Dr. Timothy Bickmore is an associate professor in the College of Computer and Information Science at Northeastern University. His research focuses on the development and study of relational agents, which are computer agents designed to build and maintain long-term, social-emotional relationships with people. Most recently, he has worked to develop animated agents with health education and health behavior change applications.
ElderBranch interviewed Dr. Bickmore to discuss his paper, “A Randomized Controlled Trial of an Automated Exercise Coach for Older Adults,” which he wrote with Dr. Rebecca A. Silliman of the Schools of Medicine and Public Health at Boston University (BU), Dr. Kerrie Nelson, Dr. Debbie M. Cheng, and Michael Winter, MPH, of BU’s School of Public Health, Lori Henault, MPH, of Boston Medical Center, and Dr. Michael K. Paasche-Orlow of BU’s School of Medicine.
What led you to research the use of automated exercise coaches with older adults? Why is this important? Does your current research augment other work in this area?
Simulated face-to-face conversation is an especially good medium for individuals who may have low health, reading, or computer literacy. Boston Medical Center (where the study was done) is an urban safety net hospital with a population that is especially high in these demographics (e.g., 40 percent have inadequate health literacy).
Each conversation with the virtual coach has a “social chat” dialog designed to build trust, rapport, and therapeutic alliance. Pilot testing at the hospital indicated that geriatrics patients were very accepting of the technology and really enjoyed this social dimension of the interactions.
Please describe your study. What methods did you use? What were your in-going hypotheses?
The trial was intended to scale up an early pilot study to test the feasibility of deploying an exercise intervention for urban, older adults from outpatient clinics and to demonstrate its efficacy (increased walking and fitness). Our secondary hypothesis was that it would be especially effective for older adults with inadequate health literacy.
All participants were given digital pedometers and paper log sheets. Intervention participants were also sent home with a touch-screen tablet computer loaded with the virtual coach software and asked to have daily conversations over the course of two months.
After the two months, intervention participants could continue having maintenance conversations with the coach on kiosks in the waiting rooms of their primary care clinics for an additional 10 months.
What were your key findings?
Acceptance, satisfaction, and use were very high during the initial two-month period. All participants demonstrated significant increases in walking over the course of the initial two-month period.
What are the implications of these findings?
It is feasible to deploy this kind of technology from outpatient clinics. (All tablets were returned and recycled.) The technology is well liked and accepted, and it leads to short-term increases in physical activity for urban, mostly minority, older adults.
Would you give any specific recommendations to the directors of out-patient clinics serving older adults as a result of your findings?
The use of pedometers alone does not appear to increase physical activity; they must be coupled with an intervention that involves regular contact with a counselor (human or virtual).
Were there any notable limitations to your research? What are the next steps to further your work in this area?
The waiting room kiosks were underutilized, due to several possible issues, and this may have resulted in lack of significant differences between the intervention and control groups after 12 months. We are exploring other methods for maintaining long-term adherence.
Is there anything else you would like to highlight regarding your research or findings that you think is important for our readership to understand?
Older adults, even those with limited health and computer literacy, are not technophobes. They can enjoy and use computer systems, as long as they are designed for their demographic.