and

Heather Comerci Reyes

Implementation of data-driven technology across the care continuum unquestionably enables clinicians to improve care quality and experience. Patient engagement and participation in self-management of wellness will require remote monitoring tools for longitudinal monitoring of one’s own health status.

The perfect monitoring system would continuously capture complete data representative of the “whole person” using unobtrusive and secure means, interpret the data, and suggest intervention at the point of highest yield for the patient. While we may easily point out the shortcomings of current systems’ ability to monitor and react to swathes of biometric data, such systems more importantly fail to capture perhaps the most vital aspects of a patient’s experience – namely the psychological and social factors. These experiences are certainly more difficult to quantify, but both indicate and influence a patient’s wellbeing as equally as their heartrate and body temperature.

In 1977, George Engel, Professor of Psychiatry at the University of Rochester proposed the biopsychosocial model, changing the way clinicians looked at health and disease. A revolution in medical thinking at the time, the biopsychosocial model provided the argument that the development of illness is the result of the interplay between biological, psychological and social factors. For example, clinicians understand that to help a patient with the biological goal of weight loss, hormones and caloric intake are key factors, but without understanding the social constructs and psychological stressors at play, behavior change is nearly impossible. The most effective weight loss programs employ combinations of counseling and medical intervention, monitoring not just the biological indicators of success, but also the sustainable nature of behavior change. To monitor this process remotely, we would fall short if we only detailed easily objectified factors such as weight change and indicators of diabetes. Monitoring systems ubiquitously make this shortcoming and will continue to provide incomplete data until we are better able to quantify the other factors driving our patients’ wellbeing.

Michael Hasselberg, PhD, Chief Digital Health Officer, University of Rochester Medical Center

The industry is beginning to envision what remote social and psychological monitoring could look like. The mainstay of such efforts revolves around use of mobile apps and questionnaires, querying mood and activities of daily living. They are intrusive and obtrusive, require active attention multiple times a day to provide non-continuous indicators of wellness, and are biased by the patients’ willingness to report truthfully. Outcomes are often limited by lack of evidence-based measurement and scalability of monitoring and intervention. The required active engagement means monitoring is only as complete as the commitment of the individual to participate.

As we approach 50 years since Engel’s breakthrough revolution, the University of Rochester continues work to understand the whole patient through a biopsychosocial lens. Researchers have teamed with Embodied, Inc. to study the use of Moxie, an artificial intelligence (AI) friend for childhood companionship and social and emotional learning. The robot, Moxie uses AI including computer vision and natural language processing to interact with a child, whose

"The industry is beginning to envision what remote social and psychological monitoring could look like."

mission as their new “mentor” is to teach Moxie what it means to be a good friend to humans. In its current form, Moxie and the child mentor work through social and emotional intelligence adventures, learning together. Moxie elicits important information from the child, including their mood, fears, and relationships. Learning through play encourages daily return and ongoing engagement. Children seek out Moxie throughout the day, and progress can be monitored through a mobile parent app.

Imagine a world where an embodied character like Moxie can talk to a patient throughout the day. A child with a new diagnosis of diabetes could meet Moxie during hospitalization and learn healthy ways to cope and socially interact with others, in addition to biological information like the importance of blood sugar monitoring and insulin injections. At home, the child is happy to teach their AI friend, Moxie, about the ways their diagnosis has changed their routine at school, how frequently they are checking blood sugars, the barriers to counting carbohydrates, and how they are explaining their diagnosis to friends and family. Imagine the complete picture of wellness available to a clinician when available data extends beyond the biological to include the psychosocial factors impacting a patient’s relationship with their diagnosis.

Until we are able to monitor the status of our patients’ psychological and social wellbeing as thoroughly as their indicators of biological wellness, remote patient monitoring will continue fall short of our expectations. In order to accomplish such a goal, we need to rethink not only how we measure and define wellness, but also how our patients are most willing to share this information and the tools we use to build a bridge between digital health and wellbeing.