Predicting Flu Outbreaks with Wearable Devices

Data published in the American Journal of Public Health found a compelling application for a wearable thermometer developed by Thermia – outbreak surveillance for influenza. Investigators at Harvard’s School of Public Health and Boston Children’s Hospital mined Thermia for reports of fever and other flu-like symptoms in China and found that Thermia was able to predict outbreak activity one month earlier than the country’s national influenza surveillance system. The Chinese health service tracks flu cases by counting cases diagnosed by a health provider at a clinic or hospital, which can create a lag in confirming cases and centrally aggregating the results; Thermia can provide the data in real-time. Google attempted something similar by using search queries to predict flu incidence (essentially predicting that searches for flu-associated terms would correlate to actual cases), but this program missed the peak of the 2013 flu season by 140% and was quietly cancelled thereafter. Thermia is different because it collects actual clinical symptoms, either through a mobile app or automatically with the use of the iThermonitor, an FDA-cleared, patch-like thermometer that is worn under the arm.

Commentary on the Boston Children’s website notes that there are 620 million mobile internet users in China, which could facilitate the collection of massive quantities of data – including public health data. The article reports that while the Thermia app and the iThermonitor are available worldwide, China has been one of the biggest adopters of the technology.