Absent fast tests, the hospitals were being forced to treat anyone who rolled into the parking lot with coronavirus-like symptoms as if the patient had the virus, when more often than not, the patient did not. Beds in the coronavirus wing were being taken by people who didn’t need them; nurses and doctors were running through scarce protective gear they needed for actual coronavirus cases. But the biggest problem of having no tests was not knowing where the virus was, and where it was not. Without fast tests you could not isolate the people who needed to be isolated, or free up the people who didn’t.
On March 12, DeRisi decided to turn the Biohub into a coronavirus testing lab. He enlisted the help of 200 University of California at San Francisco graduate students, and persuaded Newsom to issue an executive order that would allow the Biohub to operate under U.S. regulations, rather than the more stringent California ones, which prevented people not specifically trained to work in a clinical lab from doing so.
His student army organized themselves into squads kept separate from one another — so that if one person contracted the virus it would incapacitate only one squad and not the entire army. Each squad had its hierarchy, and people learned the jobs of the people immediately above them, so that they could replace them if necessary. “We built in battlefield promotions,” says DeRisi. They looked into what was going on in the various markets, all tight, for the stuff you needed to run tests — the plastic jugs of chemical reagent, for example. They designed the lab around the supplies they were sure they’d have access to.
The equipment — well, that was its own story. American medicine is littered with machines designed for testing. But they lay idle because the companies that made them were failing to supply the reagents they needed to work. UCSF itself had a machine, called the Panther, capable of processing 1,400 tests a day, but it was just sitting there unused for this very reason. “Big clinical testing machines are designed on the razor/razor blade model,” says DeRisi. “You can’t use them without their razors.” His graduate students either built the machines they needed themselves or grabbed wherever they could find that might be adapted. “Here is the frightening aspect of the global supply chain,” says DeRisi. “When there is a surge in demand, inventory goes to zero.”
Eight days after the 200 volunteers started building the new testing lab, the Biohub was up and running. Since March 20, it’s been able to process 2,500 tests a day, at no cost. Quest Diagnostics, one of the biggest diagnostic labs in the country, was saying that if you sent them a test, it would take them a week to get back to you with an answer, along with their bill. In just one day more than it took Quest to process a single test, DeRisi built an entire lab that could process 2,500 of them. A day. For free. Any given test it could process inside of three hours. DeRisi was pretty sure his lab was bigger than any lab in the country that was faster than his, and faster than any lab that was bigger.
It was also different, in an important way. The commercial labs are set up to take in the samples and spit out a simple answer: positive or negative. They aren’t set up, as the Biohub is, to sequence the genome of every positive specimen and look for variations among them. As it moves through the population, the virus replicates and, as DeRisi says, “every time you replicate something there is a chance of an error.” He’s already found tiny differences from one coronavirus genome to the next — not so great that it changes the essential nature of the virus but noticeable nevertheless.
Those tiny disparities can paint a picture of how the virus has traveled and how widely it has spread. It’s a bit like a game of telephone. A group of people sit in a circle. A person starts the game by whispering a little story into the ear of the person to his right, who then repeats it to the person on her right, as best as she can. As the story travels around the circle, tiny flaws get introduced. If you knew the flaws — both the number and the nature — you could make good guesses about the number of people playing the game, and also where they were sitting in relation to one another. That’s exactly the sort of information we’re all now starving for: How is the virus spreading, and how many people have had it.
The lab is a single room the size of half a basketball court organized into stations, and staffed by graduate students, in white coats and masks, working in shifts of roughly 60 people, around the clock. DeRisi walks me around and introduces me. The masked people are from everywhere in the world, and the only obvious pattern is that, when asked where they are from, the Americans answer with a city or a state, and the non-Americans with a country: China, Taiwan, Colorado, Tanzania, Lithuania, Florida, Canada, Phoenix, Belgium. None of them are doctors; they’re research scientists, many with PhDs; exactly zero of them have ever done the jobs they’re doing now. “It’s different from a regular research lab where people just do what they want and come and go as they please,” says DeRisi. “It’s more like a factory. A production line.”
The graduate students were still getting used to the idea that some actual human is out there on the edge of a hospital bed, waiting to hear what they’ve found. One scientist tells me she has a version of a recurring dream she used to have when she worked as a cocktail waitress. “I used to have these terror dreams that we’d have 100 people need a cocktail at the same time for happy hour,” she says. “Now I have the same dream about thousands of people needing test results at once.”
But this isn’t a dream. It’s the problem consuming American society at every level, even if it isn’t always put that way. Every question about when the country reopens is actually a question of when testing will be good enough to allow it to reopen. Every question about the virus itself — how lethal it is, how many people have it without knowing they have it — can only be answered by more testing. In mid-March, when DeRisi got started, the entire state of California was generating test results for only about 2,000 people a day — with tens of thousands of tests backed up in commercial labs. DeRisi knows he can help close this gap, doubling or even tripling his lab’s capacity to 6,000 a day from 2,500. In a sane world, the Biohub would be doing just that. Instead, it is receiving only about 200 tests each day.