Web-Based Cognitive Behavioral Therapy for Depression Among Homebound Older Adults: Development and Usability Study

Background Homebound older adults are a high-risk group for depression. However, many of them face barriers to accessing evidence-supported mental health treatments. Digital mental health interventions can potentially improve treatment access, but few web-based interventions are explicitly tailored for depression in older adults. Objective This paper describes the development process of Empower@Home, a web-delivered intervention for depression in homebound older adults that is based on cognitive behavioral therapy, and reports on the outcomes of usability studies. Methods Empower@Home was developed in collaboration with community agencies, stakeholders, and older adults, guided by user-centered design principles. User needs were assessed through secondary data analysis, demographic and health profiles from administrative data, and interviews and surveys of community partners. A comparative usability evaluation was conducted with 10 older adults to assess the usability of Empower@Home compared to 2 similar programs. Field testing was conducted with 4 end users to detect additional usability issues. Results Feedback and recommendations from community partners heavily influenced the content and design of Empower@Home. The intervention consists of 9 sessions, including psychoeducation and an introduction to cognitive behavioral therapy skills and tools through short video clips, in-session exercises, an animated storyline, and weekly out-of-session home practice. A printed workbook accompanies the web-based lessons. In comparative usability testing (N=10), Empower@Home received a System Usability Scale score of 78 (SD 7.4), which was significantly higher than the 2 comparator programs (t9=3.28; P=.005 and t9=2.78; P=.011). Most participants, 80% (n=8), preferred Empower@Home over the comparators. In the longitudinal field test (n=4), all participants reported liking the program procedures and feeling confident in performing program-related tasks. The single-subject line graph showed an overall downward trend in their depression scores over time, offering an encouraging indication of the intervention’s potential effects. Conclusions Collaboration with community stakeholders and careful consideration of potential implementation issues during the design process can result in more usable, engaging, and effective digital mental health interventions.