Development of a Voice-Activated Virtual Assistant to Improve Insomnia Among Young Adult Cancer Survivors: Mixed Methods Feasibility and Acceptability Study.

Abstract

Background: Up to 75% of young adult cancer survivors (YACS) experience chronic insomnia, negatively affecting physical and emotional health and overall quality of life. Cognitive behavioral therapy for insomnia (CBT-I) is a gold-standard intervention to address insomnia. To improve CBT-I access and treatment adherence, screen-based digital CBT-I platforms have been developed. However, even with these digital products, widespread uptake of CBT-I remains limited, and new strategies for CBT-I delivery are warranted.

Objective: The objective of this study is to understand how YACS experience insomnia and how they might incorporate technology-delivered CBT-I into a daily routine and test the feasibility and acceptability of a novel screen-free voice-activated virtual assistant-delivered CBT-I prototype.

Methods: Eligible participants-ages 18-39, living with a history of cancer (any type, any stage), self-reporting on average less sleep than National Sleep Foundation recommendations, and English-speaking-were recruited from a major urban cancer center, 2 regional oncology clinics, and 2 cancer survivorship support groups. We conducted 4 focus groups to understand the YACS experience of insomnia, their routine use of technology at home, particularly voice-activated virtual assistants such as Amazon Alexa, and input on how CBT-I might be delivered at home through a smart speaker system. We developed a prototype device to deliver key elements of CBT-I at home along with circadian lighting and monitoring of post-bedtime device use, collected YACS user perspectives on this prototype, and then conducted a single-arm feasibility and acceptability study.

Results: In total, 26 YACS (6-7 participants per group) experiencing insomnia participated in focus groups to share experiences of insomnia during cancer survivorship and to provide input regarding a CBT-I prototype. Common triggers of insomnia included worry about disease management and progression, disease-related pain and other symptoms, choices regarding personal device use, and worry about the impact of poor sleep on daily functioning. In total, 12 participants completed device prototype testing, engaging with the prototype 94% of the assigned times (twice daily for 14 days; meeting predetermined feasibility cutoff of engagement ≥70% of assigned times) and rating the prototype with an overall mean score of 5.43 on the Satisfaction subscale of the Usability, Satisfaction, and Ease of Use scale (range 4.42-7; exceeding the predetermined cutoff score for acceptability of 5.0). All participants completing the study reported they would be interested in using the prototype again and would recommend it to someone else with insomnia.

Conclusions: YACS were highly engaged with our voice-activated virtual assistant-delivered CBT-I prototype and found it acceptable to use. Following final device development, future studies should evaluate the efficacy of this intervention among YACS.

Trial registration: ClinicalTrials.gov NCT05875129; https://clinicaltrials.gov/study/NCT05875129.