Kinsa’s smart thermometer is proving social distancing works. Now it needs more data.

Kinsa Insights has long worked to improve real-time illness tracking with its flagship smart thermometer. Kinsa’s smart thermometers report users’ temperatures back to the company’s databases, and that data is then compiled into graphs and maps of fevers across the United States. In the past, this data has been used to accurately predict the spread of influenza across the country.

Now Kinsa has set its sight on the COVID-19 pandemic. By comparing current spikes in fever reporting to those typical of an area at a given time, Kinsa has created a map of “atypical” illness levels across the country. The resulting Health Weather Map presents a picture of abnormal numbers of fevers, which may be linked to localized outbreaks of COVID-19.

Kinsa notes — prominently, at the top of the map’s page — that this data does not represent COVID-19 activity. Rather, Kinsa hopes this data can be viewed alongside other indicators to show where and how COVID-19 is spreading.

Kinsa is right: live-tracking the spread of a novel virus like COVID-19 could prove invaluable in ending the pandemic. But if the prospect of doing so sounds like a futuristic dream, that’s because it is. Between a shortage of the company’s smart thermometers and scattered, incomplete data, Kinsa’s Health Weather map is more a prototype for the future than a solution for the present.

A vision for collective health — Kinsa’s approach to public health is about as future-forward and democratic as you can get. All of the company’s public data is crowdsourced from its customers, anonymized, and then aggregated into graphs and maps. What makes the data invaluable is immediacy — though by no means “live” data, it can be parsed much more quickly than data reported by medical professionals.

Take, for example, the shortage of COVID-19 testing faced during the first few weeks of the pandemic in the U.S. This shortage led to a dearth of information: without up-to-date information on where the coronavirus was spreading most quickly, politicians and health officials were left in the dark about which areas needed the harshest measures and most resources to fight the pandemic. Meanwhile, the rush to hospitals for testing only helped the virus spread.

Imagine instead that this data could be aggregated from home. In a perfect world, everyone would be taking their temperature each day during the outbreak to monitor any spikes. This data could then be aggregated and used to enact swift policy changes, preemptively redirecting resources to where they're needed, or where they will be in days to come.

A dream, for now — Kinsa’s vision is radical. It has the power to change the way we monitor infectious diseases in the future. But its uses for the present pandemic are severely limited, through no fault of its own. It's devices simply aren't sufficiently prevalent.

That lack of ubiquity means Kinsa’s most prominent limitation is its incomplete data sets. The company can only collect data from its customers — those who have deemed it essential to own a $40 thermometer. Then there is the limitation of self-reported data. Kinsa says it has sold or given away “millions” of its thermometers. But how many of those thermometers are being used on a consistent basis? Those customers most likely to use the thermometer are those already feeling ill. That means Kinsa’s data is potentially missing many, many data points from healthy customers.

Right now, the aggregated data is being plotted on two maps: one with levels of observed illness (ranging from “low” to “severe”) and another that compares these observed levels to those typically seen in a given area. Because Kinsa’s data is limited, the maps created from the data are riddled with holes and monotony. The map of atypical illness levels contains more locations with “insufficient data” than those with analyzed data. Similarly, the map of observed illness levels depicts most of the United States as “mild.” Moreover, the data can go days without being updated.

Overall, these two maps present a picture of relative health in the United States as a whole. This stands in stark contrast to other maps showing the pervasive way in which the virus has spread across the country. In fact, Kinsa’s overall data puts the U.S. below “expected” levels of illness. To put it bluntly, Kinsa’s maps simply are not an accurate portrayal of the current state of public health in the United States.

This lack of data will not be rectified soon enough to fight the spread of COVID-19. The spread of the virus has progressed too far already, and there’s simply not enough time to give the general population access to smart thermometers. Kinsa’s smart thermometers are actually all sold out at the moment, and the company is estimating a four-week wait before new orders will ship.

Let’s stick with realism — Kinsa knows it isn’t an immediate solution to the COVID-19 pandemic. The company has scattered throughout its site multiple flags: this data does not represent the spread of COVID-19, it represents all data that points to increased illness.

Despite this, Kinsa is hopeful that its data will be useful in some way to those fighting the current pandemic. Others have echoed this sentiment. The reality of the situation is that the COVID-19 pandemic will not be solved by smart thermometers. There just aren’t enough of them, and the data gleaned from the ones there are can’t be used to direct resources right now. Too much of the country is left out of the equation.

Kinsa’s vision of crowdsourced public health information will prove very useful for future planning and prevention techniques. Smart thermometers will no doubt play a role in how we learn from the COVID-19 pandemic. For now, though, we’d do best to focus our efforts on building tech that can be used to reduce the harm caused by the novel coronavirus, such as ventilators and fitted face masks. The smart thermometers are going to have to wait.