As governments around the world pushed their citizens away from populated areas to slow the spread of Covid-19, they may not have realized that they were also fighting other infectious diseases, such as the seasonal flu.
But they were, according to data from Kinsa Health, a company that sold or donated more than a million smart thermometers in the United States. Kinsa collects anonymized thermometer readings (via its app, which users connect to the device) from its active user base to estimate the proportion of sick people in different geographic areas. By comparing current thermometer readings to historical trends, the researchers used Kinsa data to predict flu epidemics weeks before the Centers for Disease Control and Prevention surveillance program, which uses hospital records.
Recent data clearly shows the spread of Covid-19. On March 19, the proportion of Americans whose temperature indicated that they had flu-like symptoms was around 4.9%, when it should normally be around 4.0%. This was probably the result of the spread of Covid-19, according to the Kinsa researchers.
But by March 23, it had dropped to 3.3%, down from 3.7% (the share of fevers drops rapidly at this time of the year due to the end of winter).
The drop – 0.9% above typical flu rates 0.4% below – in just four days, it’s the biggest Kinsa has ever seen in such a short time, according to Kinsa’s CEO Inder Singh. “There is no known precedent for this type of widespread social distancing in recent times,” said Singh. “We have nothing to compare to that, but this extreme drop is exactly what we hope and expect with the measures currently in place. “
Due to the emergence of the coronavirus, Kinsa receives two to three times more thermometer readings per day than in previous influenza seasons. Historically, Kinsa’s methods have been able to explain a large increase in the number of people taking tests, said Benjamin Dalziel, a professor at Oregon State University who studies infectious diseases. For his research, Dalziel made sure that a spike in tests did not lead to inaccurate predictions, and he therefore believes that the current decline in the disease is the result of social distancing, not a statistical anomaly. . (Dalziel has had research funded by Kinsa in the past).
Kinsa says its data showed that New York and southern Florida were possible Covid-19 hotspots before tests even proved it. The data below shows that King County, which includes Seattle, was well above average in early March when the epidemic was in full force there. New York County (Manhattan Borough), Miami-Dade County and San Francisco County were also above average at the start of the month.
In mid-March, temperatures in King County and San Francisco began to drop, while New York and Miami-Dade continued to rise. Since March 19, readings in all of these counties have dropped due to social isolation.
Unlike most places in the United States, El Paso County, Colorado has not experienced a drop in user temperatures in the past week. The data does not show us exactly why – this may be due to the fact that county residents are not yet as far from society as in other parts of the United States, or because the coronavirus epidemic is a overwhelming gain from a decrease in social contact. As of March 25, the county has 106 confirmed cases of Covid-19.
Even the places that have benefited the most from social remoteness may not see their temperatures stay low if the epidemics persist, said Dalziel. While social distance can greatly slow the spread of diseases like Covid-19, it has yet to stop it completely.