Network / Agile Projects: These projects are conducted by small teams of network members. For projects focused on questions related to personal health data (PHD), we believe agile research methodologies better fit the temporal rhythms of this fast-changing field. An agile development approach offers the research Network an opportunity to make incremental investment to open new research opportunities, create open infrastructure and data sources for the research community, and help create new training opportunities for the field.
From Self-Monitoring to Self-Experimentation: Behavior Change in Patients with MS
PI: Emil Chiauzzi (PatientsLikeMe) Co-PI: Eric Hekler (Arizona State University)
Co-PI: Carlos Rodarte (PatientsLikeMe)
Wearable devices for self-tracking are often used with a focus on personal health data collection, but it is even more important to leverage such devices in a companion behavior change program. Eric Hekler and colleagues have recently advanced the concept of a DIY self-experimentation toolkit, which enables users to develop individualized behavior change plans based on principles such as positive reinforcement, stimulus control, and self-reward. This pilot study adapts this toolkit and tests the feasibility of a behavior-change course for wearables with multiple sclerosis (MS) patients. The ultimate goal is to enable a broad range of chronic disease patients to build behavior-change programs using wearables.We will first conduct 60-minute interviews with six MS patients recruited from a previous PLM wearables (FitBit OneTM) study sample. We will assess patients use of behavior change techniques, ways in which they integrate wearables into their lifestyles, and their recommendations for the behavior change course. Next, we will adapt the self-experimentation toolkit to develop a brief Wearables 101 course that can be delivered by a trained staff member (e.g., research assistant) in a 1:1 PowerPoint presentation. Finally, we will test the program with 20 MS patients from the previous PLM study. Participants will be asked to complete the course, use their FitBit OneTM and implement the behavior change plan for one month, and complete measures of activity and experiences with the course. This data and the wearable data (e.g., steps) will be analyzed to determine the feasibility of the course.
Keeping Pace: Dynamic Assessment of Environment and Exercise Using Personal Health Data
PI: Rumi Chunara (New York University)
How does the environment around us support or inhibit our healthy behaviors Past studies have aimed to understand how the environment relates to preventative behaviors such as exercise, but have been limited by cost and labor-intensive formats involving hours of phone calls, clinic or home visits and surveys used to measure those behaviors. New efforts from health startups to large corporations reveal a massive investment in tools that measure physical activity. Accordingly, Internet and mobile connectivity potentially offer a new opportunity for behavior data collection with extremely fine spatio-temporal resolution that can also evade survey recall biases. Thus in this pilot study, we will demonstrate and learn from the use of new personal sensor data in aggregate to, inform our understanding of how the relation between the built environment and types and amounts of exercise varies over time. Our work is pertinent to the broad community of stakeholders in the personal health data domain because it brings together and leverages expertise from multiple partners and Health Data Exploration Network members: companies and industry members or others producing data, academics who are interested in using data for research, other groups interested in improving health outcomes in general, and importantly, individual members of the public. In addition to the specific examination of environment and preventative health behaviors, overall our efforts will also help understand how public engagement can improve healthcare.
Towards Privacy-Aware Research and Development in Wearable Health”
PI: Michelle DeMooy (Center for Democracy and Technology)
The Center for Democracy & Technology (CDT) seeks to develop detailed guidance for entities that collect health and wellness data on how to conduct internal research in a manner that honors the privacy and dignity of their user population. Working in partnership with one such entity, Fitbit, we will produce guidelines on privacy-protective internal research for companies using consumer-generated health data, and provide recommendations on providing customer and public benefit through research activities. In the long term, we will use these efforts to validate for companies, clinicians, and the general public the concept that responsible and ethical research using personal health data (PHD) via wearable devices can produce interesting and valuable insights on wellness and the quantified self.
Exploring Strategies to Improve Acceptability and Usability of a Just In Time Adaptive Intervention via Incorporation of Proximity Sensors and a Smartwatch
PI: Eric Hekler (Arizona State University)
Mobile technologies have the capacity to track factors that are important to health like being physically active and provide customized interventions exactly when and where it would be most beneficial for each person. While this possibility exists in theory, in practice it is very hard to know when the right time is for delivering these sorts of messages. Emerging technologies such as the smartwatch (e.g., Apple Watch) and indoor location sensors that can track where a person is inside buildings provide better opportunities to identity those exact moments when a person might both benefit from and be receptive to messages to help them achieve their personal goals. For example, a person could set up an app to keep track sitting and watching TV in his living room and then have his smartwatch vibrate and provide a motivational image at the right moment (e.g., during a commercial break/right before a new show turns on) to nudge the person to go for a walk; a behavior the person wants to do more. The purpose of this work is to examine how best to use these indoor location sensors and smartwatches to help individuals achieve their own personal health behavior change goals that explicitly balances the desire for privacy with the desire for the right help when it would be most useful.
When am I at my best? Passive sensing of circadian rhythms for individualized models of cognitive performance
PI: Julie Kientz (University of Washington)
Inter-individual differences such as age, gender, genetic variations, and personality can substantially impact the optimal time of day for performing cognitively demanding tasks i.e., the circadian rhythm of cognitive performance. Such differences can also affect individuals? abilities to physiologically and cognitively cope with sleep deprivation and irregular sleep, which have negative impacts on cognitive performance as well. In this work, we are exploring the opportunity of personal data to better understand, unobtrusively model, and reliably predict individual characteristics and patterns related to circadian variables, cognitive performance, and sleep. More specifically, we are conducting the following activities: 1) Capturing smartphone interaction patterns, web usage behaviors, and online content to develop novel user-modeling techniques capable of handling this array of big data reliably and at scale; 2) conducting a 3 week study that incorporates well-established objective and subjective measurements in order to robustly evaluate our approach to passively sense an individual’s sleep and cognitive performance, factoring in biological rhythms and external factors such as caffeine; 3) advancing our understanding of the value of personal data streams in modeling health-related characteristics and behaviors by reflecting upon the daily trends in performance that our data and approach are able to effectively model as well as our failure points; and 4) delivering contributions relevant to diverse stakeholders for instance by providing insights to behavioral scientists interested in better understanding the interplay between technology usage and health behaviors or by providing recommendations for circadian-aware behavioral intervention systems that can be pursued by designers and system builders and eventually adopted by real users.