Reopen Society Now? Let's Not Land This Plane Blindfolded

F. Perry Wilson, MD, MSCE


April 21, 2020

Find the latest COVID-19 news and guidance in Medscape's Coronavirus Resource Center.

This transcript has been edited for clarity.

Welcome to Impact Factor, your weekly dose of commentary on a new medical study. I'm Dr F. Perry Wilson, coming to you from my office here at Yale University.

This week, Iceland.

Iceland is more than fjords, Björk, and my favorite volcano, Eyjafjallajökull.

Iceland is a bit of an epidemiologist's dream. An island with only one major port of entry (the international airport in Reykjavík), a highly literate population, and a universal healthcare system provides nearly perfect conditions to study how infectious diseases spread. So it is no surprise that Iceland provides the best data we have yet on the community spread of the novel coronavirus.

Appearing in the New England Journal of Medicine, we have this study, which is really three cohort studies in one, with each cohort defined by a testing strategy—and only one mirrors what we're doing in the United States.

First, we have what you might call the standard strategy, targeted testing. You test people who you think are likely to have COVID-19: those with classic symptoms, exposure to an infected individual, or recent travel to a high-risk area. This is essentially what we are doing in the United States.

Iceland started its targeted testing program in January, 1 month before the first COVID-19 case was documented in that country. And they ramped up quickly. Though, to be fair, at 350,000 individuals, the population of Iceland is just one tenth that of my home state of Connecticut.

From January 31 to March 31, the targeted testing program tested around 9200 individuals. You can see the percentage that tested positive here, as high as 14% in the later weeks as the population prevalence was increasing.


This is close to the positive rates we're seeing in testing in the US, where currently around 20% of tests are positive.

But targeted testing doesn't tell us much about how the disease is spreading; it really just represents the tip of the iceberg. The big question is, how much of that iceberg is underwater? How many COVID-19 cases are we missing because they are minimally symptomatic?

This question is hugely important. If there is a high rate of minimally symptomatic infection, this would be great because it would imply that the death rate from the virus is much lower than the 1%-3% currently reported. It would also imply that herd immunity is developing, meaning, it may be safer to remove some social distancing provisions more quickly.

The Iceland study helps us answer this question with its two other cohorts. From March 13, all Iceland inhabitants could get a coronavirus test if they wanted one. This was a PCR test, not an antibody test; it would show active infections. But given the timing, it is unlikely that many Icelanders had already been through the disease and come out the other side. And, of course, people who had some symptoms were more likely to choose to be tested (57% of this group reported symptoms). Overall, you can see a dramatically lower rate of test positivity in this population-based testing group.


And finally, we have a randomly selected cohort, personally invited by the health service for testing, independent of symptoms. Though only 34% of the 6782 invited Icelanders responded to the invite, this still represents the best measure yet of how much of the underlying Icelandic population was infected, at least between April 1 and April 4, when this group was tested.

And so we see that the background rate of infection in Iceland in early April was around 0.6%. That would extrapolate to about 2100 people. On April 1, Iceland had 1220 confirmed cases of COVID-19. Obviously there is some lag here, but it doesn't seem crazy to think that there are probably around two individuals infected for every one whose symptoms brought him or her to medical attention.

This is not great, actually. While it implies that Iceland's social distancing policies have limited the spread of disease, it also implies that there is still a long way to go before herd immunity develops, and also that the observed COVID-19 death rate—while higher than the true death rate—is not like an order of magnitude off.

Population-based testing has helped to answer a couple of other questions that had been percolating around. We know that more men have died of COVID-19 than women, but is that because the disease is more severe in men, or are men more likely to become infected? The population-based data show that men are indeed a bit more likely to become infected than women.


Why that is an open question. It could be genetic, or maybe we just aren't as good about washing our hands.

The data also encouragingly show that kids have a lower prevalence of even asymptomatic infection. In the population screening group, none of the nearly 1000 kids under the age of 10 tested positive. If replicated, this would lend support to reopening schools.

More than anything else, the Iceland data show us the value of a robust community-testing apparatus, something that we are sorely lacking in the US. Making decisions about reopening society without information about the population prevalence of this disease is like trying to land a plane blindfolded. Oh, it can be done, but I wouldn't want to be on the hook for the outcome.

F. Perry Wilson, MD, MSCE, is an associate professor of medicine and director of Yale's Program of Applied Translational Research. His science communication work can be found in the Huffington Post, on NPR, and here on Medscape. He tweets @methodsmanmd and hosts a repository of his communication work at

Follow Medscape on Facebook, Twitter, Instagram, and YouTube


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.