Georgia Tech Health Research Reveals Online Search Patterns, New Customized Software


April 22, 2014

By creating proactive, accessible health and wellness technologies, Georgia Tech has taken significant steps in enabling people to control their own health to greater, more effective degrees.

 “People are seeking ways to integrate health care practices into their lives and improve their quality of life, so we must design health interventions that fit seamlessly with current consumer technologies developed for daily use,” says Elizabeth Mynatt, Georgia Tech’s director of the Institute for People and Technology and professor in the School of Interactive Computing.

Highlighting Georgia Tech’s health research strategy are three projects focused on personal technologies and networked computing tools designed to radically change how users interact with health care information and systems. This research will be presented at the Association of Computing Machinery’s conference on human-computer interaction, CHI 2014, in Toronto, Canada, next week.

The research includes:

  • Seeking and Sharing Health Information Online, the first large-scale study comparing what medical conditions users looked for on search engines vs. social media. For example, the study revealed that cancer was the No. 1 web search topic while headaches was No. 1 on Twitter.
  • StepStream, a computing tool to help middle school-aged students who need more physical activity through a non-competitive program and social conversations around fitness. 
  • My Journey Compass, a health management tool designed for tablets and customized for cancer patients.

For the online health information study, Microsoft Research and Georgia Tech researchers compared and contrasted health activities on search engines and social media. Researchers identified 165 prevalent medical conditions, categorized them based on symptoms and severity and developed stigma levels for each. Analyzing Twitter and web search data from a 15-month period for the medical conditions, they gathered more than 125 million tweets from half that number of Twitter users and more than 174 million web searches from almost 40 million users. They also conducted a survey of 210 Internet users to identify motives for online health activity.

“Our findings indicate that there are considerable differences in health activity between the platforms,” says Munmun De Choudhury, the lead investigator and an assistant professor in the School of Interactive Computing at Georgia Tech. “But the complementary nature of the two media – public vs. private, seeking vs. sharing – helped develop a more complete picture of the range of online health activities.”

The research results showed that people prefer search engines when seeking information on serious medical conditions, disabilities and conditions known to bear social stigma, while Twitter is used more often to share information around symptoms of different health issues, and on conditions with benign explanations.

The StepStream program resulted from a Georgia Tech social computing project on adolescent obesity. This personal health tool integrates into youths’ daily lives through an after-school program and social network site to encourage fitness. The system eschews the direct competition and direct social comparisons of other systems and instead explores how to develop social support for physical fitness within a school context.

Deployed for four weeks in a middle school, StepStream tracked the step count of individual participants who wore pedometers. When a student reached his or her baseline step count (visible only to that student), they received activity points and the social website would post a congratulatory message public to other participants. Students could also send comments to each other on the site. The points were used to play StepStream’s online game, PuddleJump, where friends worked together to unlock levels by collecting stepping stones.

“Since health behavior change theories all focus on attitude change as a prerequisite to sustained behavior change, we were particularly interested to see how StepStream may have influenced students’ perceptions of their own abilities, comparison to other students and fitness in general,” says Andrew Miller, StepStream’s creator and a Ph.D. candidate in human-centered computing at Georgia Tech.

Students averaged 6,137 steps per day in the baseline period and 6,465 steps per day while using StepStream. While the overall effect on physical activity for all students was modest, StepStream appeared to help those most in need of elevated fitness levels. Those students, with baselines below the group average (the lower 50 percent), improved their daily step counts by an average of 1,088 steps per day while using StepStream.

My Journey Compass was designed by a Georgia Tech team to increase patient engagement for improving health outcomes and create a potential gateway to impact chronic disease management. Researchers tested the tool with a small group of breast cancer patients and included a suite of preinstalled applications and health management resources on Nexus 7 Android tablets.

Designed and deployed with health care providers, My Journey Compass enabled patients to be more confident and prepared in discussing their treatment, according to the study. Patients recorded conversations with physicians (to reference later or share with family), took notes, scheduled appointments, referenced health data and did other tasks to centralize their cancer management. The high adoption rate of the tablet tool directly correlated to its customization (for health and non-health purposes), mobile use, balance of information that was relevant and not overwhelming, and privacy advantages.

My Journey Compass gave patients total control of the device, resulting in unexpected uses, notably as a relaxation and escape tool. Patients added their own entertainment apps for various activities, including making visits to chemotherapy easier and routine daily use.

“By providing a tool that shifts between a health and personal device, participants over the long term may be more likely to return to using the tool for health purposes should the need arise,” said Maia Jacobs, a researcher on the study and Ph.D. student in human-centered computing at Georgia Tech.