The Healthy Behavior Data Challenge

Searching for Tomorrow’s Health Surveillance Strategies

Innovators Participate in the Healthy Behavior Data Challenge

In their hunt for new ways to collect critical public-health information, US- and Canadian public health agencies have challenged companies, startups, academics, government organizations, and individuals to make the next big steps in health surveillance.

During the springtime Health Datapalooza 2017 that was held in Washington, DC, the Centers for Disease Control and Prevention (CDC), the Department of Health and Human Services, and the Public Health Agency of Canada announced the launch of The Healthy Behavior Data Challenge. This was a competition seeking innovative ways to collect precise wellness data from wearable devices, social media, and mobile applications—in order to monitor health and health-related behavior accurately and in real-time. These strategies could find new ways to provide public health experts and decision makers with reliable information they need to keep chronic diseases under control, prevent injury, and help everyone stay as healthy as they can be.

In the two-phase contest, participants first competed for $30,000 in prize money as they developed their prototype surveillance strategies. The most promising project proposals went on to develop and test their prototypes in Phase 2 of the Challenge, for $70,000 in prizes and CDC recognition.

At the start of 2018, the following participants started competing in Phase 2:

1. Chunara Lab

The Group: Chunara Lab develops the statistical and computational principles that transform unstructured data into a better understanding of everyone’s health. It teamed with Axis Maps, which designs and builds custom map software that clearly and effectively communicates data, bolsters outreach, and facilitates decision-making, with demonstrated public-health work experience.

The Idea: The proposed project, Keeping Pace, will aggregate physical activity data from a range of personal activity trackers and links to important information. An inviting, carefully designed interface will enhance communication with participants and data users. Direct traces of activity–captured with high levels of detail–will help avoid self-reported data-quality issues of under/over reporting and recall biases. Content and data representativeness can be assessed in real-time, allowing for group targeting and accounting for important events.

2. Temple University Institute for Survey Research

The Group: Temple University’s Institute for Survey Research (ISR) of Philadelphia, PA, is a full-service, academic survey center offering national, regional, and local survey capabilities. ISR utilizes phone, web, Short Message Service (SMS), audio computer assisted and in-person modes of data collection and specializes in working with urban and low-income populations.

The Idea: The proposed project, BeHeardBeHealthy, is designed to engage under-represented populations in an initiative of sharing and collecting behavioral health data. Team members will provide a platform with a robust set of application programming interfaces for direct connection to patient portals and personal devices, which will be used in a study of fitness device and mobile app users. The project can monitor a spectrum of behavioral risks and health status measures associated with the leading causes of preventable morbidity and premature mortality, as well as engage participants for more-detailed daily exercise and nutrition information.

3. RTI International

The Group: RTI International is an independent, 501(c)(3) nonprofit institute that provides objective research, development, and technical services to government and private-sector clients. RTI’s Digital Health and Clinical Informatics Program conducts research on patient and clinician use of technology to improve individual health, self-management, population health, and provider-based health care as part of health system transformation.

The Idea: The proposed project will demonstrate how the combination of self-generated data can enhance public health surveillance indicators to provide valid, high-fidelity measures of physical activity, sedentary behaviors, and sleep. The project will combine data extracted from recruited participants’ wearable devices, enabling a variety of individual- and cohort-level analyses not currently possible with BRFSS measures. The analyses range from descriptive statistics and temporal pattern recognition on the minute level to more complex, idiographic, multivariate assessments and modeling of physical activity, sedentary behaviors, and sleep in the sampled population.

4. Onlife Health

The Group: Onlife Health is a data-driven health solutions company. Its engagement platform empowers more than 11 million people to take an active and ongoing role in their own health and leverages sophisticated data analytics for understanding a population and engaging each of its members in meaningful programs.

The Idea: The proposed project will collect data on nutrition, sleep, physical activity and sedentary behaviors from fitness devices and mobile surveys. While fitness tracker data may not suffer from recall- or self-reporting bias, it can be incomplete, inconsistent, or lack context. This project will address these limitations with a series of short, just-in-time survey questions delivered via push notification to supplement mobile data sources. It will then use statistical analyses to cross-validate the mobile and survey data sets with each other, as well as with the BRFSS survey data, at the zip code or FIPS code level. The methodology also will reduce the human resource burden of survey administration, improve response rates, and also may permit larger-scale automated data collection without additional expense.

5. Social Etc.

The Group: Social Etc. consults with a variety of businesses and government agencies in the Washington, DC, metropolitan area. Staff focus on developing mobile applications for the iOS platform that provide value to users and clients.

The Idea: The proposed project, The Healthy Behavior Research App, will be a new way to track healthy behaviors, leveraging new and innovative activity planning, tracking, tasks and surveys to capture healthy behaviors from research participants and the general public. It relies on an Open Source Development Kit designed to capture user input and store it on the phone for later distribution to web-based data solutions. The project goal is to enhance US public health monitoring by creating a mobile application that is both useful to users wanting to lead a healthy lifestyle, as well as to researchers wanting to use data in a very large-scale study.

Please see the official challenge announcement for The Healthy Behavior Data Challenge for full details of the contest—including eligibility rules, how the participants entered the challenge, and scoring.