must prioritise and foster the development of the therapeutic alliance, both to reduce the likelihood of premature dropout and to maximise treatment outcomes for the client (Bachelor & Horvath, 1999, p. 139).

5 What can we learn about the therapeutic alliance from the SLT literature?

In comparison to the vast exploration of the role of therapeutic alliance in the psychotherapy literature, such studies remain at an emerging level in the discipline of SLT. Most of the research is based on theoretical knowledge that is primarily generated from psychotherapeutic approaches (Göldner et al., 2017). Despite this, research examining the influence of the therapeutic alliance on client motivation, engagement in the therapeutic process and outcomes is becoming increasingly evident in the discipline of SLT, in particular with interventions for aphasia (Lawton et al., 2016; Lawton et al., 2018a,b; Lawton et al., 2020), paediatric speech and language disorders (Freckmann et al., 2017), cluttering (Sønsterud, 2019) and stuttering (Connery et al., 2020a,2021; Sønsterud et al., 2019). Lawton et al. (2018b) investigated people with aphasia’s experiences of constructing and maintaining therapeutic alliances with speech and language therapists (SLTs). The results of their study highlighted the essential role the SLT plays in the construction of therapeutic alliance, in terms of the behavioural processes they employ including responding to their client’s needs and communicating information. In fact, SLTs’ own perspectives of the therapeutic alliance in aphasia rehabilitation support these findings, with a range of processes identified by this stakeholder group as being important to developing and maintaining a therapeutic alliance. These processes, including getting to know the person, showing empathy, preserving hope, promoting goal ownership, and being encouraging, are moulded by contextual factors such as the client’s family and organisational drivers (e.g., time and resource constraints) (Lawton et al., 2018a).

Systematic reviews of stuttering treatment efficacy have revealed a diverse range of effective interventions, with no significant difference in outcomes (Baxter et al., 2015; Connery et al., 2020b; Herder et al., 2006). Such findings may be explained by the presence of variables beyond the intervention technique that are common to all interventions. In their investigation of the perspectives of adults who stutter on the characteristics of therapists that contribute to an effective therapeutic experience, Plexico et al. (2010) found that over 50% of the participants identified the importance of a trusting therapeutic alliance in promoting successful change. More recent research providing client- and practice-based evidence for effective stuttering interventions offers support for this argument (Connery et al., 2020a; Connery et al.; 2021). Connery et al. (2020a) explored the perspectives of international researchers and clinical experts in the field of stuttering on the components of effective stuttering intervention using semi-structured interviews. One of the three core themes identified was “a really collaborative relationship where we are both bringing our sense of expertise to this” (Connery et al., 2020a, p.6). This theme represented the client- and therapist-related attributes that are important for enhancing the therapeutic process and treatment outcomes, as well as the benefits of collaborative work and a positive therapeutic alliance in enhancing treatment outcomes. In another stakeholder-focused study, SLTs and adults who stutter both agreed that the therapeutic alliance and variables related to the client (e.g., client motivation) and the SLT (e.g., their knowledge of stuttering) are indispensable ingredients of effective stuttering interventions (Connery at al., 2021). These studies highlight the wide-ranging variables that lead to therapeutic change and the need for SLTs to be competent in the relational aspects of therapy. However, these studies do not examine the construct of therapeutic alliance in its entirety, and further explication of the specific attributes and behaviours that are conducive to a positive alliance in stuttering intervention is needed (Connery et al., 2021; Manning, 2010).

Sønsterud et al. (2019) found significant associations between the strength of the alliance and clients’ motivation and treatment outcomes 6 months post-intervention for stuttering. The intervention focused on breath and body awareness, speech modification, mindfulness-based strategies and presentation skills. Results revealed that mutual agreement on therapy tasks and client-led goals were of particular importance to treatment outcomes, and that such factors contributed more to outcomes than the emotional bond. The authors cautioned that this finding does not indicate that the client-therapist bond should be neglected; rather, it provides a foundation for therapy tasks and client-centred goals to be established. Such findings may be further explained by considering the more task-based nature of SLT in comparison to psychotherapy. Sønsterud et al. (2019) argued that the bond between client and clinician may play less of a role in SLT than psychotherapy, which focuses more on the individual’s internal experiences. The client-therapist bond may however have increased relevance for other intervention approaches, targeting, for example, client goals on the management of the emotional and cognitive aspects of stuttering.

6 Tools for assessing the therapeutic alliance

Routine clinical use of therapeutic alliance measures can provide therapists with information required to repair problems in the alliance and identify clients with a poor response to an intervention (Manning, 2010). Most measurement tools originate from the field of psychotherapy: Elvins and Green (2008) reviewed 32 such tools, compared to just four identified in a recent review in the physiotherapy literature (Sánchez et al., 2020). The perspectives of a range of stakeholders must be obtained in order to adequately assess the therapeutic alliance. These include the client, their caregiver (if appropriate), and the therapist (Freckmann et al., 2017). The Working Alliance Inventory (WAI) is one self-report instrument that measures the strength and quality of the client-therapist relationship across Bordin’s (1979) three components of bond, goal and task (Horvath & Greenberg, 1989). A shorter 12-item version, the Working Alliance Inventory-Short Revised (WAI-SR), was later developed; it remains one of the most frequently used measures of the therapeutic alliance in psychotherapy (Hanson et al., 2002). Both have been recommended by Sønsterud et al. (2019) as useful clinical tools for evaluating the therapeutic alliance in SLT. Using the same theoretical underpinnings, the Therapeutic Alliance Scales for Children-Revised (TASC-R) was developed for use with children and their therapists, and measures therapeutic alliance across the three dimensions of task, bond and goals (Creed & Kendall, 2005). Variations of this measure such as the TASCP (Therapeutic Alliance Scale for Caregivers and Parents) also exist to collect the perspectives of other parties (Accurso et al., 2013).

In SLT, Lawton et al. (2019) recently developed the Aphasia and Stroke Therapeutic Alliance Measure (A-STAM), including both client and therapist versions. Although no stuttering-specific measure of the therapeutic alliance exists, the clinical use of rating scales to assess the quality of the therapeutic alliance in stuttering intervention has been recommended (Plexico et al., 2010; Zebrowski & Kelly, 2002). One example is the Session Rating Scale (SRS) which consists of four interacting items in visual analogue form and in line with Bordin’s (1979) classical definition of the therapeutic alliance: Relationship, Goals and Topics, Approach or Method, and Overall (Miller et al., 2002). Each item is rated from 1-10, and if the total score is lower than 36, a discussion with the client is required to identify areas requiring change moving forward (Miller et al., 2002).

7 Strategies to foster a therapeutic alliance with an adult who stutters

SLTs play a key role in establishing and maintaining the therapeutic alliance with adults who stutter. Based on the literature reviewed above, a range of recommendations to bolster this in clinical practice will now be presented. Although these recommendations are specific to intervention with adults who stutter, they also have relevance to clinical work with clients with other communication difficulties.

8 Recommendations for future research

Although psychotherapeutic constructs of the therapeutic alliance offer valuable knowledge to the discipline of SLT, the development and maintenance of alliances with clients with communication difficulties are more nuanced. This is due to clients’ specific deficit in communication (which is an essential skill for developing a therapeutic alliance), and the fact that the processes involved in communication rehabilitation are different to those commonly used in psychotherapy (Lawton et al., 2018b). It is therefore fundamental that the evidence base for the therapeutic alliance specific to communication difficulties be expanded so as to include the perspectives of all key stakeholders.

Client-based evidence is an under-represented form of knowledge in the SLT literature, and this is echoed in the paucity of research examining client perceptions of the therapeutic alliance (Fourie, 2009; McCurtin et al., 2019). It is likely that there are differences in the experiences of the therapeutic alliance between different communication disorder cohorts, given the chronic nature of some conditions (e.g., stuttering) and the more acute nature of others (e.g., aphasia post-stroke). Therefore, a conceptualisation of the therapeutic alliance as it relates to specific communication disorders is warranted. This has recently been demonstrated in the aphasia rehabilitation literature, with a stakeholder-focused exploration of the experience of therapeutic alliance from the perspective of SLTs and adults with post-stroke aphasia (Lawton et al., 2018a, Lawton et al., 2018b; Lawton et al., 2020). The stuttering-specific literature would benefit from a similar stakeholder-focused exploration and conceptualisation of the therapeutic alliance. Obtaining the perspectives of both stakeholder groups (SLTs and adults who stutter) is essential due to differences in clients’ and therapists’ perceptions, with clients viewing the therapeutic alliance as stable and therapists indicating more change over time (Horvath et al., 2011; Martin et al., 2000). Additionally, the therapeutic alliance has been characterised by an imbalance of power between healthcare professionals and clients, with power disproportionately weighted toward therapists and highlighted by reduced collaboration and therapist-led goal-setting (Lawton et al., 2016). It is therefore essential to obtain the perspectives on what the therapeutic alliance means to both stakeholder groups, to develop a conceptualisation that can inform further research in the area and offer guidance for those working in clinical practice.

9 Conclusion

The therapeutic alliance is a multi-faceted and dynamic construct that has been recognised as an active and essential ingredient of effective psychotherapy for several decades. In more recent times, its role in healthcare more generally has gained traction, with research concluding that the strength of the therapeutic alliance is related to treatment outcomes, adherence and satisfaction. Research into the range of person-related and contextual variables that influence the therapeutic alliance highlights the complexity of this construct that goes beyond Bordin’s original tripartite conceptualisation of bond, goal and task. Therapeutic alliance is a complex construct that is influenced by factors within and beyond the therapist-client dyad. Current conceptualisations of this construct in the literature have primarily evolved from the discipline of psychotherapy. They therefore fail to capture the true meaning of the therapeutic alliance as it relates to the context of disciplines such as SLT.

Despite significant advancements made in the experimental validation of SLT treatment methods and techniques, less attention has been paid to other relational aspects of the therapeutic process that influence outcomes. It is therefore essential that the field of SLT directs its attention to the investigation of variables such as the therapeutic alliance to ensure equal value is placed on the clinical and relational elements of SLT. Stakeholder-informed research exploring the conceptualisation of this construct as it applies to specific cohorts of individuals with communication needs, such as adults who stutter, is recommended. This will support its application in clinical practice and ensure optimum outcomes for clients with communication difficulties are achieved.