Fitness tracker devices show potential as predictors of flu outbreaks

By Sally Robertson, B.Sc.Jan 17 2020

Researchers at Scripps Research Translational Institute have found that fitness tracker devices – wearable tracker devices that measure heart rate and sleep duration – are better predictors of real-time flu outbreaks than current surveillance methods.

Image Credit: Syda Productions/Shutterstock.com

A study that assessed data available for more than 47,000 fitness tracker devices users across five U.S states, showed that the devices improved and accelerated the prediction of flu at the state level.

As well as helping people to tell whether they are coming down with the flu, the data could warn health authorities that they need to get ready to help, potentially increasing response time to outbreaks.

"Responding more quickly to influenza outbreaks can prevent further spread and infection, and we were curious to see if sensor data could improve real-time surveillance at the state level," says study author Jennifer Radin.

In the United States, about 7% of working adults and one-fifth of children aged under 5 years are infected with the flu every year and, according to estimates from the World Health Organization, the infection kills as many as 650,000 individuals globally every year.

Current surveillance reporting

Current surveillance methods to predict ILI rates have mainly used data reported to the Centers for Disease and Control (CDC), but this is often delayed by at least 1–3 weeks, and data is often revised months later.

This delay means response measures such as deploying vaccines or anti-viral medications, are often slow and outbreaks quickly spread to new vulnerable populations and geographical regions.

Previous studies have tried to use crowdsourced data such as Google Flu Trends and Twitter to provide real-time influenza-like illness (ILI) information - a method referred to as “nowcasting.”

However, these approaches result in variable success, especially at the state level, partly because they are influenced by outside factors. For example, media coverage of influenza alerting people to an outbreak makes it impossible to separate the behavior of people who have the flu from the so-called “worried well” who search for information online during epidemic periods.

“Use of sensor-based data would offer the first objective and real-time measurement of illness in a population that could potentially reduce the effect of overestimation during epidemics,” said Radin and team. The ability to harness this data at a large scale “might help to improve objective, real-time estimates of ILI rates at a more local level, giving public health responders the ability to act quickly and precisely on suspected outbreaks,” they add.

What did the study involve?

Sleep and resting heart rate (RHR) are likely to differ from the norm in response to acute infection, especially when it is accompanied by fever: “When someone is unwell, their RHR increases, their total sleep is likely to increase, and their activity is likely to decline,” explain the authors.

The team, therefore, set out to evaluate whether population trends of seasonal respiratory infections such as influenza, could be identified through wearable sensors that collect RHR and sleep data.

To our knowledge, this is the first study to evaluate the use of RHR and sleep data in a large population to predict real-time ILI rates at the state level.”

The team de-identified sensor data from 200,000 people who used the same wearable fitness tracker, for at least 60 days between March 2016 and March 2018 and focused on the top five states with the most fitness tracker users in the dataset, which were California, Texas, New York, Illinois, and Pennsylvania. The average age of participants was 43 years and 60% were female.

After excluding daily measurements with missing RHR, missing wear time, and wear time less than 1000 min per day, the team identified 47,249 users in the top five states who wore a fitness tracker consistently during the study period, including more than 13·3 million total RHR and sleep measures.

A fitness tracker user’s’ resting heart rate and sleep duration were considered abnormal if their average weekly heart rate was higher than their overall average, and their weekly average sleep was not lower than their overall average.

The sensor data was compared with weekly estimates of influenza-like illness (ILI) rates at the state level, as reported by the CDC.  

What did the results show?

As reported in The Lancet Digital Health, the fitness tracker data significantly improved ILI predictions across all five states, with the authors citing improvement of 6% to 33% over baseline models.

In most cases, week-to-week changes in the proportion of people with abnormal data were associated with week-to-week changes in ILI rates.

“This study shows that using RHR and other metrics from wearables has the potential to improve real-time ILI surveillance,” said Radin and team.

“This information could be vital to enact timely outbreak response measures”

The use of activity and physiological trackers is becoming increasingly popular in the United States and globally. The authors say that by accessing these data, it could be possible to improve real-time and geographically refined influenza surveillance. Moreover, it might be possible to identify ILI rates on a daily, instead of weekly, basis, providing even more timely surveillance, they add.

“This information could be vital to enact timely outbreak response measures to prevent further transmission of influenza cases during outbreaks,” concludes the team.