Virtual AppLication-supported ENvironment To INcrease Exercise During Cardiac Rehabilitation Study

The VALENTINE Cardiac Rehabilitation Study will evaluate a digital intervention to supplement cardiac rehabilitation for low- and moderate risk patients, integrating a mobile application with physiologic and contextual information from wearables to provide incremental support to patients. The study will be performed in conjunction with planned home-, hybrid-, or center-based cardiac rehabilitation at Michigan Medicine. Participants will be randomized to the control or telehealth group. Both groups will receive a smartwatch and usual care. Participants in the telehealth arm will additionally (1) have access to activity tracking and goal setting through the VALENTINE app; (2) receive micro-randomized, contextually tailored notifications, and (3) receive weekly activity summaries via email, which will also be provided to their exercise physiologist while enrolled in cardiac rehabilitation. Contextually tailored notifications will be one of two types: walking and exercise notifications. Walking notifications are designed to encourage participants to be active though at lower than their target heart rate zone. Participants will receive 1 notification per day, on average, at one of four times. Notifications will be randomized on four dimensions of context. Exercise notifications are designed to encourage participants to exercise within their target heart rate zone. Notifications will be delivered each evening at a time of the participant's choosing. Participants will have a 50% probability of receiving a notification each evening. Notifications will be randomized on two dimensions of context. Participants will be followed for 6-months. Participants in both arms will be asked to complete a 6-minute walk test at baseline, 3-months, and 6-months using their mobile phone and smartwatch. They will additionally complete general and disease-specific quality-of-life questionnaires at baseline and at 6-months.

To account for potential measurement error between devices, for primary and secondary analyses, we will perform regression analysis to jointly test the null hypothesis H0: B0(F) = B0(A) = 0 for change between baseline and 6 months where (F) and (A) refer to Fitbit and Apple watches respectively. A secondary analysis will subsequently be performed to determine whether to reject the individual null hypotheses for each of the two devices. For all statistical analyses, the level of significance will be set at p <0.05.