Designing Intelligent Assistants to Support Older Adults' Personal Health Information Management at Home

Abstract

Recalling and managing health information tasks at home is crucial for older adults who seek independent living. However, organizing and handling large amounts of information can be challenging, particularly when coordinating with caregivers. The range of tasks involved in health information management encompass accessing doctors’ notes available in patients’ portals, addressing health-related queries, and scheduling reminders. These tasks often demand a substantial time commitment from both older adults and their caregivers. To tackle these challenges, intelligent assistants (IAs), such as advanced voice solutions, have emerged. These solutions acknowledge the complexities inherent in health information management, offering features like voice transcription to generate encounter notes from patient-physician conversations or conversational technologies, such as voice input, to streamline health management tasks. Research indicates that these systems hold promise for empowering older adults to independently oversee their health, alleviating some of the burdens on caregivers and minimizing the need for frequent doctor follow-ups. Yet, we are still learning about older adults’ needs and expectations for intelligent assistants that support their personal health information management at home. To address older adults’ broader needs for health information support, I performed four qualitative studies and designed prototypes. These studies aim to uncover older adults’ needs and beliefs about IAs that support organizing and recalling medical instructions provided during doctors’ visits and receiving tailored answers to health-related questions once at home. I identified overlap in older adult and caregiver perceptions of supportive and threatening interactions in care relationships and of emerging IAs designed to support health information management practices. Based on the findings, I generated design implications that led to the development of a high-fidelity prototype. The prototype includes automated and non-automated features for an IA that support older adults’ health information management once at home.