Contribute Try STAT+ Today

Early in the coronavirus outbreak, as the first infected patients trickled into Zuckerberg San Francisco General Hospital, Diane Havlir noticed a troubling trend. Most were Latinx, most were men, and most were young. The infectious disease specialist wanted to understand why — and what it meant about how this new virus was traveling through her city.

Havlir, who runs the HIV/AIDS division at the University of California, San Francisco, knew she needed data, and a lot of it. While it seemed nearly impossible, she wanted to test thousands of people all at once, maybe even a whole census tract. She wanted to test those with symptoms and those without, in a Latinx population with a deep fear and distrust of the government. Many of her colleagues were incredulous.

“When we proposed testing everyone in a census district, it was met with enormous obstacles. Where? Everything’s shut. Who? There’s no volunteers. There’s no protocols, there’s no supplies, there’s no labs. There’s no systems to get data back to people,” Havlir recalled. “And think of the paradox of this. We’ve told people to shelter in place, to stay home, and now we’re saying, ‘Oh no, please come out and get a test.’”


She persisted. And in a matter of weeks, a hastily created coalition of hundreds of university, hospital, and community volunteers called Unidos en Salud pulled together one of the largest coronavirus testing studies in the nation. In a four-day blitz at the end of April, they swabbed and drew blood from 4,160 adults and children, including more than half of the residents in the 16 square blocks that make up San Francisco Census Tract 229.01. In the heart of the Mission District, it is one of the city’s most densely populated and heavily Latinx neighborhoods.

Support STAT: If you value our coronavirus coverage, please consider making a one-time contribution to support our journalism.

While Havlir expected to see the Latinx community hit hard by the virus, the actual numbers came as a shock. About 2% of people tested positive for the coronavirus. Nearly all of them — 95% — were Latinx. The other 5% were Asian or Pacific Islander. Not a single white person tested positive, though 34% of the tract’s residents are white, according to the U.S. Census; 58% are Hispanic.


“When we all saw these results most people were a little spellbound,” Havlir said. “It was sobering.”

The testing project is proving to be a national model, because of the logistical challenges the organizers overcame and what the results revealed about the dynamics of the Covid-19 epidemic in San Francisco. Team leaders say scientists have called from around the country, asking how they might accomplish something similar in their regions.

One of those was Richard Bennett, a professor of microbiology and immunology at Brown University who is part of a Covid-19 research task force looking into how his university can help the state of Rhode Island and how their labs could overcome barriers to testing populations widely. “We want to know how much virus is out there, what percentage of the population currently has it, and how many have already had it,” Bennett said. “If you can do more testing, including longitudinal testing, it’s a real help. UCSF is definitely ahead of the curve in this regard.”

One of Havlir’s motivations for the testing was to understand how the virus was being transmitted even after the city had been locked down for six weeks. Questionnaires administered with the tests gave her an answer: 90% of those who tested positive could not work from home. Most were low-income, and most lived in households with three or more people.

“What really comes out of these data is that low-wage essential workers are victims of this disease,” Havlir said. Many of those infected were working in food service, making deliveries, or cleaning offices despite shutdown orders. “These people were out working the entire time,” she said.

“Anecdotally, we knew this, but the hard data is heartbreaking,” said Susana Rojas, executive director of the Calle 24 Latino Cultural District and a leader of the Latino Task Force for Covid-19 that partnered with UCSF to run the study. “Our community was out working, keeping the city moving and fed. Of course they were more exposed and getting sick.”

The study also found that 53% of those who tested positive were asymptomatic, possibly transmitting the virus to others without realizing they had it themselves. This asymptomatic transmission is something Havlir calls the “Achilles’ heel of Covid-19 pandemic control.”

The results have led elected officials to call for changes in the city’s response to the coronavirus. San Francisco Supervisor Hillary Ronen, whose district includes the Mission, is pushing for the city to provide replacement wages, food, supplies, and hotel rooms for those who need assistance to self-quarantine.

“This study shows us that it actually takes resources to quarantine,” said Kirsten Bibbins-Domingo, a physician, epidemiologist, and vice dean of population health and health equity at UCSF who has been treating Covid-19 patients, a majority of them Latinx restaurant, delivery, custodial, and construction workers. “It speaks to the many ways risk happens in this pandemic.”

From the start, the leaders of Unidos en Salud were committed to the idea that their project was not only about testing, but about providing care to the people who tested positive. The project reached 71 of 74 people found to be infected. Only 30 had primary care physicians, so the team linked them with doctors. The project also provided hot meals, tubs full of cleaning supplies, masks and gloves, groceries, gift cards, and hotel rooms for those who needed them. Rojas said it was often difficult to get people to accept help even if they were sick. “We’re a proud community,” she said. “A lot of people said, ‘Oh you shouldn’t take the trouble.’”

Community volunteers would stay on the phone for hours, speaking Spanish and building relationships, to ask people about their symptoms and what they might need in order to recover and stay home. “We wanted people to have dignity while they were healing,” Rojas said. “These weren’t just transactions. People started to feel like family. They started asking, could you bring me a few lemons? And a new toothbrush?”

Mission_Latino Task Force w/ Diane
Members of the Latino Task Force for Covid-19 with UCSF’s Diane Havlir (left) in the Mission District. Courtesy Barbara Ries

While the Mission project shows mass testing of communities is possible, it’s certainly not easy. The first challenge for project volunteers was figuring out how to encourage a marginalized and minority population to agree to be tested. It’s a community where many people are frightened about the virus but also lack health insurance and are mistrustful of the government.

“People were asking, ‘Could this impact my status if I’m undocumented? Who gets this information? Are they going to use this to try vaccines on us?” said Jon Jacobo, a leader of the Latino Task Force for Covid-19 who helped lead the study. “And then people were saying, ‘Dude, you’ve been telling us for weeks to stay inside and now you want us to come outside?’”

It became clear almost immediately that the outreach strategies often used for medical studies, such as websites for signing up, were not going to work. “Right away, we noticed the wealthy, well-informed people who live in the Mission had taken the first 1,000 testing spots. They were using Nextdoor and social media to inform each other and pass on the URL,” Rojas said. “Once we saw that, we knew we needed to change our strategy if we were going to reach the most vulnerable and fragile people. We called in everybody we knew and started going door to door.”

The team brainstormed on how to reach people who had no internet access. They provided all information in Spanish. They leafletted. They went to bodegas and parks. They brought food, along with flyers about testing, to day laborers. They gave “golden tickets” to people who sheltered in the area but had no physical address so they wouldn’t be turned away from test sites. They rode around in cars with megaphones and P.A. systems. “We went old school,” Rojas said.

Jacobo, an affordable housing advocate who lives in the census tract and served as the Latino press secretary for Bernie Sanders’ presidential run, likened the effort to political work. “This was really employing a campaigning methodology, thinking of it as if it were a ballot measure or a candidate we were pushing, but really it was the testing,” he said. “What that translates into is door knocking on all 1,400 doors of this census tract, flyering all 1,400 homes in this census tract, and phone banking.”

The testing days were a whirlwind, with more than 1,500 tests given on the single busiest day. Testing occurred outside in parks and elementary schoolyards, in tents festooned with balloons, with plenty of social distancing and masks.

Collecting so many blood samples was one challenge. Running them and getting fast results was another. (A mass testing project in Telluride, Colo. was less successful because it didn’t return results for three weeks.) The team wanted a 72-hour turnaround so results would be useful for those who tested positive and for public health officials who wanted those people to self-quarantine and be followed up by wellness teams and contact tracers.

For this, they turned to Joe DeRisi, an expert on viral diseases whose “ViroChip” helped identify the coronavirus that caused the SARS outbreak in 2003. “Diane said this would require processing 4,000 samples in a very short period of time,” said DeRisi, a professor of biochemistry and biophysics at UCSF and one of the leaders of the CZ Biohub, a research center funded by Facebook CEO Mark Zuckerberg and his wife, physician Priscilla Chan. “I said, ‘Bring it on.’”

In the early days of the pandemic, with reagents and swabs in short supply, UCSF had been able to run only between 60 and 140 tests a day at its clinical labs. After Gov. Gavin Newsom relaxed requirements on who could staff diagnostic labs on March 12, the CZ Biohub could tap skilled graduate students and postdocs to run Covid-19 tests as well. In just over a week, DeRisi converted empty UCSF lab space near the Biohub into a testing facility, equipped with eight robots that could run thousands of tests a day.

The expanded lab is offering free test analysis to any of California’s 60 public health districts and has run some 20,000 coronavirus samples. But the Mission project, with so many samples coming in together, was a true test of the lab’s capability and volunteer army of 50 graduate students, postdocs, and even faculty.

“It was crazy here. We had tons of tubes showing up here late into the night,” said DeRisi, who pitched in to help volunteers organize racks of tubes as samples flooded in. “It was such a crush. It was all hands on deck and it was awesome.”

In the days before the Mission project, the lab warmed up by processing in four days nearly 2,000 samples from another UCSF testing project in Bolinas, a semi-rural coastal community in Marin County, 13 miles northwest of San Francisco. The contrast between the studies ended up proving informative. In the largely white community, teams in tents tested 1,847 people who walked up or drove through in cars, bicycles, and “a really sweet Vespa from the 1960s,” said the project’s leader, Bryan Greenhouse, a physician and malaria researcher at UCSF.

Not a single person in Bolinas tested positive for coronavirus. ”It’s not too far away, but a world of difference,” Greenhouse said.

Both projects are waiting for the results of antibody tests that were also administered to see how many people may have been previously exposed to the virus but no longer have active infections. Expected in early June, the results could help clarify how widely the virus has spread in the Bay Area, an issue that remains hazy after a contentious antibody study from Stanford suggested that up to 4% of people in Santa Clara County had been infected with the novel coronavirus.

The Unidos en Salud team wants to keep tracking the virus and its spread through the Mission; it’s planning follow-up studies in three and six months. Many of the team members, who’ve been working nonstop for weeks, are still processing the results, and the anger they feel about how few resources their community has to deal with the virus even as they risk themselves doing the essential work San Franciscans rely on. “As rich as this city is, as expensive as the housing is, it would not function without our community,” Rojas said. “If Latinos stopped working, who would have food? No one.”

  • For Pete’s sake STAT. I was expecting more than this from you. I agree with most other poster’s here. The “virtue signaling and political pandering does not look or feel right in a publication like STAT. And most of us see right through what you are really attempting to accomplish in this instance. We are not a bunch of useful idiots. You are becoming what C. S. Lewis referred to as “moralizing busy body’s”. Stop it. Please…Latinx, give me a break.

  • This article is an excellent example of why no figures provided are reliable. The information provided is all emotional/political.

  • Did you test every single person in this tract? 4000 is not the entire population of that neighborhood. Political pandering at its worst.

    Also I work in the mission and after several visits to local liquor/grocery stores during my lunch break, I can see why Hispanics were hit hardest. The individuals I saw in these business were not social distancing, not using protective covers, nor were the employees monitoring store occupancies. This was over the last two weeks…

  • What was the racial mix, age, gender, and pre-existing conditions of the 4,0000+ tested and those that tested positive?

  • This is what happens when media and government is controlled by corporations. It looks almost genocidal to send people back to work at these low wage jobs.

    • I notice that all these areas where this disparity exists are controlled by Democrats. Could it be Democrats are causing these problems and disparities? There aren’t any Republicans in LA.

    • BillyG, some folks fail to see the forest for the trees. They will say anything to promote a greater agenda.

  • The most important piece of data is that 2% of all people tested were found positive for corona virus, especially in an area of homelessness and lack of adequate health care. It sounds like someone is doing something right. This is good news. In Atlanta of homeless persons tested 4% were positive and in Seattle 5%.

    It is concerning that almost all of that 2% were Latinx and good to hear that people with the virus were followed up with and given medical care and help with food, clothing and housing. We aren’t given any information about fatalities or hospitalizations. We can hope since most of those had the virus are purported to be young that they recovered well.

    This 2% infection rate should be studied further.

  • Does everything have to be political? I am not surprised that more Latin-Os (I will use LatinX when over 50% of the Latin-Os use it, as that is just plain courtesy, but not because someone whose gender politics do not align with probably 95% of Latin-Os, an “Usha McFarling” tells me to)

    I appreciate the Latin-Os are doing a lot of work to keep the city moving. They have a good attitude and strong work ethic and are taking on significant risk to work.

    HOWEVER- political explanations for everything are not only not science, they are antithetical to science. These folks ARE exposed at their jobs, I am sure, one can see that clearly -but the idea that explains all or nearly all the difference in infection rates is pretty specious. If about 1/3 are asymptomatic, perhaps mostly because of youth, then it is easy to see how 95% got infected – and, while their jobs on average are more dangerous, and they kept going to work when more privileged people stayed home – that is not all due to social inequities – cultural differences can be important – and of course, more crowded housing, which, most of the time, is not a medical issue.

    This is not really medical news – at least, not beyond the stats – and from death rates at least, it appears the Latin-O people are weathering it better than the blacks or whites, so it is not really clear it even has much medical importance. Mostly just political harangue.

    • Forgot this – thanks to all involved for getting the testing done and trying to help public health.

  • 1. Could this not have used sewage testing to narrow down the virus to the mission district more economically and more anonymously since the subjects are afraid of the government and protecting their immigration status?
    2. We can assume employers followed the masking, hand washing, and distancing rules when low paid workers are on the job cleaning and food preparing and stocking. Does this indicate the workers don’t comply or that it is not effective when used by Latino races? How do Latino servants and gardners of those well paid white ocean front “lockins” working at home not infect their employers? Not even one rich white Bolinas testee in a rural area hiring low cost field and ranch help? Really?

    • Bolinas is a community known for being reclusive, plus it’s north of the Golden Gate Bridge and difficult to get to. That may be why it was chosen as a control.

      I live not far from an area known as Little Mexico — the stores there are blowing off the restrictions, restaurants are open and crowded, no masks, no distancing.

  • Is there a racial breakdown of the people actually tested in the study? The article states that 34% of the census tract is white, but doesn’t state how many of the people tested were white.

    • that’s exactly what i was looking for too. if 0 (or only a handful of) white people showed up for the testing, then it wouldn’t be surprising that 0 white people tested positive; but that wouldn’t prove that 0 white people had the virus. it’s a bit suspicious that they did not include that data, since it could change the narrative, and it’s a simple stat for them to include.

    • You’d think a publication with the word “stat” in it would have included such a critical detail, but that would have distracted from the sensationalism of the headline that “no white people” tested positive.

  • Great job to the entire movement and collaborations amongst the best in the area. Data driven policies help secure the future.

    • The data here is specious and missing in chunks, such as how many white people were tested? How many black people were tested? Are we to believe that the Mission is only Latino and Asian? If so, then it’s not really all that surprising that none who were infected were white or black.

Comments are closed.