By 2020, the total of amount of digital healthcare data worldwide is projected to exceed 2,000 exabytes (1 exabyte is equivalent to 1 billion gigabytes). Nearly unfathomable a decade ago, this explosion of information has the potential to revolutionize healthcare. By combining patient clinical data from electronic health records, genomics data, and wearables data, we can imagine a vision of the not-so-distant future that includes a healthcare system driven by personalized medicine.
As a doctoral candidate in the BIG IDEAS Lab in Biomedical Engineering and as the newest Duke Forge Predoctoral Scholar, I am incredibly excited about research in the space of personalized medicine, especially when it comes to chronic disease management.
According to the U.S. Centers for Disease Control and Prevention (CDC), 90% of the nation’s $3.3 trillion in annual health care expenditures are for people with chronic and mental health conditions. Not only is the economic burden of chronic diseases astronomically high, but there are social, behavioral, and psychological consequences both to individuals and their families. For this reason, addressing the prevention and management of chronic diseases is a critical priority in healthcare.
One area that shows an immense amount of potential for addressing the challenges of chronic diseases revolves around data gathered from “wearable” digital devices. Currently, approximately 60 million people in the US are using a total of 117 million wearable devices – a figure that’s expected to double in the next 3 years, according to eMarketer reports. This rapid increase in the accessibility of wearable devices and recent improvements in mobile health technologies provide an unprecedented opportunity to revolutionize chronic disease detection and intervention through the development of digital biomarkers, a kind of digitally collected data that provides an objective measurement of “...normal biologic processes, pathologic processes, or biological responses to therapeutic interventions.” Using “big data” analytical techniques and machine learning models, we can engineer, analyze, and model digital biomarkers to predict health outcomes and monitor health over time.
Another opportunity for the application of personalized medicine is in mobile application development using data from electronic health records (EHR) with Fast Healthcare Interoperability Standards (FHIR) standards. When integrated with patient input, these systems could provide a direct interface between patient and healthcare professional, allowing for increased accessibility to healthcare, prevention strategies, and management of disease. Combined with wearables data and digital biomarker development, a mobile app would help enable efforts to manage and prevent chronic disease in global underserved populations who have limited access to healthcare.
According to the U.S. Centers for Disease Control and Prevention (CDC), 90% of the nation’s $3.3 trillion in annual healthcare expenditures are for people with chronic and mental health conditions. Not only is the economic burden of chronic diseases astronomically high, but there are social, behavioral, and psychological consequences both to individuals and their families. For this reason, addressing the prevention and management of chronic diseases is a critical priority in healthcare.
In my doctoral work under Dr. Jessilyn Dunn in the BIG IDEAS Lab, I’m exploring these opportunities for advancing personalized medicine in chronic disease management. My thesis work focuses on engineering digital biomarkers of prediabetes from wearable devices and using machine learning to predict outcomes and monitor prediabetic states over time. Working with a clinical team in the Duke Endocrinology and Lipids Clinic, we are developing new tools for prediabetes management. Because prediabetes is reversible, it is the first step in preventing diabetes, as well as other chronic diseases such as heart disease and obesity.
Another project I am developing with a team of data science students is a monitoring mobile application for asthma management. Working with Dr. Ed Hammond from the Duke Center for Health Informatics, we used FHIR interoperability standards to access EHR data. Merging EHR data with patient input data, our mobile application tracks the health of patients with asthma over time and can make predictions of asthma health based on environmental data from drawn from application programming interfaces (APIs). Our mobile application was recognized as a finalist in the FHIR DevDays Hackathon competition, and we will be pitching our app at DevDays this summer.
Although this application is still under development, it marks the beginning of a wave of health data mobile applications. With key players in the field, such as Apple and Google, becoming more involved with health data, the entire mobile development space is moving toward health management. With this evolution will come new challenges related to data storage and security. However, health data mobile applications are poised to make significant contributions to the future healthcare system and expand access to tools that will let people better monitor their health status and manage their symptoms.
In the future, I hope to become even more involved with the initiative to develop a healthcare system driven by personalized medicine from big data. I’m honored to be a Forge Predoctoral Scholar, and I’m excited for the opportunities to collaborate with the passionate community of people at Duke Forge.
If you have an interest in my work, please follow me on my professional profiles: