Over the past few years, a California-based tech startup has repeatedly made headlines for beating public health agencies at their own game.
Kinsa, which makes internet-connected thermometers, has routinely detected the spread of seasonal flu weeks before the Centers for Disease Control and Prevention. And when COVID hit last year, the company saw unusual spikes in fevers about 18 days before states recorded peaks in deaths.
“The difference is not that we’re smarter,” said Kinsa founder and CEO Inder Singh. “We’ve got better data.”
Many disease-tracking efforts, including the CDC’s flu surveillance system, rely on data — patient symptoms, test results, inpatient admissions and deaths, for instance — reported by hospitals, laboratories and other health care facilities. But Kinsa’s devices provide an illness signal as soon as someone feels sick enough to use a thermometer. “In simple terms, we talk to mildly symptomatic patients,” Singh said. “The health care system misses them entirely.”
Now, the company is putting its pandemic prognostication skills to a new test in a partnership with the New York City Department of Health. Over the coming months, Kinsa will distribute as many as 100,000 free smart thermometers through the city’s elementary schools and will make the resulting data available to local health officials. The goal is to create a citywide early-warning and early-response system for outbreaks of COVID, the flu and other infectious diseases.
“One of the critical lessons that we have learned during the COVID epidemic is how important it is to have as accurate information as we can possibly get, in real time, about how diseases spread through communities,” said Dr. Jay Varma, senior adviser for public health to Mayor Bill de Blasio.
The Kinsa partnership, he added, is “going to help us strengthen our ability to understand new and emerging diseases that may pop up in the school community.”
This is not Kinsa’s first foray into schools. Since 2015, it has distributed thermometers through more than 4,000 individual schools across the United States as part of its FLUency program. But the New York City initiative will be its first citywide rollout.
The first phase began last month, when the company began distributing 5,000 free thermometers to teachers, staff members and families at 50 elementary schools in city neighborhoods that had been particularly hard-hit by COVID. In the fall, Kinsa hopes to open the program to all of the city’s elementary schools.
The program is entirely voluntary, officials at both Kinsa and City Hall stress. Schools that opt into the program will send Kinsa’s brochures home to students’ families; if parents want to participate, they can download Kinsa’s app and order a free thermometer.
“There’s nothing about this program that is mandated,” Varma said. “Schools are not required to participate. Families are not required to participate, and of course they can kind of discontinue it at any time.”
The program, partly funded by Lysol, will be free to schools and families.
Readings from the thermometers will be sent to the accompanying app, which also asks users to log any other symptoms they may be experiencing. Depending on what they report, the app may recommend that a child stay home from school, suggest a visit to the doctor or direct users to a nearby COVID testing site. School administrators and families can view information about grade-level trends at their own schools — that there are four ailing fourth-graders, for instance.
City health officials will also have access to this aggregated, anonymized data, which they hope will help them identify unusual illness clusters earlier than is currently possible. “It’s measuring something that we’ve never really been able to measure before,” Varma said. “This is information about people’s biological measurements, being taken by somebody in their home before they’ve actually, in many situations, sought care.”
If officials see an illness beginning to spread through a school, they could marshal a targeted response, Varma said — perhaps by doing some deep cleaning, temporarily closing a school building, or helping students and staff members access care.
And while the rollout focuses on schools, Kinsa says the system could benefit all New Yorkers. “The reason that we’re targeting elementary schools is because we know parents of elementary-age children are power users of thermometers,” said Nita Nehru, vice president of communications at Kinsa. “In a normal illness season, illness spreads rapidly within schools, and that’s where you want to start gathering data earlier, if you want to be able to stop the spread of illness more broadly.”
The data has limitations. Thermometers, of course, will not catch ailments that are not accompanied by fevers, and many cases of COVID, especially in children, are asymptomatic. Moreover, schools and families that opt into the program may not be representative of the city’s population at large.
And then, of course, there are the inevitable privacy concerns. Kinsa emphasizes that all data provided to the city will be aggregated and anonymized. “None of the individual data is going to anyone other than to that individual,” Singh said. “They own the data, and we’re really adamant about this.”
While digital-privacy experts say that these are important safeguards, they also note that information about children and health is particularly sensitive. “It’s really important to balance the public health benefits and needs with the social or societal risks,” said Rachele Hendricks-Sturrup, health policy counsel at the Future of Privacy Forum, a think tank focused on data privacy.
For instance, even de-identified data can sometimes be re-identified. “Even if it becomes ‘A fourth-grader at this school in this neighborhood,’ that could narrow it down,” said Hayley Tsukayama, a legislative activist at the Electronic Frontier Foundation, a digital-privacy group. “It doesn’t take a lot of data points to re-identify something.”
The data, aggregated by ZIP code, will also be incorporated into illness signals that Kinsa makes available in its public HealthWeather map. The company sometimes shares this ZIP-code-level information with pharmacies, vaccine distributors and other companies. Clorox, for instance, has used Kinsa’s data to determine where to target its ads. (Lysol will have no special access to the data, Kinsa says.)
Both Kinsa and the city need to be transparent with families about how the data will be used, stored and shared and how long it will be retained, experts said. City officials are “essentially putting their stamp of approval on this,” said Amelia Vance, director of youth and education privacy at the Future of Privacy Forum. “They need to make sure that they are living up to the trust that parents will have that this program has been fully vetted and is safe for their kids and their families.”
Over the coming months, city officials will keep close tabs on how well the program is working, Varma said. How do families feel about the program? Is there enough uptake to produce useful data? Can they actually catch outbreaks earlier and slow the spread of disease?
“Our goal is to try to see whether or not, in the real world, whether it really does have that impact that we hope it does,” Varma said. “It’s also possible the system may not detect anything abnormal or unusual, but that it still proves to be successful because it provides people with information that they find useful and builds their confidence in having their kids at school.”This article originally appeared in The New York Times.