Don't waste coronavirus tests on those already showing symptoms. There's a smarter way.

Universal testing in the United States may be on the way, but it is not around the corner. Acknowledging that regrettable reality means that, in the meantime, we have to decide how to allocate wisely our scarce testing resources. Unfortunately, with limitations on testing we seem to be testing the wrong people for SARS-CoV-2, the virus that causes COVID-19.

Most authoritative sources, including the Centers for Disease Control and Prevention, recommend testing people with symptoms. Though that sounds reasonable, it leads to squandering our testing resources on people who we already know are probably infected. Moreover, with high false negative rates from some of the tests, even a negative test in someone with classic symptoms should be assumed to be infected.

Until we can achieve universal testing, in order to control the spread of this virus, a better approach must be taken than the one in use today. Starting now, we need to spend vast numbers of available tests on symptom-free essential workers with a high number of daily contacts.

To understand why this is, we first have to understand the two reasons why we test to begin with: surveillance and intervention.

Test smarter

Surveillance testing aims to tell epidemiologists how common a disease is. Attempting to glean this simply by tracking the percent of all positive tests being done in a community is prone to sampling bias. If most tests are done in a hospital intensive care unit, the numbers will artificially high. Alternatively, if too many are performed in a relatively unaffected parts of town, the numbers will be artificially low.

Proper surveillance can be achieved with random sampling, using with just a small fraction of our testing resources. Such an approach is used all the time in polling data, which are generally accurate within a few percentage points, when done correctly.

Intervention testing, in contrast, is designed to break the chain of transmission by lowering the number of infected people who in turn infect others. The key to this virus is that many people are infectious before they develop symptoms — and many never develop symptoms. They become highly efficient spreaders. We need to take people out of circulation who are shedding virus without knowing it. The way to do that is to pick people who are out and about — and interacting with others and test them often to be sure they are not silent carriers.

People with symptoms should be taken out of interacting with others reflexively, regardless of any test result. Testing them should not change anything. It is those who do not know they are spreading the virus that we most urgently need to find.

Enough! Coronavirus: I was in the stay-home-until-it's-safe camp. But I just can't take it anymore.

In either case, we also should notify people that infected individuals have seen recently and warn them that they might be infected — and if they have symptoms they should self-quarantine. If not, they should be tested too and, in an abundance of caution, separate out for a bit.

The virus has a big advantage

Here’s a starting framework for how we believe our finite number of weekly tests ought to be expended.

First, a large number of tests should go towards universal and frequently repeated testing of nursing home residents and staff. A staggering proportion of fatal cases of COVID-19 have originated in long term care facilities, and we must monitor these facilities meticulously. Correctional facilities also deserve routine and widespread testing.

From there, we would mostly test symptom-free essential workers who interact with high numbers of people. While some of these tests might go to health care workers, we need to focus intently on other professions as well, including transit employees, front-line police and fire personnel, grocery store workers, and others in high contact service industries. Even more than health care workers, who are wrapped in protective gear, these other groups are at risk and can be efficient spreaders of the virus.

Sen. Rand Paul and Rep. Andy Biggs: Anthony Fauci wants America closed until there's nothing to reopen

This approach reflects an important reality; infected individuals who do not have symptoms are a major reason this pandemic has occurred and also why it will be so hard to quash. It is this evolutionary advantage that SARS-CoV-2 exploits so well.

Unlike other respiratory viruses, this one seems to have a remarkable ability to evade detection by our immune systems, at least initially. Instead of making all of its hosts to “get sick” within a couple of days, causing us to voluntarily self-isolate as we usually do while riding out a fever or a nasty cough — a remarkably advantageous evolutionary strategy that we have developed that protects our own communities as much as any personal immunologic functions these misery-inducing strategies may serve — this virus can circulate in our bodies in stealth for many days.

Our strategy identifies hot zones more quickly and allows for earlier intervention. Shutting down one grocery store for a few days where one or more symptom-free case has been detected (while all employees can be tested and the store can be cleaned and abandoned for long enough for trace particles to die out) would be an example of an actionable intervention directly resulting from our strategy that other strategies simply omit wholesale.

As we move forward, we predict that the public’s willingness to cooperate with system-wide shelter-in-place orders will diminish. However, there will be utility in doing some of this. Our strategy identifies hot zones sooner, and allows a more surgical approach to future shutdowns. If we can limit shutdowns to relatively smaller areas, we will inhibit spread of the virus without disrupting other parts of the economy that can reasonably carry on. We believe this approach will save many lives while permitting our system to slowly come back online in a stepwise fashion and that can respond to on-the-ground circumstances.

Dr. Jeremy Samuel Faust is an emergency physician in the Division of Health Policy and Public Health at Brigham and Women’s Hospital in Boston and an instructor at Harvard Medical School. Follow him on Twitter: @jeremyfaust

Dr. Harlan Krumholz is a professor of medicine at Yale and director of the Yale New Haven Hospital Center for Outcomes Research and Evaluation, one of the nation’s first research units dedicated to improving patient outcomes and promoting better population health. Follow him on Twitter: @hmkyale

You can read diverse opinions from our Board of Contributors and other writers on the Opinion front page, on Twitter @usatodayopinion and in our daily Opinion newsletter. To respond to a column, submit a comment to letters@usatoday.com.

This article originally appeared on USA TODAY: Make COVID-19 testing more strategic — don't test those with symptoms