Skip subnavigation and go to article content

COVID-19: What the Numbers Mean

By David Tuller, DrPH

 Faculty Headshot for John Swartzberg

John Swartzberg, MD

This article originally appeared on the Berkeley Wellness website.

In our continuing series of interviews with John Swartzberg, MD, about the coronavirus pandemic, we asked him to take us inside the COVID-19 case and mortality numbers we’re seeing reported each day. That is—what do they really reflect? Are enough people even being tested to give an accurate view of what’s happening on the ground? What can we expect in the coming weeks? Here is an edited version of our conversation on April 4.

Can you put the numbers we’re seeing in some context?

The first thing to say about the numbers we have today is that they’re not very accurate. If we look at the numbers of people reported as infected, that’s supposed to be based on the total number of people tested. But that assumes that all of the data are being reported to public health departments, and that assumption is not what is happening in many cases. A lot of labs are only reporting positive tests, some are not reporting at all. They’re so stretched just getting the results out, and reporting to the public health authorities is another layer of work.

And the labs are also so backed up that last week it was taking between five and seven days to get results back. And that means wasted resources. You’ve got patients in isolation so they’re taking up a needed bed, you think they might have COVID-19, and you need to wear personal protective equipment (PPE) while treating them. And it takes you up to a week to find out if they’re negative or positive. If they’re negative, that means you’ve wasted that isolation room and that PPE on someone you don’t need to use it for. And this epidemic started in flu season, so many patients in that situation had influenza, not COVID-19.

What is the impact of this shortage of testing?

Basically we’re still not testing enough people to ascertain how many really are infected, and we’re not reporting reliably to public health authorities. That means our surveillance data is sub-optimal. And without knowing where you are in the epidemic, how do you develop public health policy? Two major problems account for the current situation. The first is the complete fiasco with testing—that involves the failure of the CDC and the FDA under the Trump administration. The other is the larger failure of our society to fund public health over several decades, so that it can do the job it needs to do.

What should we have been doing?

The U.S. should have had the tests and testing equipment available on a timely basis. We had the time to prepare for this. We should have been testing everybody we could, and a month ago we were hardly testing anybody. That would have allowed us to do contact tracing and isolating people when needed, but that didn’t happen. South Korea and Taiwan both did that. By extensive testing and robust contact-tracing, and having any contacts shelter in place, they were able to minimize the spread and avoid shutting down completely.

Can you explain why the death toll is expected to keep rising even though so many of us have been sheltering at home for weeks now?

You can actually count your way backwards to understand this. The time from infection to becoming symptomatic is from two to 14 days, but the average is five days. The average time from developing symptoms to presentation at the hospital is seven to nine days—so that’s at least 12 days. And when a patient dies, it is typically after a two-week stay in the hospital. So that means that deaths taking place today represent infections that took place around a month ago. That means we will keep seeing deaths from infections that have occurred in the past weeks.

Can you explain the current understanding of “asymptomatic transmission”?

We know people can be infected with this virus but not have symptoms. Some of those people who are infected may go on to develop symptoms a few days later, or some might never become symptomatic at all. Actually, a lot of people we’re calling asymptomatic are really minimally symptomatic but might not even notice it. We don’t know what that number is, but it wouldn’t be unreasonable to think that half of those infected might not have symptoms or have symptoms so mild that they don’t realize they might have COVID-19.

What we don’t know is how effectively many of those people can transmit the virus to other people. Some evidence shows that this does occur. We don’t yet know enough about this route to be able to say how common it is, but it seems to be less of a factor than other routes.

So what is known currently about the most frequent forms of transmission?

The data are clear—the most common way is through respiratory droplets, which symptomatic people emit, especially when they sneeze or cough. These droplets can potentially infect people within six feet—that’s why we emphasize staying that far away from other people. The second most important form of transmission is likely through inanimate objects on which the virus can stay viable for a period of time. If someone infected touches an inanimate object, which is then touched by someone else, and if the second person then touches their mouth or nose or eyes with that hand before washing it, they can become infected. In epidemiology, we call these inanimate objects that can be vectors of infection “fomites.”

There is also evidence that in certain circumstances viral particles can remain suspended in the air for a period of time and travel farther than six feet. That means someone could possibly get infected by being at a distance greater than six feet away from someone infected. From what we know now, this (airborne transmission) and asymptomatic transmission are vying for third and fourth place in how people get infected.

We’ve heard that antibody tests will be available soon. What will these tell us?

When people are infected with a pathogen they have never previously encountered, they start creating antibodies to fight it. An antibody test for the coronavirus that causes COVID-19 was just cleared for use by the FDA, and there should be many more available soon. Since no one in the U.S. would have had antibodies to this coronavirus before January, the test can tell us who was infected in the last three months. Where this is really going to be helpful is in telling us how many people have been infected overall. The numbers could be 100-fold or more than we currently think.

Does having antibodies from past infection confer immunity? There’s a good chance it will, but there are infectious diseases where it doesn’t, like syphilis. And assuming it is protective, how long will that last? Is it like measles and many diseases, in which immunity after an infection is lifelong? There are four coronaviruses that cause the common cold in humans, and when we get one of those our immunity lasts between one and three years. Our immunity to SARS appears to last about three years.In the case of this new virus, we don’t yet know.