Over the past month, the number of new COVID cases in my social circle has become impossible to ignore. I brushed off the first few—guests at a wedding I attended in early April—as outliers during the post-Omicron lull. But then came frantic texts from two former colleagues. The next week, a friend at the local café was complaining that she’d lost her sense of smell. My Instagram feed is now surfacing selfies of people in isolation, some for the second or third time.
Cases in New York City, where I live, have been creeping up since early March. Lately, they’ve risen nationally, too. On Tuesday, the national seven-day average of new COVID cases hit nearly 49,000, up from about 27,000 three weeks earlier. The uptick is likely being driven by BA.2, the new, more transmissible offshoot of Omicron that’s now dominant in the United States. BA.2 does seem to be troubling: In Western Europe and the U.K. in particular, where previous waves have tended to hit a few weeks earlier than they have in the U.S., the variant fueled a major surge in March that outpaced the Delta spike from the summer.
At least so far, the official numbers in the U.S. don’t seem to show that a similar wave has made it stateside. But those numbers aren’t exactly reliable these days. In recent months, testing practices have changed across the country, as at-home rapid tests have gone fully mainstream. These tests, however, don’t usually get recorded in official case counts. This means that our data could be missing a whole lot of infections across the country—enough to obscure a large surge. So … are we in the middle of an invisible wave? I posed the question to experts, and even they were stumped by what’s really happening in the U.S.
For a while, COVID waves were not all that difficult to detect. Even at the beginning of the pandemic, when the country was desperately short of tests, people sought out medical help that showed up in hospitalization data. Later, when Americans could easily access PCR tests at clinics, their results would automatically get reported to government agencies. But what makes this moment so confusing is that the COVID metrics that reveal the most about how the coronavirus is spreading are telling us less and less. “Why we’re seeing what we’re seeing now is one of the more challenging scientific questions to answer,” Sam Scarpino, the vice president of pathogen surveillance at the Rockefeller Foundation, told me.
Not only is our understanding of case counts limited, but all the epidemiological data we do have in the U.S. is rife with biases, because it’s collected haphazardly instead of through randomized sampling, he said. The data sets we rely on—case counts, wastewater, and hospitalizations—are “blurry pictures that we try to piece together to figure out what’s going on,” Jennifer Nuzzo, an epidemiologist at Brown, told me.
An invisible wave is possible because cases capture only the number of people who test positive for the virus, which is different from what epidemiologists really want to know: how many people are infected in the general population. That’s always produced an undercount in how many people are actually infected, but the numbers are becoming even more uncertain as government testing sites wind down and at-home testing becomes more common. Unlike during past waves, each household can request up to eight free rapid tests from the federal government, and insurance companies are required to reimburse Americans for the cost of any additional rapid tests they purchase. These changes in testing practices leave even more room for bias.
Sheer pandemic fatigue probably isn’t helping, either. People who are over this virus could be ignoring their symptoms and going about their daily lives, while people who are getting reinfected may be getting milder symptoms that they don’t recognize as COVID, Nuzzo said. “I do believe we are in a situation where there’s more of a surge happening, a larger proportion of which is hidden from the usual sort of sensors that we have to detect them and to appreciate their magnitude,” Denis Nash, an epidemiologist at the City University of New York, told me. He was the only expert I spoke with who suggested that we might be in a wave that we’re missing because of our poor testing data, though he too wavered on that point. “I wish there was a clear answer,” he said.
Instead of relying solely on case counts to gauge the size of a wave, Nash said, it’s better to take into account other metrics such as hospitalizations and wastewater data, to triangulate what’s going on. Positivity rate—the percent of tests taken that have a positive result—can be more informative than looking at the raw numbers, too. And right now, the nationwide positivity rate is telling us that an increasing number of people are getting sick: Nationwide, 6.7 percent of COVID tests are coming back positive, versus 5.3 percent last week.
Unlike traditional COVID testing, wastewater surveillance, which is a process of detecting SARS-CoV-2 in public sewage, doesn’t reveal who exactly might be infected in a particular community. But by analyzing sewer data for evidence of the coronavirus, it can provide an early signal that a surge is happening, in part because people may shed virus in their feces before they start feeling sick. Nationwide levels of COVID in wastewater have climbed steadily in the past six weeks, suggesting more of a wave than the case counts indicate, though they vary greatly by region and can’t account for the chunk of the population who doesn’t use public utilities, says Gigi Gronvall, a senior scholar at the Johns Hopkins Center for Health Security at the Bloomberg School of Public Health. Scarpino noted a rise in certain areas, including Boston and New York, but he didn’t characterize them as a wave. “Multiple data sets are showing [a] plateau in some places,” he said. “It’s that combined trend across multiple data sets that we’re looking for.”
If America is indeed not experiencing a big wave at all, that would be breaking with our recent history of following in Europe’s path. One possibility is that “the immunological landscape is different here,” Scarpino said. At the peak of Omicron’s sweep across the U.S., in January, more than 800,000 people were getting infected each day, partly a function of the fact that just 67 percent of eligible Americans are fully vaccinated. Most of those who recovered got an immunity bump from their infection, which might now be protecting them from BA.2. Even with all the data issues we have, the relatively slow rise in new cases “does raise the possibility of there being less population vulnerability” in the U.S., Nuzzo said. But, she noted, this doesn’t mean people should think we’re done with the pandemic. States in the Northeast and Midwest are seeing far more cases than the South and the West. As this wide regional variation suggests, many pockets of the country are still vulnerable.
In all likelihood, we’re seeing elements of both scenarios right now. There could be many more COVID infections than the reported numbers indicate, even while the situation in the U.S. may be unique enough to prevent the same pattern of spread as in Europe. Regardless, the course of the pandemic would be far less uncertain if we had data that truly reflected what was happening across the country. All the experts I spoke with agreed that the U.S. desperately needs active surveillance, the kind that involves deliberately testing representative samples of the population to produce unbiased results. It would tell us what percentage of the general population is actually infected, and how trends differ by age and location. Now that “we’re moving away from blunt tools like mandates, we need data to inform more targeted interventions that are aimed at reducing transmission,” Nuzzo said.
In some ways, not knowing whether we are in an invisible wave is more unsettling than knowing for certain. It leaves us with very little to go on when making personal decisions about our safety, such as deciding whether to mask or avoid indoor dining, which is especially frustrating as the government has fully shifted the onus of COVID decision making to individuals. “If I want to know what my risk is, I just look to see if my friends and family are infected,” Scarpino said. “The closer the infection is to me, the higher my risk is.” But we can’t continue flying blind forever. It’s the third year of the pandemic—why are we still unable to tell how many people are sick?
But our inability to nail down whether we’re in a wave is also an indication that we’re closer to the end of this crisis than the beginning. An encouraging sign is that COVID hospitalizations aren’t currently rising at the same rate as cases and wastewater data. Nationally, they’re still close to all-time lows. Hospitalization data, Nuzzo said, is “one of our more stable metrics at this point,” though it lags behind the real-time rise in cases because it usually takes people a few weeks to get sick enough to be hospitalized.
Even if BA.2 is silently infecting large swaths of the country, it doesn’t seem to yet be causing as much severe illness as previous waves, thanks to immunity and perhaps also antiviral drugs. If that trend holds, it may mean we are seeing a decoupling of cases and hospitalizations (and, thus, with deaths too). “This is the kind of thing we really want to see—we can absorb a big surge without a lot of people having severe infection and dying,” Nash said. Still, it’s impossible to say for certain. For that, yet again, we’d need better data.