Where are we most likely to catch COVID-19?

We know a lot more about transmission risk than we did a year ago.
a person wearing a backpack and holding textbooks against a blue background with several red coronavirus illustrations
We've learned that sanitizing surfaces is less important than keeping college students masked and far apart. Unsplash, Element5 Digital

Click here to see all of PopSci’s COVID-19 coverage.

One in four Americans had received at least one dose of a coronavirus vaccine as of March 10. After a year of living through the COVID-19 pandemic (and now that the end may finally be in sight) those who have already received their shots can start to cautiously expand their social bubbles. But for the many Americans who aren’t sure how many weeks or months it may take to get their own shots, it’s more crucial than ever to exercise care lest they risk infection during the home stretch—and help new coronavirus variants gain a foothold in the process.

With the US seeing upwards of 50,000 cases a day, what do we know now, one year into the pandemic, about where and how people get infected?

The lack of contact tracing makes it difficult to know where COVID-19 lurks

Before getting into specifics, it’s key to acknowledge that this is a challenging question to answer, especially in the US. One major reason: the country is not great at contact tracing.

Contact tracing—a practice in which trained public health workers call someone who has tested positive for COVID-19 and ask about their recent activity in order to identify other potential cases—is a well-known public health strategy. Other countries, such as South Korea, have used high-tech versions of the process to stop infections from becoming outbreaks.

In the US, though, the practice is done unevenly, if at all. Most states and local jurisdictions, struggling from years of underfunded public health departments leading up to the pandemic, have not been able to hire and train the contact tracers needed to keep tabs on every case.

Many states have attempted to supplement their limited contact tracing workforces with exposure notification apps, which are theoretically able to notify users when they’ve come into contact with someone who tested positive. Though these apps became more widespread in the US this past winter, they’re still not used widely enough to provide useful information. New Jersey, one state that provides data on its app use, reports that about 574,000 state residents have downloaded the app as of March 6—out of a population of 8.9 million.

It should come as no surprise, then, that there is no national repository linking COVID-19 cases to specific events or settings. Most states don’t publish such data, either, and those that do attribute many cases to unknown origins. All of that said, we can learn a lot from what data we do have.

[Related: Can we ever hug again?]

Outside is always less risky when it comes to catching COVID-19

COVID-19 experts generally agree that outdoor activities are safer than indoor ones. In an extensive FAQ document on coronavirus spread through air, a group of aerosol scientists and engineers write, “All data show that outdoors is far safer than indoors, for the same activity and distance.” Koen Swinkels, an independent data researcher who runs the Superspreading Events Database, has analyzed more than 2,000 of these situations—instances in which a few people with COVID-19 quickly infect a significantly larger number. None of them occurred outdoors.

That’s not to say that popping up a tent outside of your venue makes a packed gathering safe. While the evidence suggests that events that occur completely outside are pretty safe, Swinkels says, he has seen events featuring both an outdoor and indoor component show up in his database.

Also worth noting: Less risky doesn’t mean not at all risky. It’s still important to wear a mask in outdoor spaces when you’re within several feet of others, and a smaller gathering—one where there’s enough room for people from separate households to stay far apart from each other—is always safer than a larger one. For more information on what kind of structures count as “outdoors,” for epidemiological purposes, check out this explainer on pop-up dining options.

Surfaces are not a common source of COVID-19 transmission

You may have seen this finding before, but it bears repeating: You probably will not catch the novel coronavirus from touching your groceries or your mail.

That same aerosol FAQ document referenced above notes that the virus is much more likely to be transmitted by the air, whether that’s from being in close proximity to someone’s cough or stepping into a poorly-ventilated room with lingering virus particles (like outdoor dining igloos). Extensive cleaning procedures touted by some businesses may look impressive, but actually have minimal impact.

The New York City subways, for example, are still closing every night to deep-clean each train, in a move that epidemiologists say is an example of “hygiene theater.” You should feel safe taking public transportation because you can see that your fellow riders are wearing masks and you can feel regular airflow, not because millions of dollars went into soaping down the plastic seats.

Washing your hands, especially when you’re out in public, is still a healthy practice that is often recommended by COVID-19 scientists. But masks, physical distancing, and ventilation are now known to be much more important in protecting yourself and others from coronavirus.

[Related: COVID-19 cases are moving—slowly—in the right direction]

Facilities like nursing homes tend to be COVID-19 hotspots

The majority of the superspreading events in Swinkels’ database occurred in communal living sites and other congregate locations, where many people gather for an extended period of time. These include nursing homes, prisons, rehabilitation centers, and food processing plants.

Sites like these are perfect for coronavirus spread because they involve large numbers of people in close proximity, often in environments where physical distancing is not possible. Nursing homes have been particularly dangerous, as elderly residents face higher risks of severe disease or death if they are infected. Luckily, after they were prioritized early in the vaccine rollout, millions of these residents and staff have received doses; infection rates are now going down.

Prisons continue to be major hotspots, though. The Marshall Project reported in December that one in every five state and federal US prisoners had tested positive for COVID-19. In Colorado, one state which reports highly detailed data on COVID-19 outbreaks, 6,900 out of a total 28,000 cases in active outbreaks are linked to state prisons as of March 10. Another 3,900 current outbreak cases are linked to jails.

The next-highest outbreak setting in Colorado is colleges and universities, representing 6,700 cases. College dorms are another type of communal living site, and careless students can make matters worse. Town police had to break up a mask-less party at the University of Colorado in late February.

Even if you’ve avoided mass gatherings, indoor dining, and keggers, their very existence puts you at higher risk of catching COVID-19 at relatively safe locations like grocery stores and parks. Parties and other superspreading events aren’t just dangerous for the people directly involved—they send COVID-19 out into a broader community. Students who attend a maskless party or guards who work in a prison bring cases out to their local shops, restaurants, healthcare centers, and other establishments. One December 2020 study from the Prison Policy Initiative found, for example, that COVID-19 caseloads grew more quickly during the summer in rural counties with larger incarcerated populations.

“The same types of settings that occurred most frequently [as superspreading events] at the start of the pandemic still do now,” Swinkels says. He questions why some businesses that employ large numbers of people have not adopted better ventilation or allowed for more distancing.

Swinkels also notes, however, that communal settings may be overrepresented in a database like his because they may have more capacity to test people and actually link those cases to each other. His team’s website is careful to warn researchers that the page is not a comprehensive repository of superspreading events; instead, it represents as many events as they’re able to compile from papers and media reports.

Indoor dining—and similar settings—pose COVID-19 risks

In a new study published by the CDC last week, researchers provided evidence for a statement that may seem self-evident: Eating inside a restaurant is a great way to catch and spread COVID-19.

When a state allows indoor dining, the study found, COVID-19 cases and deaths go up in counties across that state several weeks later. While it is difficult to link specific cases to specific restaurants due to gaps in contact tracing, this county-by-county study shows how the practice impacts an entire community—even when only a small fraction of the population actually participates in that indoor dining.

[Related: When it comes to COVID-19 risk, what counts as ‘outdoor’ dining?]

Though this study focused on indoor dining and mask mandates, it has implications for other settings, says lead author Gery P. Guy Jr., a CDC scientist. He urges caution in any setting where “physical distancing of six feet or more is difficult to maintain and the consistent use of masks is not possible.” This can apply to clubs, travel hubs like airports, retail stores, and gyms, depending on local mask mandates and the conscientiousness of your fellow patrons.

Gyms, in particular, may be superspreading sites if masks and physical distancing aren’t used, as exercise leads to higher respiratory exertion. A CDC study published in February found that, among 81 participants in an indoor high-intensity fitness class at a Chicago gym, 55 caught the coronavirus. Two dozen of those people had attended other classes after noticing they had COVID-19 symptoms, and masks weren’t worn consistently.

The risk of catching COVID-19 at work depends on your job

Essential workers are called essential for a reason: They are asked to risk catching the coronavirus in order to perform duties that keep society running. They are the waiters in those restaurants open for indoor dining, or the workers clocking in at those food processing plants.

While the US does not provide data linking COVID-19 risk to specific occupations, the UK does. In an analysis of COVID-19 deaths from March to December 2020, the country’s Office for National Statistics found that the highest rate of death among men was in “elementary occupations”—jobs that require some physical labor, such as security, construction, and farming. Among women, the highest death rates were in machine operators and those in caring, leisure, and service work—jobs that include taking care of sick patients, the elderly, supervising children, and providing travel services.

These occupations tend to be riskier because they involve more face-to-face interaction with other people. In the US, people of color are more likely to be essential workers—and are also more likely to become infected. The COVID Tracking Project’s COVID Racial Data Tracker reports that 7,600 out of every 100,000 Hispanic/Latino Americans and 5,800 out of every 100,000 Black Americans have tested positive, compared to 4,500 out of every 100,000 white Americans. The numbers are even higher for indigenous communities and Pacific Islanders. [Editor’s Note: The author volunteered at the COVID Tracking Project.]

Don’t rule out COVID-19 risk in small gatherings

While gathering in your home with trusted family and friends may seem like a safer option, it can still be risky if you don’t take precautions.

Massachusetts is another state which publishes data on case origins, sorting cases into “clusters” which are defined as two or more infections that share a common source of exposure. Out of about 20,000 cases connected to active clusters in the month of February, 19,000 were tied to individual households—people who live at the same address infecting each other.

These data may be biased, though, because it’s easier to trace cases when you catch the coronavirus from someone you know (or live with). The 20,000 cases in Massachusetts’ active cluster report account for only one-third of cases the state reported that month, meaning a lot of other transmission events are going unexplained.

Still, this finding suggests that those who are not yet vaccinated need to continue proceeding with caution when holding small gatherings. The CDC now recommends that vaccinated individuals can go unmasked with other vaccinated people or just one un-vaccinated household; it’s not time for big parties yet.

Avoiding COVID-19 risk means layering up precautions

You may have seen a diagram that disease experts call the “Swiss cheese model.” It functions as a metaphor for pandemic risk mitigation: Each slice of cheese is one choice you can make to keep yourself and those around you safer. Even if there’s a hole in one piece of cheese—say, your mask slips off your nose while you’re at the grocery store—other layers will ensure that you still have some protection.

Four important layers of protection, according to the CDC’s Guy, are masks, ventilation, physical distancing, and the surveillance measures like contact tracing that can tell you to stay home if you feel sick or are exposed to someone who tests positive.

Swinkels similarly emphasizes the value of ventilation and physical distancing. He says the worst superspreading events are crowded, indoors, feature poor ventilation, and keep people together for long periods of time—especially if they’re vocalizing, and even more so if they’re singing or shouting.

These same layers of mitigation measures can help protect against the viral variants now circulating widely in the US. “The emergence of variants really just highlights the increased importance of doubling down on these measures to help prevent further spread,” Guy says.