U of I research shows "smart" thermometers could predict flu outbreaks earlier

By  | 

IOWA CITY, Iowa (KCRG TV9) -- This current flu season is likely to go down as one of the worst in recent years. But could early warning of how the flu is spreading, in real time, help doctors do more to disrupt outbreaks?

A smart thermometer made by Kinsa, Inc. on Thursday, Feb. 8, 2018 U of I researchers found information link from the thermometer to a smart phone application could be used to predict the spread of flu activity. (Dave Franzman/KCRG-TV9)

University of Iowa researchers just published a study that analyzed eight million readings from one company’s “smart” thermometers. They found it’s statistically possible to forecast the spread of the flu two or three weeks earlier than other methods.

University of Iowa researchers cooperated with Kinsa, Inc—a manufacturer of smart thermometers that collect data on temperatures, symptoms and other information through a smart phone application. Customers who allow the data to be collected by the company were not identified in the study.

Aaron Miller, a postdoctoral scholar in computer science and one of the study’s authors, says medical providers typically rely on the Centers for Disease Control (CDC) to gather statistics on the Influenza outbreaks and figure out just where the flu is hitting and how hard.

And right now, that takes some time.

“On Friday, there’s going to be a new (flu) report published by the CDC and it will give us a sense of Influenza activity. And that’s actually going to be delayed by 12 days when it’s released,” he said.

Miller and Dr. Philip Polgreen, an associate professor of internal medicine and epidemiology, co-authored the study for the University of Iowa. They wondered if the information from the smart thermometers, stored by the smart phone app and sent to the manufacturer, could tell them more about what the flu is doing sooner.

They studied readings from 450,000 devices from all 50 states over a two and a half year period and found the information tracked what the CDC would report several weeks later about the spread of the flu.

For doctors, and public health workers, an early warning about the spread of the flu could open up a lot of possibilities.

Dr. Alicia Gerke, a U of I Pulmonary Critical Care physician, explained how doctors might use the information.

“Having more real-time prediction about when influenza might occur might motivate more people to get vaccinated or seek health care earlier. You can use appropriate antivirals at the right time and avoid unnecessary antibiotic use,” Dr. Gerke said.

Researchers also say early flu warnings might prompt medical clinics to increase staffing if they see a rise in fevers reported by the smart thermometers. They could also isolate some patient populations sooner to try and prevent the spread of Influenza. It may also be possible to zero in on flu-like activity by looking at the data on a regional basis or even school by school.

The research was published in the online medical journal Clinical Infectious Diseases on Thursday.