Author Identifiers

Sarah D'Souza

Date of Award


Degree Type


Degree Name

Doctor of Philosophy


School of Medical and Health Sciences

First Advisor

Erin Godecke

Second Advisor

Natalie Ciccone

Third Advisor

Deborah Hersh

Fourth Advisor

Elizabeth Armstrong

Fifth Advisor

Heidi Janssen


Background: Aphasia is an acquired communication disorder that affects approximately 30% of first ever stroke survivors and persists one-year post-stroke in up to 61% of survivors. Aphasia impacts on all communication modalities with significant negative consequences for social participation, interpersonal relationships, autonomy, capacity to work and quality of life. It is recognised that the environment can influence neural remapping during early stroke recovery. However, patients with aphasia (PWA) following stroke have been observed to spend less than 30% of their day communicating and 44% of their day alone during their first weeks of in-patient rehabilitation. Inadequate opportunities for communication places PWA at risk of developing maladaptive behaviours such as learnt non-use of language. This can negatively impact on aphasia language recovery through lack of language use with adverse consequences for healthrelated quality of life. An enriched environment (EE) refers to conditions which promote physical, cognitive and social activity and has been shown in animal models of stroke to enhance neuroplasticity, promote better learning and memory, and contribute to significant improvements in motor function. The human equivalent model in an acute and a rehabilitation unit results in patients following stroke spending more time engaged in activity and less time sleeping and alone, however is yet to demonstrate positive effects on clinical outcomes. Aphasia is a complex language impairment and PWA may need support within an EE. This pilot study explores the development, implementation and investigation of an adapted model of an EE, a Communication Enhanced Environment (CEE) model, as a strategy to provide PWA and patients without aphasia (PWOA) greater opportunities to engage in language activities during in-patient rehabilitation early after stroke.

Method: This before-and-after mixed methods pilot study involves one mixed acute and slow stream rehabilitation ward and one rehabilitation ward in a metropolitan private hospital in Perth, Western Australia. A hospital-based CEE model was developed, implemented and investigated. As a basis for implementation of an individual and systems-level behavioural change intervention, the study design aligned with implementation science principles. The study included:

i. the before phase which involved observation of patients following stroke (the control group; n=7; PWA=3, PWOA=4). Behavioural mapping was completed during the first minute of each five-minute interval over 12 hours (between 7am and 7pm) to determine patient engagement in language activities. Semi-structured interviews which incorporated a qualitative description approach were conducted with patients (n=7) to determine factors that were perceived to facilitate or create a barrier for communication. A qualitative description approach was also used throughout focus groups that were conducted with hospital staff and volunteers (n=51) to explore their perceptions of: their knowledge of, skills with, and attitude towards aphasia and communication; opportunities for potentially enhancing communication and language activities for patients; and additional aspects that could be included in the CEE model.

ii. the implementation phase where the CEE model was developed and embedded in usual care.

iii. the after phase which involved repeated data collection with a different cohort of patients (the intervention group; n=7; PWA=4, PWOA=3). The availability of the CEE model was monitored by hospital site investigators (a senior physiotherapist and a speech pathologist). Comparisons of patient engagement in language activity levels were conducted. Patient interviews (n=7) and staff and volunteer focus groups (n=22) were conducted. This was to determine differences following the implementation of the CEE model in: patient experiences of communication; hospital staff knowledge of, skills with, and attitudes towards aphasia and communication; and staff experiences of the implementation and use of the CEE model.

Results: A total of 29 of the 41 (71%) CEE model initiatives were reported to be available to the intervention group. A total of 24 of the 29 (83%) CEE model initiatives were reported to be available for PWA. A total of 5 of the 12 (42%) CEE model initiatives were reported to be available for PWOA. The intervention group engaged in higher, but not significant (CI 95%), levels of language activities (600 of 816 observation time points, 73%) than the control group (551 of 835 observation time points, 66%). Patients described variable experiences accessing different elements of the CEE model, which were influenced by individual patient factors, staff factors, hospital features as well as staff time pressures. Patients who were able to access elements of the CEE model described positive opportunities for their engagement in language activities. Staff perceived the CEE model increased their knowledge of aphasia and developed their skills and confidence in using communication supportive strategies. After the implementation of the CEE model, staff reported embedding communication within usual care tasks and rehabilitation activities, and perceived communication as a shared responsibility within the multidisciplinary team. There were several unforeseen factors that occurred which may have influenced the implementation and use of the CEE model including: a reduction in stroke admissions at the hospital site; a reduction in nurse-to-patient ratio; a ward reconfiguration; and reduced access to communal dining opportunities. Staff identified a range of factors which influenced the implementation and use of the CEE model. These included: the hospital context; and individual staff, volunteer and patient characteristics; the ease of use for both staff and patients and the implementation approach.

Conclusions: Consideration of implementation science approaches in this pilot study informed the development of a CEE model. This individual and service-level multidisciplinary team intervention was successfully implemented in clinical practice in a mixed acute and slow stream rehabilitation ward and a rehabilitation ward in a private hospital. This study demonstrated that the implementation of a CEE model within this hospital setting was feasible, with patients, staff and volunteers reacting positively to the CEE model overall. The unforeseen contextual challenges that occurred during the study period were beyond the control of the research team and demonstrated the everchanging and challenging nature of the hospital environment. The reduced availability of the CEE model for PWOA requires further attention to determine if the elements of the CEE model could be better applied to meet the needs of this population. Differences between levels of patient language activity before-and-after the implementation of the CEE model did not reach statistical significance. However, some individual increases taken together with the positive feedback suggest that a CEE model has value in enhancing the ward environment for staff and volunteers, the hospital system, and patients following stroke. This study highlights the complex and dynamic nature of the hospital environment which should be considered in future studies investigating individual and hospital service-level interventions such as EE or CEE models. Staff perceptions of factors contributing to the implementation and use of the CEE model provide valuable insights which may inform the implementation approach of future innovative interventions and subsequent development of the CEE model. Results from this study highlight the need to further explore the question of feasibility of a CEE model and patient access to the intervention across multiple ward contexts before we can explore the question of effectiveness. Future iterations of a CEE model should co-designed with patients and their family members.

Access Note

Access to Chapters 6 and 9 of this thesis is not available.