Smart thermometer, app provide real-time influenza tracking
Temperature readings from devices help to accurately predict flu levels up to three weeks into the future.
Using a "smart” thermometer connected to a mobile phone app, University of Iowa researchers have successfully demonstrated that they can track influenza activity in real time at both population and individual levels, and in the process accurately predict flu activity as much as three weeks in advance.
The Centers for Disease Control and Prevention monitors influenza-like illnesses (ILI) in the United States by gathering information from physicians’ reports about patients with these illnesses who seek medical attention—one of the most widely used methods for tracking flu. While CDC’s data provides useful estimates of influenza activity, there’s a one- to two-week delay in release of formal reports.
“Using simple forecasting models, we showed that thermometer data could be effectively used to predict influenza levels up to two to three weeks into the future,” says Aaron Miller, a University of Iowa postdoctoral scholar in computer science. “Given that traditional surveillance systems provide data with a lag time of one to two weeks, this means that estimates of future flu activity may actually be improved up to four or five weeks earlier.”
Also See: Northwell Health automates efforts to fight flu epidemic
UI researchers analyzed de-identified data from FDA-approved Kinsa Smart Ear Thermometers and an accompanying app, which recorded more than 8 million temperature readings from users generated by almost 450,000 unique devices from August 2015 to December 2017.
Results of their study, published online on Thursday in the journal Clinical Infectious Diseases, show that researchers were able to make accurate predictions of influenza activity that matched subsequent CDC reporting. In fact, Miller contends that early on he and his colleague Philip Polgreen, MD, UI associate professor of internal medicine and epidemiology, predicted that the 2017–2018 flu season was going to be much more severe than in the past few years.
“There are two main benefits of data coming from smart thermometers,” observes Miller. “The first is real time advanced warning of flu outbreaks and how severe a flu season will be. The second benefit of smart thermometers is that we actually have a platform for public health interventions. You can imagine in the future public health alerts that could appear on a device and give users real-time feedback on how severe the flu season is in their area, and provide reminders of the importance of vaccination and where they can get a flu shot.”
According to the CDC, Influenza-related hospitalization rates so far in the 2017–2018 season are exceeding milestones set during the 2014–2015 season.
“Smart thermometers represent a novel source of information for influenza surveillance and forecasting,” conclude the authors, who included Kinsa’s Inder Singh and Erin Koehler. “Not only do thermometer readings capture real-time ILI activity at a population level, they can also be used to generate improved forecasts. Moreover, the widespread deployment of these smart thermometers may also allow for more rapid and efficient surveillance at the household level.”
In the study, temperature readings were gathered from all 50 states and were aggregated to provide region and age-group specific flu activity estimates. Going forward, Miller says he and Polgreen are working on the ability to make predictions at the state level, as opposed to national and regional levels.
Neither Miller nor Polgreen have a financial relationship with Kinsa, the San Francisco-based manufacturer of the smart thermometers.
The Centers for Disease Control and Prevention monitors influenza-like illnesses (ILI) in the United States by gathering information from physicians’ reports about patients with these illnesses who seek medical attention—one of the most widely used methods for tracking flu. While CDC’s data provides useful estimates of influenza activity, there’s a one- to two-week delay in release of formal reports.
“Using simple forecasting models, we showed that thermometer data could be effectively used to predict influenza levels up to two to three weeks into the future,” says Aaron Miller, a University of Iowa postdoctoral scholar in computer science. “Given that traditional surveillance systems provide data with a lag time of one to two weeks, this means that estimates of future flu activity may actually be improved up to four or five weeks earlier.”
Also See: Northwell Health automates efforts to fight flu epidemicUI researchers analyzed de-identified data from FDA-approved Kinsa Smart Ear Thermometers and an accompanying app, which recorded more than 8 million temperature readings from users generated by almost 450,000 unique devices from August 2015 to December 2017.
Results of their study, published online on Thursday in the journal Clinical Infectious Diseases, show that researchers were able to make accurate predictions of influenza activity that matched subsequent CDC reporting. In fact, Miller contends that early on he and his colleague Philip Polgreen, MD, UI associate professor of internal medicine and epidemiology, predicted that the 2017–2018 flu season was going to be much more severe than in the past few years.
“There are two main benefits of data coming from smart thermometers,” observes Miller. “The first is real time advanced warning of flu outbreaks and how severe a flu season will be. The second benefit of smart thermometers is that we actually have a platform for public health interventions. You can imagine in the future public health alerts that could appear on a device and give users real-time feedback on how severe the flu season is in their area, and provide reminders of the importance of vaccination and where they can get a flu shot.”
According to the CDC, Influenza-related hospitalization rates so far in the 2017–2018 season are exceeding milestones set during the 2014–2015 season.
“Smart thermometers represent a novel source of information for influenza surveillance and forecasting,” conclude the authors, who included Kinsa’s Inder Singh and Erin Koehler. “Not only do thermometer readings capture real-time ILI activity at a population level, they can also be used to generate improved forecasts. Moreover, the widespread deployment of these smart thermometers may also allow for more rapid and efficient surveillance at the household level.”
In the study, temperature readings were gathered from all 50 states and were aggregated to provide region and age-group specific flu activity estimates. Going forward, Miller says he and Polgreen are working on the ability to make predictions at the state level, as opposed to national and regional levels.
Neither Miller nor Polgreen have a financial relationship with Kinsa, the San Francisco-based manufacturer of the smart thermometers.
More for you
Loading data for hdm_tax_topic #better-outcomes...