Using Smart Displays to Implement an eHealth System for Older Adults with Multiple Chronic Conditions: A Randomized Controlled Trial.

Background: Smart displays and smart speakers offer ease of voice interaction, which may be more accessible and appealing for older adults experiencing chronic pain with other multimorbid chronic conditions. Previous ElderTree trials found that the socio-emotional benefits of ElderTree (vs control) were stronger for older adults with high rates of primary care use and for older adults with multiple chronic conditions and that the socio-emotional benefits were stronger for women than for men.

Objective: This study tested whether older adults suffering from chronic pain, in the context of multiple other chronic conditions, would show more use and benefits of ElderTree, a digital/eHealth intervention targeting pain and quality of life, if they were given the intervention on a smart display rather than a laptop, and whether both the smart display and laptop groups would show benefits relative to a control group given no device or access to the intervention.

Methods: In a non-blinded, randomized controlled trial, 269 participants with chronic pain and at least 3 high-risk chronic conditions were recruited from the University of Wisconsin-Madison Department of Family Medicine and General Internal Medicine system (UW Health) as well as community organizations in the Madison, Milwaukee, and Beloit, Wisconsin, areas and were assigned 1:1:1 to (1) a smart display with Internet access and ElderTree, along with their usual care; (2) a touchscreen laptop with Internet access and ElderTree, along with usual care; or (3) usual care alone. All participants were age 60 or older, had a chronic pain diagnosis and/or reported chronic pain, and had at least three common chronic conditions. Primary outcomes were pain interference and psychosocial quality of life. Proposed mediators of effects of study arm (ElderTree vs active control) on changes in primary outcomes over time were ElderTree use at 4 months (for ElderTree arms only), competence, relatedness, motivation, exercise, and pain intensity. Moderators were gender, number of chronic conditions, and barriers to technology use. Data sources were surveys at baseline, 4, and 8 months and continuously collected ElderTree usage.

Results: There was not a significant difference between the laptop vs smart display groups for changes in pain interference over time (arm x time interaction b = -0.11, 95% Cl -1.07 to 0.85; p = .82) or psychosocial quality of life over time (arm x time interaction b = -0.21, 95% CI -0.96 to 0.55, p = .56). There was also a non-significant difference between the combined laptop + smart display group vs control group for change in pain interference over time (arm x time interaction b = -0.41, 95% Cl -1.23 to 0.41; p = .33) and psychosocial quality of life over time (arm x time interaction: b = 0.04, 95% Cl -0.61 to 0.69; p = .90). No mediation was tested on the primary outcomes since the effects of study arm on primary outcomes were non-significant. Gender did not moderate the effect of laptop versus smart display groups on changes over time in pain interference (b = -1.56, CI -3.56 to 0.44, p = .13). Gender did moderate the effect of the combined laptop + smart display group versus control group (b = 1.91, CI 0.11 to 3.71, p = .037). Women in the combined laptop + smart display group showed a significant decrease in pain interference over time (b = -0.69, CI -1.29 to -0.10, p = .022); those in the control group showed no significant change (b = 0.25, CI -0.53 to 1.04, p = .53). Men in the combined laptop + smart display group showed a nonsignificant decrease (b = -0.67, CI -1.47 to 0.14, p = .10); those in the control group showed a significant decrease (b = -1.61, CI -2.88 to -0.35, p = .013). Participants assigned to the laptop (vs smart display) used ElderTree more often and had more favorable perceptions. However, those in the smart display group used it for more hours. Analyses of secondary and exploratory outcomes showed no significant differences between the laptop and smart display groups or between the combined laptop + smart display group versus control group.

Conclusions: We found no significant differences between the combined ElderTree group and the control group on changes over time for any of the primary, secondary, or exploratory outcomes. The moderation analyses indicated that only gender moderated the effects of study arm, and only for the effects of laptop + smart display vs control group on changes over time in the two primary outcomes.

Clinicaltrial: ClinicalTrials.gov NCT04798196.

International registered report: RR2-10.2196/37522.