Editor’s Note: Every Wednesday, James Hamblin takes questions from readers about health-related curiosities, concerns, and obsessions. Have one? Email him at paging.dr.hamblin@theatlantic.com.
Dear Dr. Hamblin,
I’m an attorney and I’ve been working from home. Yesterday I had to visit several courthouses to pick up files. At the security checks at the entrance, they had some kind of infrared camera to check people’s temperature. At one courthouse, I came in through the employee garage and someone stopped me and put a thermometer to my forehead. I live in Miami, my car’s AC is broken, and I had been stuck in traffic and was soaked in sweat. Yet my temperature apparently was 97 degrees, and I was waved inside.
Are these temperature checks just “security theater”? How effective are they?
Megan
Miami, Florida
Yes, what you’re describing is theater. As logical as mass temperature checks may seem, the practice hasn’t clearly been shown to help contain the coronavirus. The practice is sort of like spraying down the sides of buildings, showering football players in hand sanitizer, or deep cleaning an office carpet. These things might make us feel safer, but they may not keep us safe if they actually cause us to let our guard down.
Fever screens seem part of America’s ongoing failure to learn an abiding lesson of modern medicine: Screening tests often backfire. They have value only in specific contexts. Doctors have repeatedly seen this play out with lifesaving tests like mammograms, CT scans, and colonoscopies, which can lead to unnecessary biopsies and surgeries when overused.
Body temperature is a vital sign that is monitored continually in hospitals, where it can be an invaluable window into how a patient is faring. But screening entire populations is a different matter, and temperature screening for the coronavirus is now widely used. Employers around the United States are attempting to order daily temperature checks to determine who is allowed to enter an office and do their job. As you noticed, access to courts may be based on this test. Fever screens have even been implicated in many plans to open schools: New York City’s Department of Education announced last week that kids would be randomly screened before being allowed to enter school buildings, despite the fact that kids with the virus rarely develop fevers.
Temperature checks were first deployed in January amid China’s “throw everything at the wall and show the world we’ve got this locked down” attempt to stop the virus. News reports showed people being ordered out of their car at tollbooths to have infrared thermometer guns placed against their forehead. For a period in February, the Chinese government required some 11 million people to report temperatures to the government daily. By late February, though, researchers had begun reporting that the tactic wasn’t an effective screening tool for this new disease. Unlike Ebola and some other infectious diseases, this coronavirus is contagious before symptoms appear. Many infected people never develop any symptoms at all. While a fever is one of the most common symptoms for people who do get sick, it’s still far from universal among those infected. In June, a study at an Australian hospital found that a minority of patients admitted with the virus had a fever. The researchers concluded that “using fever as a screening tool for COVID-19 may provide a false sense of security.”
But what’s the harm? In this moment, that’s a question we need to take seriously—with regard to tests, hygiene practices, or anything else we might be doing to try to stay safe. Anything with the potential to benefit our health also has the potential to undermine it. If people are reassured by a fever check and compromise on the basics—wearing a mask, distancing, hand-washing—they put themselves and others at risk. This test has no ability to reassure people, but a real ability to mislead.
A useful test is both sensitive and specific. If you haven’t encountered these terms yet, they can seem esoteric, but we should all know them. A test is considered sensitive if it catches most (and ideally all) cases. A test is specific if it doesn’t misidentify too many cases—that is, if it doesn’t catch too many more than all.
Temperature checks for coronavirus infections don’t seem to be sensitive or specific. An elevated body temperature can be caused by many factors other than COVID-19—or illness at all. Some people will see an elevation after exercise or during anxiety. When you’re sweating, your body is working to expel heat through the skin. I’m surprised you registered at 97 degrees. But I still couldn’t comfortably say you aren’t spreading the virus.
Your temperature also may not have been 97 degrees. Infrared scanners are imperfect compared with measuring temperature under the tongue or in the rectum. When they’re pointed at your forehead, they’re pretty reliable. But when scanning a crowd, as some public-transit agencies and school districts have started to do, the accuracy fades. Such technology is well suited to detect the presence of humans, as in disasters or military scenarios, but subtle variations and mild fevers are likely to be lost in the crowd. Discerning those is especially difficult because, as you note, people’s normal body temperatures vary on their own. What may constitute an early elevation in one person could be a normal temperature for another.
The irony that all of this is happening in a courthouse is probably not lost on you. The state collecting medical information about people raises all sorts of legal issues, only heightened now that the outcome of an inaccurate test could deny access to basic services like courts or schools. Because scanning people for a fever is typically considered a medical test, the act also typically requires consent. And consent or no, employers aren’t normally allowed to demand medical tests of employees.
When the pandemic hit the U.S, the Equal Employment Opportunity Commission developed emergency precautions that specifically allow employers to monitor employees’ body temperature. But not everyone agrees that this is sound legal policy. As the American Civil Liberties Union argued in response to fever screening in May, privacy is always valued relative to the value of the intervention. Some privacy may need to be sacrificed during a crisis, but we must be deliberate about how and where those compromises are made: for tests that will deliver actionable information proven to help protect the population.
Temperature checks may be useful in specific circumstances, as opposed to mass fishing expeditions. But people who are found to have a fever are likely being told what they already know. When fevers do come on, people typically already feel ill. They’re likely to also have a cough, shortness of breath, or other symptoms. They should have had access to a proper test that would tell them if they had the virus. They should have been able to seek medical attention. They should have been able to isolate at home.
The real spirit of mass temperature checks, as they’re currently being used, is punitive. If, amid the many false positives and false negatives, fever screening does successfully catch a few sick people who went out anyway, then this may be a symptom of a societal ill. Feverish people are not likely venturing out because they have a death wish or want to infect others; they’re likely doing it because they can’t afford to stay home, or they need to pick up their kids, or they don’t have sick leave. The solution to this isn’t to pour more resources into infrared cameras in an attempt to detect and detain these people. It’s to put those resources into taking care of them in the first place.
“Paging Dr. Hamblin” is for informational purposes only, does not constitute medical advice, and is not a substitute for professional medical advice, diagnosis, or treatment. Always seek the advice of your physician or other qualified health provider with any questions you may have regarding a medical condition. By submitting a letter, you are agreeing to let The Atlantic use it—in part or in full—and we may edit it for length and/or clarity.
