JMIR Publications

ABSTRACT

Background:

Breast cancer (BC) is a significant public health burden. Despite its critical role in preventing the recurrence of BC, rates of long-term adherence to endocrine therapy (ET) remain low among certain breast cancer survivors (BCS). Utilizing embedded sensors in smartphones and wearables, ecological momentary assessment data, and health behavior theory may facilitate a richer understanding of the real world context of medication-taking behaviors, which can aid in the development of personalized interventions.

Objective:

The objective of the current study is to describe the development of a multiscale modeling intervention (MMI) system to facilitate adherence to daily oral ET for BCS.

Methods:

To develop the MMI system, we conducted usability interviews with 25 BCS prescribed ET, conducted a review of commercial wrist-worn and medication event monitoring system (MEMS) sensors, conducted a secondary analysis of ET adherence data from 32 BCS, and reviewed research literature of constructs and measurement of constructs associated with ET adherence among BCS. All usability interviews were audio recorded, transcribed, and summarized. Secondary data analysis included the use of randomized neural network (RNN) analysis to predict factors strongly associated with medication adherence.

Results:

Usability interview findings suggested that participants were willing to use an EMA smartphone application, a smartwatch and associated smartphone application, a smart pill bottle or smart pill box and associated smartphone application, and the entire MMI system for the duration of a six-month study period. Twenty-six commercially available wearable sensors and 18 pill MEMS devices were reviewed. Of the devices reviewed, the Fitbit Sense wrist-worn sensor and RxCap MEMS device were selected to be included in the MMI system. RNN analysis identified 104 survey items with significant contribution to four week medication adherence using a threshold of 70th percentile for feature importance. Literature review findings identified 42 surveys as predictors of ET medication adherence. When combined, 34 surveys were identified to be included in the MMI system. Due to participants' recommendations to shorten the survey, the final survey for the MMI system consisted of 32 surveys and demographic items.

Conclusions:

Overall findings suggest that participants were willing to use each component of the entire MMI system for the future six-month study duration (Phase 2). The study team identified 32 constructs that will be included in the baseline survey, 3-month survey, and 6-month survey for the MMI system. The next phase of the study will include deployment of the entire MMI system with 20 BCS. Our research highlights the use of theory, data-driven models, and participant feedback to inform development of a medication adherence monitoring system, which may lead to improved interventions for ET adherence.