What’s going around

Clammy with a chance of sniffles: New health apps track sickness hotspots

Health surveillance apps are being touted as Doppler maps for disease. Do they live up to the hype?

Topics

At the touch of an icon, a smart phone can tell you if light showers or heavy wind gusts are headed your way. In the not-too-distant future, mobile apps designed to track illness could be used to accurately forecast waves of congestion, surges of fever and floods of fatigue.

Several health apps emerging on the mobile market tout themselves as Doppler maps for disease, drawing on data aggregated from search engines and social media platforms like Facebook and Twitter to track sickness hot spots in communities across the nation.

Although the technology is still developing and current data can be somewhat unreliable, such apps could soon play a key role in tracking and even preventing the spread of contagious diseases.

“Social media has really been an underutilized platform for tracking illness,” says Jay Bhatt, DO, chief health officer at the Illinois Health and Hospital Association. “It could potentially identify an illness before it really takes hold in an area.”

Dr. Bhatt and his team enhanced and managed Foodborne Chicago, an award-winning web application that uses Twitter to identify individuals who believe they may have food poisoning. The app flags public tweets for possible follow up from the city’s health department, offering the capability for health officials to track foodborne outbreaks in real time.

“This kind of tracking is a tool that offers tremendous potential for improving public health,” Dr. Bhatt says.

The reliability factor

Although the dependability of mobile apps like Sickweather, FluView and HealthMap continue to improve, many illness trackers rely heavily on self-reported, unsubstantiated data. Contributing further toward unreliability, too few users are engaging or contributing on many of the mobile platforms to make the data conclusive.

Aaron George, DO, walks a patient through an app.

Farhad Modarai, DO, a family medicine resident at Duke University Medical Center in Durham, North Carolina, stresses that the number of people using illness trackers plays a significant role when it comes to reliability.

“In an ideal world, you’d want close to one-third, if not one-half of a city’s population reporting their symptoms,” he says.

Cole Zanetti, DO, points out that older surveillance trackers, such as the now-defunct Google Flu Trends tool, relied on search engine activity to predict illness trends. However, the tool’s algorithms weren’t able to discriminate between users who were searching flu symptoms or treatment options and those searching for other reasons.

“The validity of the data used to perform predictions was inaccurate,” he says. “Tools like Google Flu Trends tended to overestimate.”

New flu trackers, like ARGO, do a better job of interpreting user search behavior, resulting in more accurate estimations. Developed by Harvard statisticians, ARGO relies on Google search trends, but also incorporates CDC historical data and seasonality factors. Work is under way to make the tool more widely available.

Putting it into practice

As popularity of sickness trackers continues to increase, patients could potentially misinterpret or overreact to app data, leading to unwarranted health concerns.

“If a patient believes cases of the flu are popping up all over their neighborhood, they might be convinced they need a Tamiflu prescription when they’re really just battling a cold,” explains Aaron George, DO, a family physician in Chambersburg, Pennsylvania.

Ultimately, any app that leads to open dialogue between physician and patient is a good thing, Dr. George stresses. “I love when my patients use apps that lead to thoughtful questions about health concerns,” he says. “It gives us an opportunity to have a discussion and partner together on a solution.”

Leave a comment Please see our comment policy