Poster The Joint Annual Meeting of the Stroke Society of Australasia (SSA) and Smartstrokes 2023

Co-designing an embodied computer agent to self-manage stroke recovery: The Taking Charge Intelligent Agent (TaCIA) Project (#147)

Heidi Janssen 1 2 3 , Paulo S Maciel 4 , Deborah Richards 4 , Harry K McNaughton 5 , Vivian Fu 5 , Christopher Levi 6 7 8 , Michael Nilsson 1 9
  1. Heart and Stroke Program , Hunter Medical Research Institute, New Lambton Heights, NSW, Australia
  2. School of Health Sciences, University of Newcastle, Newcastle, NSW, Australia
  3. Hunter Stroke Service, Hunter New England Local Health District, New Lambton Heights, NSW, Australia
  4. School of Computing, Faculty of Science and Engineering, Macquarie University, Sydney, NSW, Australia
  5. Stroke/Rehabilitation Research, Medical Research Institute of New Zealand, Wellington, New Zealand
  6. Hunter New England Health Local Health District, New Lambton Heights, NSW, Australia
  7. School of Medicine and Public Health, University of Newcastle, Newcastle, NSW, Australia
  8. Heart and Stroke Program , Hunter Medical Research Institute, New Lambton Heights, NSW, Australia
  9. Centre for Rehab Innovations, University of Newcastle, Newcastle, NSW, Australia

Background: The conversation-based Take Charge tool is an evidenced-based self-management intervention shown to significantly improve independence and quality of life after stroke. Take Charge is delivered face-to-face by a facilitator. Adaption of the tool to be delivered via an embodied computer agent may reduce barriers to access that post-stroke impairments and/or delivery preference impose.

 

Aims: Work with key stakeholders to co-design and test the feasibility of an embodied computer agent to support self-management after stroke.

 

Methods: Co-design with stakeholders (i) Take Charge researchers, (ii) previous Take Charge facilitators and (iii) stroke survivors was conducted to build and then test the feasibility and experience of an embodied computer agent. Feasibility measures collected included (i) motivation (Autonomy, Mastery, Purpose and Connectedness, 1-4), (ii) usability (System Usability Scale, 0-100), (iii) therapeutic alliance (Session Rating Scale, 0-10) and (iv) acceptability.

 

Results: Two Take Charge researchers (50% female) and four Take Charge facilitators (100% female health professionals) contributed to refinement of the research team built embodied computer agent prototype which was named ‘TaCIA’. Stroke survivors (n=8, 63% male, 2 years post-stroke) with median age of 53 yrs [IQR 49.3-59.5] completed (n=7) the prescribed session (mean=28min) with TaCIA. Participants scored average motivation (3.2/4 [SD 0.6], good system usability (71/100, [SD 13]) and a satisfactory therapeutic alliance (8.0/10 [SD 2.3]). Stroke survivors agreed that: TaCIA could be used by other stroke survivors, they would like to keep using TaCIA and they would recommend use of TaCIA to other stroke survivors. 

 

Conclusion: This is the first known attempt to build an embodied computer agent for use after stroke. Results suggest that use of this medium to support self-management is acceptable to stroke survivors but future iterations tested on a larger sample of people with a broader range of time post-stroke and age should focus on improving motivation, usability and therapeutic alliance.