Answers are grounded in this paper and cite their sources.
Analysis reveals mHealth engagement influences healthcare utilization in patients with cardiovascular disease, highlighting key physiological predictors.
BACKGROUND While prior research has explored the use of software applications such as patient portals to study patient engagement with healthcare providers, there are major gaps in our understanding of the impact of patient engagement with mobile health (mhealth) technologies on their utilization of healthcare resources. It is unclear whether physiological data collected from mobile wearable devices can be used to predict future health encounters of patients across different modes of healthcare delivery. OBJECTIVE This study evaluates the viability of a novel mhealth infrastructure developed to analyze cardiovascular disease outcomes among patients by empowering them to embrace behaviors that can improve cardiovascular health through utilization of innovative digital health strategies. METHODS Patients with a history of coronary artery disease (CAD) were recruited from the cardiology clinic at a large academic medical center in Texas. Twenty patients were trained on the use of Garmin devices by the study coordinator, prior to data collection. The data collected from their use of wearable devices was analyzed together with data from their electronic health records (EHR), including their pattern of health encounters and utilization of healthcare services. We tracked data on patient health outcomes over a two-year period which includes approximately 1,500 health encounters that span patients’ hospital and outpatient visits, telehealth visits, as well as other visits such as prescription refills, procedural visits, and infusions. We used statistical regression models to evaluate the relationships between patient physiological data and their frequency of different types of health encounters. We also used machine learning methods to predict the frequency of hospital and outpatient visits, and evaluated the importance of different types of physiological data on model prediction. RESULTS Results indicate that patients exhibit significant reduction in their weekly outpatient visits during the period of the study, when compared to the 18-month period prior to study recruitment. Patients also exhibit a significant increase in medication refills and medical procedures during the study period. Study participants were more likely to use the wearable devices to monitor their health conditions, which in turn, was associated with lower frequency of future outpatient visits, but also incurred greater utilization of other healthcare services, such as procedures and prescription refills. CONCLUSIONS Our results indicate a significant association between the use of mhealth wearable devices and patient engagement based on their pattern of utilization of healthcare resources, after controlling for patient-specific characteristics. Specifically, they provide evidence of a significant association between patients’ physiological data, such as minimum heartbeat rate, active time, maximum stress level, and total sleep time, and their frequency of future health encounters. Based on prediction models developed using random forest methods, these types of physiological data are identified as important features in predicting hospitalizations and outpatient visits. CLINICALTRIAL Not Applicable
No takes yet. Share an insight, caveat, or question.
Bardhan et al. (2025) studied this question.
Synapse has enriched 8 closely related papers on similar clinical questions. Consider them for comparative context: