This study identified and explored the core components of a strong therapeutic alliance and those factors perceived to influence its development within an inpatient stroke rehabilitation setting. The therapeutic alliance has been a consistent predictor of outcome in psychotherapy and mental health research and there is growing evidence that it may facilitate beneficial change in other health settings. A systematic review of the brain injury rehabilitation literature (completed as part of this research) identified associations between the therapeutic alliance and a range of rehabilitation outcomes including improved productivity, self-awareness and emotional regulation. However, the review highlighted very few studies specifically exploring the therapeutic alliance in stroke rehabilitation, and a lack of clarity around whether psychotherapy-based alliance concepts and measures apply to brain injury rehabilitation research.

The empirical study then undertaken used Interpretive Description methodology to explore client and clinician perceptions of their therapeutic alliances. Clients (n=10) with a range of stroke-related difficulties and clinicians (n=7) from a range of professional backgrounds (nurses and therapists) were recruited. Semi-structured individual (clients) and focus group (clinicians) interviews were the primary source of data collection. Interviews were transcribed and uploaded to NVivo data management software, and data analysed using thematic analysis with careful attention to rigour.

Two themes emerged from the data. The first theme, everyone is different, described the core components of a strong therapeutic alliance. These were: a personal connection, a professional collaboration and family/whānau collaboration. Client participants appeared to prioritise each component to varying degrees and these priorities could change over their rehabilitation stay. Further, a range of personal factors and competing demands seemed to influence alliance development. This meant that therapeutic alliances were often dynamic and complex. The second theme, relationship disruptions, outlined the factors that could compromise alliance quality. The strength of the pre-existing alliance and the management of the ‘disruption’ seemed to impact on whether client participants continued to engage in these alliances and therapy, or whether a relationship breakdown would occur and impact on future work with the practitioner.

The therapeutic alliance model proposed in this thesis extends current understandings of how alliances may be conceptualised and operationalised in an inpatient stroke rehabilitation service. Based on these findings and other relevant literature, a number of practice-based recommendations are discussed. These include the need for a person-centred, flexible approach to determine, regularly review and respond to each client’s alliance needs and preferences, rather than assume what these may be. In this research, strong alliances were considered to alleviate distress and promote hope, rehabilitation engagement, wellbeing and progress. As such, explicit consideration of each person’s therapeutic alliance needs and preferences should arguably be considered a crucial component of clinical work, in order to augment stroke rehabilitation processes and outcomes. Intervention studies testing different approaches to developing and maintaining therapeutic relationships are needed, with this work informing the key components of those interventions. Meantime, the study makes clear that the way clinicians connect with each client matters, and that how we might best connect varies between clients and over time.