Improving communication and coordination between older adults and their caregivers

AI Caring Lab

Date

May-Aug 2023

Role

User Interviews, recruit participants, literature review, affinity diagramming

Team

John Zimmerman: Professor, Supervisor

Alicia Lee: PhD student

Sreehana Mandava: Research Assistant

Destiny Deshields: Research Assistant

Tools

Miro, Zoom

Problem Statement

With the rising prevalence of dementia and chronic diseases supporting informal caregivers becomes increasingly important, but minimal technological interventions have been created to support this group. Our study aims to create AI driven technology to improve communication and coordination between older adults with MCI (Mild Cognitive Impairment) and their care network. 

Existing research highlights the critical need for effective coordination among caregivers 

Existing research shows that caregivers need to work well together to handle unexpected care gaps. It also stresses the importance of using real-world data to understand the experiences of people with mild cognitive impairment (pwMCI) and their care networks.

However, there are gaps…

  • Past research primarily focuses on dementia patient care networks within middle-class families, often overlooking people in lower socioeconomic status (SES) communities.

  • Our study addresses this gap by focusing on the unique challenges faced by these people, who frequently remain undiagnosed and rely heavily on friends, neighbors, and community networks rather than formal services.

We interviewed older adults facing cognitive decline along with their informal caregivers to better understand their constraints

Participants: 

We interviewed 24 individuals, organized into 12 pairs, consisting of older adults experiencing cognitive decline and their informal caregivers. The pairs included:

  • 4 spousal pairs

  • 2 parent-child pairs

  • 1 grandparent-grandchild pair

  • 5 friend/neighbor pairs

Procedure

We conducted multisession semi-structured qualitative interviews. The interviews were held at the participants' residences and followed up either in person, via phone call, or through Zoom.

Focus Areas

During the interviews, we explored:

  • Current challenges faced by participants

  • Successes in managing cognitive decline

  • Needs regarding the coordination of day-to-day support

  • Future plans and strategies for managing cognitive decline

Our research indicates that older adults in lower SES communities have a diverse care network involving friends, neighbors and community members

Older adults in lower socioeconomic status (SES) communities rely significantly on friends, neighbors, and community members for daily support and emotional well-being. This broader care network contrasts with traditional findings that focus primarily on family members like spouses or children or professionals.

Pair08: “So if there was something that I had forgotten to order I might ask L[CG]. And there's a woman who was across the hall and someone upstairs who has also offered."

No tension or feelings of burden exist within the friend-based older adult and caregiver relationships. Older adults often felt comfortable seeking help from friends and neighbors compared to family members (who live far or are busy), because supporting is seen as a mutual (two-way) relationship among peers, reducing the perceived burden.

Pair01: “If I'm going to the store or if D's going, I can ask her, well, do you mind getting this when you go or something? She does me the same way.”

Pair 04’s care network helps her with transportation to doctor’s appointment

We also found that specific areas in the home serve as central coordination points where communication and task management occur

  • These “hotspots” were often located in kitchens, bedside, and desks , and facilitated efficient care coordination

  • Older adults use creative methods to remember tasks; digital calendars, reminder apps, paper receipts on refrigerators, and calendar in bathrooms.

Pair07 (OA-G)’s physical kitchen calendar 

Pair02 (OA-G)’s digital calendar

Pair01 (OA-B)’s sticky notes as a reminder of tasks

AI-driven solution to improve coordination and communication between older adults and their care network

Since we found a diverse care circle for older adults with cognitive impairment in lower-SES communities which could be difficult to coordinate with, we decided to design an AI agent that would help with this specific problem area.

  1. The AI agent would have a deep understanding of the tasks that the older adult needs help with.

  2. The agent would have a thorough understanding of the availability of the people in the older adult’s care network. It would take into account factors like their routine schedule and proximity to the older adult to determine their availability.

  3. The agent would then assign tasks to specific people in the care network so that it would be accomplished in the most efficient way possible. This agent would help the older adult by reducing the burden coordinating with multiple people to complete a task.

Discussions about the details of the AI agent are still on-going. Described above is one potential form the agent can take that was discussed early on.

Collaborations and Future Direction

  • Interviews raised the following unanticipated questions:

    • Do older adults care more about feeling in control than completing tasks? 

    • Are we overestimating the importance of task completion in caregiving relationships? 

  • Our next step is to conduct recurring interviews to gain deeper insights into caregivers' experiences, focusing on understanding their comfort levels and boundaries with support requests, while assessing the balance between task completion and maintaining a sense of control in caregiving relationships. 

  • This effort will be enhanced by collaborative studies involving teams from GT-UIR, CMU-UIR, and CMU-RI to develop comprehensive support systems for informal CGs helping OAs with cognitive decline.