Fitbit data could help predict flu outbreaks in real time

January 22, 2020

That Fitbit on your wrist could be doing a lot more than tracking the strides you make each day: Researchers at the California-based Scripps Research Translational Institute reviewed de-identified data from 200,000 users of Fitbit exercise and activity trackers in five states—and found that they were able to use data like rising heart rates and changes in sleep patterns to predict flu outbreaks in real-time, according to a report by CNN.

Indeed, the scientists were able to calculate the proportion of users falling above set thresholds for average heart rate and sleep duration—and to compare that data to weekly flu rates determined by the Centers for Disease Control—in order to predict flu outbreaks in real time.

The finding shows the potential for the soon-to-be Alphabet-owned brand to predict disease outbreaks —which could open an opportunity to propel Google-sister company Verily’s population health efforts:

With the flu affecting an estimated 35.5 million and driving 490,600 hospitalizations in the US in the 2018-2019 flu season alone, according to the CDC, the ability to predict outbreaks would be welcomed by an already overburdened healthcare system. And the potential savings could be significant: During the 2015-2016 U.S. flu season, an estimated $10.4 billion was spent on direct costs for adult hospitalizations and outpatient visits, according to CNBC.

And should Alphabet get the regulatory go-ahead for its Fitbit purchase, the potential to predict disease outbreaks would be a huge value-add to Verily’s population health efforts.For example, CNN suggests, “We could see Verily integrate health data collected from Fitbit users into its Project Baseline initiative, which is aimed at developing technologies to help researchers architect a map of human health and gain a deeper understanding of prevalent conditions like Parkinson’s Disease, rheumatoid arthritis, and inflammatory bowel disease.”

However, there are some flies in the proverbial ointment: While promising, the impact of the finding could be lessened due to the limited nature of the data collected — and Alphabet will need to be clear about its data-sharing policies or risk losing more consumer trust if it seeks a partner for future Fitbit endeavors, CNN notes.

What’s more, Fitbit users aren’t necessarily representative of the general population: For example, U.S. consumers who use wearables skew younger and tend to have higher incomes, as eMarketer noted in its Wearables 2019 report, which means there are likely gaps missing in the data collected.

Research contact: @CNN

Leave a Reply

Your email address will not be published. Required fields are marked *