Flu data may help COVID-19 battle
Unusual patterns likely caused by coronavirus
Mike McMahon vacationed with his family at Disney World in late January. Back in Boston a few days later, he started to feel lousy.
He “couldn’t go from laying on the bed to sitting up without being out of breath,” he said.
A flu test came back negative, but an urgent care doctor prescribed Tamiflu anyway. The next day, a different doctor prescribed antibiotics and an inhaler for pneumonia. A second flu test came back negative. It was unclear what had caused his lungs to fill with fluid.
“None of it seemed to help,” McMahon said. “I never felt better.”
At that point, the USA had confirmed only about a dozen coronavirus cases. It was a month more before Massachusetts reported its first case.
McMahon is among thousands of Americans who reported flu-like symptoms but might’ve instead had COVID-19 – particularly in the early weeks of the pandemic when testing wasn’t widely available.
A USA TODAY analysis of historical flu and pneumonia surveillance data shows unusual patterns this year that experts suspect are the result of the coronavirus pandemic.
The data shows more people than usual reported flu-like symptoms while positive flu tests declined. Pneumonia deaths continued to rise. Typically, all three metrics rise and fall together.
That gap, health experts said, is probably COVID-19.
Researchers hope to use that data to detect and respond to the next big coronavirus outbreak as early as possible.
States such as Wisconsin are using it as a benchmark – along with coronavirus testing – for decisions about how quickly to end lockdowns.
Because so many people with COVID-19 develop symptoms similar to the flu, experts can monitor the weekly
surveillance reports for spikes like those seen in many places in February and March and quickly respond.
State officials can tell which hospitals are seeing more coughs and fevers, giving them an early signal that the coronavirus might have arrived in a particular area. It could be most useful for tracking the disease through communities without widespread testing.
“These systems, which don’t require a laboratory diagnosis of COVID-19, are going to be really useful,” said Ben Lopman, an infectious disease epidemiologist at Emory University.
Flu or not
Nationwide, as early as January, patients such as McMahon flummoxed their doctors, arriving in clinics and hospitals with flu-like symptoms but testing negative for influenza. Many did not qualify for a coronavirus test, even if one was available, but they were still counted by state officials in syndromic surveillance.
In Massachusetts, the flu season seemed to peak in the first week of February. Roughly 7% of people visiting hospitals reported flu-like symptoms, including McMahon.
By the first week of March, less than 4% of patients reported flu symptoms.
“The prevalence of flu in the community dropped off dramatically,” said Saul Weingart, chief medical officer at Tufts Medical Center in Boston. Yet, Weingart said, patients were still arriving with serious respiratory symptoms.
State data shows another spike in the third week of March as the percentage of patients reporting a cough and fever grew to 7%. Only 570 tests came back positive for the flu that week, compared with about 6,000 the first week of February and 2,000 the first week of March.
Positive flu test rates dropped low enough that many doctors stopped testing for the flu altogether, Weingart said.
National flu data mirrors that of Massachusetts.
In the first week of February, influenza-like symptoms accounted for 7.1% of hospital visits, then declined and later spiked again in late March to 7.4% even as fewer flu tests came back positive.
In March, nearly 300,000 Americans reported flu-like symptoms, which is 55% higher than in 2019 when the number of hospitals reporting figures was similar. Only 12% of flu tests were positive compared with 24% the year before – about 64,000 more negative results.
Although some people who tested negative for the flu later tested positive for COVID-19, doctors said many did not qualify for a coronavirus test. Many states rationed testing supplies through March and April.
“There were definitely patients that I saw in January and February who I tested for flu, and the flu test was negative … that very well may have had COVID-19,” said Arvind Venkat, an emergency physician at Allegheny General Hospital in Pittsburgh.
Pneumonia deaths grow
At the same time positive flu tests declined, more people were dying from pneumonia, which is typically caused by the flu or related bacterial infections.
“Given that pneumonia is one of the complications from having COVID-19, you’d anticipate undetected COVID cases that were listed as pneumonia,” said Timothy Lant from the BioDesign Institute at Arizona State University.
“It certainly happened in China. It certainly happened in Italy,” Lant said. “And there’s evidence it’s happening here.”
Provisional death certificate shows the trend most clearly.
In a typical year, about 11,300 Americans die from pneumonia in the last week of March and first two weeks of April. This year, nearly twice that many people died from it during the same period. In at least 31 states, pneumonia deaths were higher than average in March and early April.
More than 87,400 death certificates listed pneumonia as the cause of death but did not specify which pathogen triggered the condition, according to state data reported to the National Center for Health Statistics since February. data
“Without a doubt,” some of the unclassified pneumonia deaths from this year include uncounted COVID-19 cases and could help explain the unusually high number of fatalities, said Matthew Boulton from the University of Michigan School of Public Health.
States with the most elevated figures of pneumonia deaths in March are known COVID-19 hot spots. In New York, pneumonia deaths were 76% higher than usual. They were 65% higher in New Jersey, 48% higher in Washington and 31% in the District of Columbia.
Montana’s count of flu and pneumonia deaths for March – 36 – is the highest in at least five years and 64% higher than average, but the state reported five COVID-19 deaths in March.
“I don’t think we truly know what the penetration rate was across the U.S.,” said Ted Ross, director of the Center for Vaccines and Immunology at the University of Georgia. “Different states had different capacities for testing. Ones with larger metro areas were able to concentrate testing to that location. Rural areas often were not touched by testing.”
The signal will be clearest this summer when the flu is not being spread.
Although death data is delayed compared with symptom surveillance, experts said it still is important to study and could influence public health responses.
By reviewing deaths missed in initial COVID-19 tallies, officials can better understand how the coronavirus spread and identify gaps in testing, community outreach and health care access. Scientists can target studies that reveal more about how the virus attacks the body, who is most at risk for severe symptoms and how to predict its spread more accurately.
“Looking at that number (of unusually high deaths), some people may say that that’s not a lot,” said Angela Clendenin, whose research at the Texas A&M University School of Public Health includes pandemic response and communication. “But the value of looking at that number and placing an importance on that number is letting their deaths mean something. What can we learn from it? What can we figure out from it that will help us save lives in the future?”