Smart Technologies and Online Digital Data Promote Faster and Better Flu Tracking

Smart Technologies and Online Digital Data Promote Faster and Better Flu TrackingThe Centers for Disease Control and Prevention (CDC) tracks the spread of flu and illness with flu symptoms across the nation. Data is made available every Monday morning about the spread of the illness as well as hospitalizations in each state during flu season, from October through April. Medical transcription services help providers maintain accurate and timely electronic medical records, allowing public health organizations to quickly identify people who have flu symptoms and other illnesses that increase their risk for complications. A recent Health Tech Magazine report says that smart technologies are providing more timely predictions about the spread of the flu and assessments of seasons in progress. With these smart technologies, public health officials no longer have to wait for the weekly CDC estimate and can stay updated about the spread of flu as it happens.

Smart Thermometers: The CDC’s estimates of flu activity are based on data from hospitals and clinics that report on the number of influenza-like illness (ILI) they treat. There can be delays if the data is not delivered quickly. In January, the New York Times reported on smart thermometers that track the spread of flu more effectively. According to the manufacturer, the FDA-approved Kinsa smartphone connected oral and ear thermometers can instantly identify fever spikes in states, including cities and neighborhoods and provides 25,000 readings a day. At the time the report was published, more than 500,000 households had these thermometers.

Researchers from the University of Iowa reported on the efficacy of this smart thermometer in tracking flu activity in real time at both population and individual levels. They compared the data from the smart thermometers to ILI activity data collected by the CDC from health care providers across the country. The team found that the de-identified smart thermometer data were highly correlated with ILI activity at national and regional levels and for different age groups.  Their study, which was published in the journal Clinical Infectious Diseases, noted that the data can be used to significantly improve flu forecasting by predicting influenza levels up to two to three weeks into the future, much faster than standard approaches.

The advantages of the smart thermometer are:

  • Captures clinically relevant symptoms (temperature) likely even before a person goes to the doctor
  • Predicts influenza activity at least three weeks in advance
  • More advanced information regarding influenza activity helps alert health care professionals about the spread of influenza which improves coordination of response efforts
  • Allows health officials to start public health awareness campaigns earlier
  • Helps predict clinic and hospital staffing needs
  • Real-time information allows public health officials to allocate the right resources where and when they are necessary

Online Digital Data: Data from Twitter, Wikipedia and Google Trends are widely used to analyze flu trends. Digital data includes personal tweets about being sick, searching flu-related pages on Wikipedia, and Googling flu-related topics.

In 2017, researchers led by Northeastern’s Alessandro Vespignani developed a unique computational model using Twitter posts to project the spread of the seasonal flu in real time. The team found that the posts, along with key parameters of each season’s epidemic, such as the incubation period of the disease, the immunization rate, how many people an individual with the virus can infect, and the viral strains present, could accurately forecast the evolution of flu up to six weeks in advance.

According to the New York Times report, researchers at the University of Chicago Medical Center studied insurance claims, demographic data and 1.7 billion geolocated Twitter messages to understand the flu and people’s movements. They found the flu tends to originate in warm, southern parts of the United States and spreads northward.

Data Visualization Tool: In Missouri, the Springfield-Greene County Health Department (SGCHD) has been using LiveStories, a data visualization tool to inform residents and media about the flu.

With interactive charts, videos and benchmarks against peers, the regularly updated flu LiveStory webpage provides information on cases, a guide to symptoms and treatment resources. The new data visualization in 2018 allows the SGCHD to communicate with the public in a relevant way. The tool tracks cases by week, flu type and age of victim, and offers tips for the ailing, parents and pregnant women. SGCHD has reported a decline in calls from media and residents and credits this to the tool.

Sequencing flu variants: Experts predict that sequencing flu variants holds the most promise for trying to predict the size of an epidemic, though it will require an improved statistical framework to do so. Deep sequencing can identify rare mutations. The flu virus is continuously evolving and there are many different mutations among the millions of viruses in any single patient. Deep sequencing in infected people could help scientists monitor emerging mutation. This technique is superior to current flu surveillance which typically uses regular sequencing and identifies only the common variant.

However, digital tools are not without challenges. For instance, in the 2012-2013 and 2011-2012 seasons, Google Flu Trends (GFT) overestimated the prevalence of flu by more than 50%. From August 2011 to September 2013, GFT over-predicted the prevalence of the flu in 100 out 108 weeks. Moreover, as digital data sources are available only for the past 10-15 years, their use in longitudinal studies is limited.

Making accurate, comprehensive, and accessible data is necessary to improve predictions about disease. Medical transcription outsourcing will continue to play a key role in making complete patient information available in electronic medical records, allowing health personnel to quickly identify people who have flu symptoms and other illnesses that put them at high risk for complications.