Michael Mina, MD, PhD, assistant professor of epidemiology at Harvard T. H. Chan School of Public Health, assistant professor in immunology and infectious diseases at Harvard T. H. Chan School of Public Health, tweets: "*speculative* – given very limited testing and relatively short windows of time to capture virus in a nasal swab, I really won't be surprised if even 50x more people have acquired the virus than cases confirmed If so ~11 million in US could have acquired #COVID19."
If that is so (a big if), it is great news, because with total U.S. deaths attributed to the disease at roughly 8,500, (people can argue whether that's an overcount or an undercount, but for now, let's leave that aside), it would mean that people who have Covid-19 have only a 0.077 percent chance of dying from it. (Okay, maybe some of those who have it now will eventually die of it, but even so, it would mean that so far, if you have Covid-19 in the United States, your chances of having died of it were 8,500 out of 11 million, or, very roughly, less than one in a thousand. That doesn't mean you should go out and lick the sneeze guard at the salad bar or otherwise take this coronavirus risk casually, especially if you have underlying medical conditions or are of an age that makes you at risk of getting a more severe case, but it's a different story from what is being more widely reported.
The only way to know whether it's true is to stop testing only very sick people or people who have had a known contact, and instead start doing some random, population-based sample testing. I've heard and read plenty of people be dismissive, or at least highly skeptical, of this idea that this virus is so widely out there already (Here is Dr. Scott Gottlieb doing that), but it would seem to me (and to some other people, including some physicians) that this would be a good use of testing—rather than using it only on very sick people on the way into the hospital or doctors and nurses and ambulance crews who have been exposed to those people, to test more of the general population who may have had a mild or asymptomatic case in January or February or early March.
As Aaron Carroll, a professor at Indiana University School of Medicine, put it, "We do need to be able to test many people who don't have symptoms, though, and at the moment we seem to be focusing all our testing on the sick and high risk. Epidemiologists are really good at this. They can tell you (for a population) how many people we'd need to sample to see a how much virus there is in a community. It would have to be random, though, and it would require a totally different strategy than we are employing now."