To understand a new disease, start at the beginning

Would you know if you had COVID-19?

The symptoms, we’re told, are dry cough, shortness of breath and fever. But a recent presentation by the doctors leading UCSF’s clinical response to COVID-19 made clear that symptoms vary quite a bit. Just three-quarters of all COVID-19 patients experience fever, they said, and only half have a fever when they come into the hospital. Dry cough is common, but a wet cough is also possible.

With media reports also circulating that some have lost their sense of smell and others have had cognitive confusion, one can see why an editorial in the New England Journal of Medicine recently called for a natural history of the disease. A natural history would include an accurate estimate of a virus’s incubation period; how best to diagnose COVID-19; the full picture of how and for how long it can spread from an infected person; how, when, and why the virus causes much more severe illness in some patients; and if and for how long recovered patients are immune.

“There's a lot of confusion,” said Dan Kelly, MD, MPH, assistant professor of epidemiology and biostatistics, who is proposing to develop a natural history. “There are important case studies that are being published right now around one person. One study found the virus shedding in one person for 49 days. Forty-nine days is much longer than people are being quarantined for!” Shedding virus doesn’t necessarily mean the person was contagious, which shows just how young the science on this disease still is.

To develop an evidence-based natural history, you need to identify patients who have recently been exposed and then take blood samples and swabs from their orifices every few days to see how much virus they are shedding and where. Normally the Institutional Review Board (IRB) that approves human-subjects research concerns itself with safety to patients, but in this case the resesarchers’ safety is a bigger stumbling block.

“You’re going out to go into somebody’s home where coronavirus is everywhere and taking a ton of samples from them and, you know, trying to find the infection,” Kelly explained. “People are like, ‘Tell me more about how you're not going to get infected.’”

If approved and funded, the study would track patients for 30 days. In 3-6 months, the researchers could publish a rough analysis, which would, even so, be an important contribution to what is known for certain about COVID-19. The final product would take about a year.

Others are likely also working toward a natural history: With most of the world’s scientists sidelined from their existing research due to social distancing and/or a focus on clinical responsibilities, the effort going into COVID-19 is massive.

“Basically, the whole scientific community is converging on this, which is really exciting. It makes me optimistic that we’re going to have answers at an incredible speed,” he said.