False Negatives

26 Mar 2020

A few days ago I wrote a post on how easy it is to have “false negatives”—negative for virus, but the person is sick. I made a table to show how easy it is to overestimate the efficacy of a test.

Paul Romer a co-winner of the Nobel Prize in Economics has a simple blog post where he shows that if we can increase testing, I don’t need to be so worried. He runs 50 simulations for the spread of the virus, based tests of different false-negative rates.

“The simulated data here contrast policies that isolate people who test positive using four different assumptions about the quality of the test. Even a very bad test cuts the fraction of the population who are ultimately infected almost in half. And when I say bad, I mean bad – an 80% false negative rate, which means that 4 out of 5 of people who are truly infectious will get a negative test result – i.e. a result saying that they are not infectious.”


There are big, important limitations to this. These simulations assume that each day 7% of the population is randomly selected for testing and completes that testing. The first limitation is practicability. It’s not likely that we’ll achieve (1) agreement on and (2) implementation of such a policy of random testing. Also, (3) compliance with testing might be low. However the biggest challenge could be (4) compliance with self-isolation requirements. The (5) scale of testing is also an important limitation. Romer’s model relies on the US testing 23.1 million people EVERY DAY. Cumulatively, the US has conducted 432,655 tests. (104P, EDT, 3/25/2020)

Nevertheless, this is strong evidence that my worry about insufficient oversight can be overcome if enough firms can—basically without impediment—develop, produce and distribute lots of test.

It also means that we might be able to rely on thermometers. Even if 40% of infected people are asymptomatic—see my post on Iceland’s testing—that’s still enough to reduce overall infections so we can manage the crisis.

Here we run up against what I think is the biggest obstacle of all: if people aren’t sure the test is reliable, they’re not likely to be willing to immediately accept the draconian measure of self-isolation.

No less an expert than Dr. Deborah Birx illustrates this. At a briefing Birx mentioned that she had a fever over the weekend and was careful and got a test. But when her fever receded, Birx figured she was okay to go back to work—before she got the results of her test. Coronavirus causes an intermittent fever (https://www.cdc.gov/coronavirus/2019-ncov/hcp/clinical-guidance-management-patients.html). And I can’t imagine anyone is more conscious of this than Birx. Of course, I’m not saying she was irresponsible—she might’ve taken extra precautions and she might not have let her guard down until she’d reached a sufficient time to be sure her fever wasn’t returning. But if Birx can skirt the requirements a bit, unreliable tests, unfortunately, may not do the trick.

False Negatives is built on the principles we teach in our live, online Product Science Bootcamp.

Stuck on a business problem? Don’t have time to attend our bootcamp?
Bring it with you to our Unstuck Assist session and Get Unstuck.

Free Consult