We rush to dismiss it as “mild.”

In February 2020, when the then-novel coronavirus still seemed far away, a reassuring statistic emerged: 82 percent of cases were mild—milder than SARS, certainly milder than Ebola. This notion would haunt our response: What’s the big deal? Worry about the flu! Since then, we’ve learned what mild in “most” people can mean when the virus spreads to infect hundreds of millions: 5.4 million dead around the world, with 800,000 in just the U.S.

This coronavirus has caused far more damage than viruses that are deadlier to individuals, because it’s more transmissible. A milder but more transmissible virus can spread so aggressively that it ultimately causes more hospitalizations and deaths. Mild initial infections can also lead to persistent, debilitating symptoms, as people with long COVID have learned. The notion of a mostly mild disease became entrenched so rapidly that the experience of many long-haulers was dismissed. We’ve seen how such early concepts can lead us astray, and still the idea of Omicron as an intrinsically mild variant has already taken hold.

We don’t know yet if Omicron is less virulent than Delta. We do know it’s far more transmissible in highly immune places. That’s enough for worry. We can expect Omicron cases to be milder in vaccinated people than unvaccinated. And because the variant is able to infect many vaccinated people that Delta cannot, the proportion of infected people who need to be hospitalized will look lower than Delta’s. What’s less clear is if Omicron is intrinsically any less virulent in unvaccinated people. Some early data from South Africa and the U.K. suggest that it might be, but confounding factors like previous immunity are hard to disentangle. In any case, Omicron does not appear so mild that we can dismiss the hospitalization burden of a huge wave.