SPECIAL SEMINAR
Monday, May 22, 2017
4:00 – 5:00 PM
177 Huntington Avenue, 9th Floor Shared Space
The use of real-world data (RWD) for
understanding health behaviors
Presented
by
Elina Helander, PhD
Postdoctoral Fellow
Tampere University of Technology, Finland
Wide-spread use of wearable and connected sensors enable collection of massive amounts of health related data from the daily lives of a large number of individuals in the wild. In this talk,
we will discuss the use of real world data (RWD) for research. RWD can be described as data collected under normal day-to-day circumstances outside of a traditional (randomized) clinical trial. RWD has potential to reveal new predictors for health outcomes,
aid in generating hypotheses for future RCTs, and training and testing models and allow development of timely interventions. However, by itself, RWD is meaningless; it must be analyzed within the context of its collection. Still, analysis and interpretation
of such data is not without challenges.
We discuss two cases of using RWD. The first addresses weight self-monitoring data of 10,000 connected weight scale users in free-living settings. We will show how we used the data to study
holiday weight gain in three countries (US, Germany, and Japan) resulting in article published in NEJM. In addition, we describe how adherence to weight self-monitoring affects weight change at the sample and individual level. The second dataset consists
of heart rate variability measurements of thousands of employees recorded over a few days in real-life settings. We will show how reported doses of alcohol affect heart rate and heart rate variability in a dose-dependent manner during the first hours of sleep
even in this non-controlled data.
At the end of the talk, we will describe a few examples of current ongoing research projects and interests, namely decision support systems for neonatal sepsis and neuro-physiological maturation
and Digital Health Revolution project for collecting a massively deep genomic, physiological and behavioral data for ~100 subjects.
All are welcome to join!