1 - (SYM 99) Towards Personalizing Medication Adherence Support and App Engagement Strategies Among Adolescents and Young Adults with Cancer

Background: Medication adherence is essential for optimal treatment response and influenced by psychosocial factors that can fluctuate within individuals and across days (e.g., coping with symptoms). While digital health tools are promising for increasing access to adherence interventions between face-to-face clinical encounters, “one-size-fits-all” approaches have not yielded robust effects. Moreover, suboptimal user engagement is a pervasive implementation challenge that threatens the real-world effectiveness of digital tools, including among adolescents and young adults (AYAs).

Methods: With AYAs with cancer, we have been developing an open-source app called ADAPTS that will ultimately deliver a just-in-time adaptive intervention (JITAI) to promote oral chemotherapy adherence while maintaining app engagement. This has involved two formative steps: 1) a 6-month ecological momentary assessment (EMA) study to pre-screen potential tailoring variables with n=28 AYAs with cancer (M Age=17, 36% Black, 14% Hispanic) and n=26 caregivers, with an embedded micro-randomized trial of low-cost engagement strategies (e.g., memes, altruistic messages), and 2) human-centered design methods to co-design the JITAI with n=10 AYAs and 4 caregivers from the EMA study, and 4 AYA survivors who formed a research advisory board. 

Results: Emerging data shows that AYAs and caregivers completed an average of 67% and 71% of daily surveys, respectively. When AYAs demonstrated a 2-unit increase in their average daily fatigue, there was 3.34 greater odds of missing their oral chemotherapy that day (p=.006). Our human-centered design process revealed preferences for positively-framed and tailored adherence messages (nudges toward adherence by linking it to context, life goals, and promoting remission rather than preventing relapse) and inspirational engagement strategies (quotes and altruistic content preferred over funny memes).

Conclusions: The results suggest that sending an adherence-promoting message when an AYA is especially fatigued may be an important consideration for a future machine learning algorithm. Aligned with this finding, AYAs desire messages that are tailored to their day-to-day experiences. Inspirational content was viewed as important for both medication adherence and app engagement. If accepted, we will present visual prototypes of our JITAI app design and results from the micro-randomized trial of app engagement strategies (data collection recently completed and analyses forthcoming).