A thought inspired by this story about what business leaders apparently told Trump today. You want the economy to reopen? Then let’s see more testing. Lots more.

In the first phone call convened between President Donald Trump and some members of his newly formed business council, industry leaders reiterated to the President what public health experts and governors have been telling him for weeks: that there would need to be guarantees of ramped-up coronavirus testing before people return to work, according to one person briefed on the discussions.

The call, one of a series with various sectors on Wednesday, was the first task force teleconference aimed at devising a strategy for reopening the country.

Ideally we’ll have enough testing available by mid-May that we won’t need to guess who’s sick and who isn’t based on how they’re feeling on a given day.

But nothing about our predicament is ideal. There’s no reason to believe the feds will be on the ball with testing in May when they blew it at the start of the epidemic and are bedeviled by it intermittently even now. If they can’t get up to speed soon, are we supposed to keep the economy shuttered until they can?

It’s not just diagnostic tests they’ll need either. This is a full-spectrum challenge.

The U.S. must be able to perform four categories of tests, Giroir said — testing people with symptoms of the illness to confirm they’re infected; testing their close contacts to control the spread of the virus; randomly testing asymptomatic people to head off new outbreaks; and testing for antibodies in healthy people, to identify Americans who have already been infected and recovered…

Giroir said that if the U.S. was only seeing about 10,000 new cases a day, for example — the current rate is triple or quadruple that number — health authorities would have to perform at least 300,000 tests a month just to confirm those infections. To contain the disease, another five to 10 additional people should be tested for every positive — as many as 3 million per month…

But that doesn’t include a wide range of other testing — asymptomatic people, health workers, law enforcement personnel, or even the ordinary work force as a condition for companies to reopen their workplaces. It doesn’t include contact-trace testing, of people who’ve been in contact with confirmed cases — on average, 5 per case, or another 1.5 million tests monthly if there are 10,000 new cases daily. And it also doesn’t include antibody testing.

They’ve had trouble providing testing implements like swabs and reagents in the volume required by labs, enough so that testing by commercial labs actually declined by 30 percent last week in the U.S. What are we going to do if we run into another testing glitch a month from now, when businesses are cautiously reopening their doors in the expectation that sick people can be rapidly identified and removed from society?

A thought: Could we empower people to simply report the symptoms they’re having to local authorities in lieu of tests? It’s an imperfect solution, but a months-long shortfall in testing is also … highly imperfect.

As I understand it, doctors want widespread testing available when the country opens back up for two reasons:

1. At a micro level, they want to be able to identify infected people quickly and trace their contacts so that all who are infected can be quarantined.
2. At a macro level, they want to be able to spot clusters of infection, new “hot spots,” in communities before a small wildfire becomes a forest fire.

Both of those goals could be achieved (again, imperfectly) with self-reporting. Imagine if there were a smartphone app that allowed anyone to notify the CDC and the local health department if they woke up with two or more of the now familiar symptoms — fever, dry cough, sore throat, loss of sense of smell, chills, eye pain, chest pressure. Instead of the authorities having to wait hours or even days until a test was taken and the result processed, they’d know immediately that they have a probable COVID case and could act. If the app were widely available, a cluster of people reporting similar symptoms in the same small geographic area would alert them to a probable localized outbreak.

This is how the Kinsa system works, right? The difference is that Kinsa’s data on fevers is relayed automatically by its thermometers to the company’s databases. But it doesn’t need to work that way. It’s convenient, but the same process could be performed with analog thermometers by a group of people willing to faithfully type their daily temperature readings into the Kinsa app.

It’d be even better if our hypothetical app could quickly notify all of the people with whom the sick person has recently been in contact that they’ve been in proximity to someone who’s probably infected with COVID-19. But guess what: That app is already in the works. Seems like it’d be easy to build a self-reporting feature into it and then have major cultural influencers, starting with Trump, encourage the public to use the app en masse to report symptoms before they get tested. All contacts of a probable case could be warned by the app to self-quarantine until the probable case was either confirmed as infected or cleared by testing. If speed is important here, nothing’s as speedy as self-reporting.

The obvious objection is that we’ll have many false positives this way. Lots of illnesses produce fever and chills, including the good ol’ flu that’ll return in spades this fall. Many probable cases of COVID won’t actually be COVID. It’s a fair point, but let me gently remind you that much of our diagnostic testing is also producing inaccurate results. And not just false positives. False negatives. Would you rather have people err on the side of self-quarantine if they have a fever or have them out and about spreading the virus because a diagnostic test incorrectly told them they weren’t infected?

Testing would have a fantastic, unbeatable advantage over self-reporting, though, *if* it ever becomes so ubiquitous that we can test a chunk of asymptomatic people daily too. Someone who’s infected but not (yet) experiencing symptoms is potentially very dangerous:

1/3 New study shows that people with #COVID19 may be most infectious before their symptoms develop. The authors estimate that – on average – about 44% of secondary cases were infected from people who had not yet developed symptoms. https://t.co/C4KEeP1cMz

— Scott Gottlieb, MD (@ScottGottliebMD) April 15, 2020

There may be a lot of asymptomatic carriers out there. Self-reporting won’t work for them. But unless and until we reach the point where everyone with symptoms is getting tested quickly plus a batch of asymptomatic people (what criteria would be used to test someone who’s not showing any symptoms?), self-reporting seems like an obvious way to make up the shortfall in testing. It could work in tandem with other strategies to allow community monitoring, too. For instance, testing of local wastewater would also offer a clue as to whether local infections are increasing or decreasing.

Self-reporting seems like an obvious fallback solution, or even supplement, to widespread testing but I haven’t heard it discussed amid all the complaints about the testing shortfall. Why not? Is it just because the Apple/Google contact tracing app wasn’t officially announced until a few days ago?

There is some good news about testing out there, by the way. Antibody tests should be much easier to come by than they have been, and sooner than we think.