Coronavirus data have always been incomplete—but the situation in America is particularly murky now.





The numbers are remarkable. More than 100 million people in the United States have likely been infected by SARS-CoV-2 and 167 million people are fully vaccinated. Yet despite this huge population of people with at least some level of immunity, the Delta variant has sent case and hospitalization numbers soaring. Florida is on its way to having twice as many people hospitalized now than during any previous wave, when essentially no one was vaccinated.

One way to think about it, as the epidemiologist Ellie Murray has laid out, is that if Delta is as transmissible as the CDC thinks, we need a much higher percentage of our population vaccinated for immunizations and natural infection alone to cause the virus to peter out. Even when the huge majority of people in a given place have gotten the coronavirus or a shot, there might still be outbreaks, as the Brown University public-health expert Ashish Jha fears will happen in South Dakota after the Sturgis Motorcycle Rally.

These realities have already smashed the more optimistic projections of late spring, including my own. Having stared at these numbers for months and months with the COVID Tracking Project at The Atlantic, I never thought that we’d see hospitalization numbers higher than they were during the winter peak in any state. But here we are.

It’s time for a data-driven reset on the basic knowns and unknowns of this pandemic, a task that must be undertaken with great humility. The virus keeps changing, and so does our understanding of the social and biological components of the pandemic. But in exploring both the knowns and the unknowns, we can see how complex the pandemic has become—and that we’re still lacking crucial data because of the failings of state and federal government.

read the full story