Photo by Lola M. Chavez

Every morning at 9 a.m., the San Francisco Department of Public Health updates the confirmed number of COVID-19 cases in our city, and the confirmed number of deaths. 

And that’s all. These two statistics, which could be shouted out a window, are the extent of the data publicly disclosed and readily available on the city’s website

The Bay Area region has won praise for its proactive steps to shelter-in-place and take seriously the world-altering potential of the COVID-19 pandemic. Six counties jointly issued sweeping health orders days and weeks before other municipalities and states. Thousands of lives may have been saved by this forward-thinking action. 

But with the data San Francisco publicly discloses, no one can truly know if these measures are working. Are we winning the war? Are we flattening the curve? It’s impossible to say with the numbers on hand; San Francisco omits the data you’d most want to know. 

In this we are something of an outlier. While the shelter-in-place order carries collectively across multiple Bay Area counties, the individual counties have taken a balkanized approach to data reporting. As such, Solano, Sonoma and Santa Clara Counties in particular report far, far more information than does San Francisco — including key measures such as hospitalization rates, hospital capacity, COVID-19 tests administered, and many other vital data points. 

San Francisco’s Department of Public Health offers daily updates on the number of confirmed COVID-19 cases and deaths … and that’s it.

“To have meaningful data, we need to know how many people are infected, how many are ill, how many are so ill they must be hospitalized, and how many of those people require ICU care,” says Dr. John Swartzberg, a clinical professor emeritus at UC Berkeley and UCSF specializing in infectious diseases. 

San Francisco discloses few of those measures — and listing the number of confirmed cases in this or any city is only so useful, considering the ongoing national fiasco regarding spotty testing. 

“We have a poor idea of how many are infected because we are not doing sufficient testing,” Swartzberg continues. “We have no idea how many people are ill but not hospitalized with SARS-CoV-2.” 

Health policies aren’t Peter Khoury’s stock in trade. But good data is. The Mission District data scientist has applied his expertise to crunching every number being reported by each Bay Area county. 

In San Francisco, that’s not really possible. There’s simply not enough data being disclosed to come up with any meaningful analysis. “You cannot know how San Francisco is doing,” he says. “And there are vulnerable populations.”

Anticipated outbreaks at Laguna Honda Hospital and among the city’s homeless population will likely soon drive our number higher — and, with the meager data this county reports, it’s hard to gauge if or when we’ll outstrip the capacity of our health system. 

Khoury critiqued the data presentations from each Bay Area county — an independent and parallel write-up to one published by Berkeleyside

Khoury (and Berkeleyside) rank San Francisco toward the bottom of the barrel. 

Parsing Santa Clara’s publicly available data, for example, Khoury created a chart revealing scary, exponential growth in the number of people hospitalized for COVID-19. Considering the haphazard and patchwork testing practices, hospitalizations are a far more definitive tally, he says. 

Santa Clara discloses this data. And since it also discloses its hospital capacity (and ICU capacity), Khoury could plot the day the county’s sick patients would figure to overwhelm its ability to care for them. 

While on a Thursday phone call with Mission Local, Khoury bemoaned that Santa Clara had, inexplicably, recently ceased listing hospitalization totals. 

In fact, the news site San Jose Inside last week reported on the abrupt disappearance of this vital data — and it has since reappeared. Informed of this, Khoury quickly checked Santa Clara’s webpage and gave a shout of joy. “This is good! This is awesome!” 

Based on a graph he produced last week — and barring intervention — Santa Clara County should’ve had some 353 people hospitalized by by now. But the present Santa Clara numbers show 245 — a sizable deviation downward off the projected curve and an indication that shelter-in-place measures are working

“They have put off their surge. It’s still up — but we are losing by less,” Khoury says. “The flattening is working. Getting this information consistently, across the Bay Area, is incredibly important.” 

In Solano County, Khoury continues, not only is all this data presented, so are case-by-case breakdowns by age. That way, we know that roughly half of people over age 65 diagnosed with COVID-19 have been hospitalized — but only one-fifth of those in the younger age group. This data is evolving and allows people to personally grasp what this pandemic means for them, their children, or their parents. 

Mission Local’s messages to the San Francisco Department of Public Health about the data it chooses to disclose — and not disclose — have not been returned. But a source familiar with the department and its practices and protocols tells us that “the DPH 100 percent has this data; they’re looking at it every day, they have it organized in spreadsheets, they can not not have been doing that.” 

So why sit on the data? Our inside source speculated it might simply be because “they don’t value putting it out.” They also fell upon an explanation other experts in the field came to: If the curve is flattening, even a little, there’s a fear that positive news will lead to complacency. 

“I think public health officials in California want to make sure people keep their heads down and continue following all these social distancing measures,” said Dr. Jake Scott, an infectious disease physician and assistant clinical professor at Stanford University’s School of Medicine.

“I get a sense that public health officials don’t want to give too much in the way of good news. They don’t want people slacking off.” 

That’s one way of looking at things. But failure to disclose pertinent data also breeds distrust. It also may harm coordinated efforts. To wit, San Francisco General Hospital and UC San Francisco have made efforts at transparency by disclosing patients in ICUs and other pertinent information — while California Pacific Medical Centers refuse to disclose totals of COVID-19 patients, patients in ICUs, or patients on ventilators. 

“Out of respect for patient and employee privacy we are not able to disclose the number of positive COVID-19 patients in our facilities or under our care,” a spokeswoman e-mailed us. “Even in times of emergency HIPAA still applies and it is our duty to protect patient, staff and clinician privacy.” 

Considering San Francisco General and UCSF’s positions, this is confounding. And, possibly, detrimental to coordinated efforts. 

“Making data inconsistent hampers an understanding,” Khoury says. “Having consistent information across our region matters; that was one of the lessons from Italy, too.” 

To be useful, Khoury says, county health departments should be disclosing the following: COVID-19 hospitalizations; available hospital beds; available ICU beds; total number of tests conducted (including multiple tests on individuals); total number of test conducted on unique individuals; and total number of tests that were positive for COVID-19. 

“Without good data,” adds Swartzberg, “you can imagine how difficult it is to formulate solid health policies.” 

Loi Almeron and Julian Mark contributed reporting to this article. 

Keep Mission Local on the ground. Support us today.

 

Joe Eskenazi

Joe was born in San Francisco, raised in the Bay Area, and attended U.C. Berkeley. He never left. “Your humble narrator” was a writer and columnist for SF Weekly from 2007 to 2015, and a senior...

Join the Conversation

26 Comments

  1. Thank you! The chron won’t challenge the authorities here, so thanks for the logical, in-depth reporting! The more data the better, San Francisco supes and the mayor had better step up some leadership here or they’ll be history soon enough!

  2. This numbers are a joke and completely meaningless. Part of the reason they’re able to keep numbers so low is that they’re refusing to test almost anyone!
    I’m currently watching my 41 year old, in great health, never sick husband get sicker each day and there’s nothing to do but wait and see if he’ll get better or worse.
    He’s currently on day 7 of symptoms and they refuse to test him unless he winds up in the hospital with respiratory distress but say he’s got presumptive Covid. In order to test, they need to be 65 or older AND have symptoms, but also he high risk, or have symptoms AND be a healthcare worker. So most of the population is out.
    So we are in quarantine now, and can’t leave the house for any reason. I’m terrified.

    Their numbers and tests are meaningless.

    1. Erica, I am aware of a case similar to your husband’s. That would be my niece’s significant other/boyfriend, who also had severe symptoms but was finally tested; he was turned away from testing at first but then was later tested. They live in the East Bay — in Oakland, I believe. I suggest that you persevere and press the powers that be to give him the test for the virus, for the sake of your own sanity as well as your husband’s health. I wish you all the best.

  3. I’m not one to trust the SFDPH to be honest about how it shuffles dollars to its nonprofit “community partners.”

    But when it comes to the Public Health operation, I’m prepared to give them more leeway.

    So long as they do not promulgate false information, I’m prepared to give them the benefit of the doubt on using communication to change behavior to elicit the best public health outcomes.

    We do not know, cannot know now, if the apparent flattening that we are seeing is noise at the left end of the incipient curve or whether we are actually flattening the curve by social distancing.

    So long as public health officials don’t lie, the purpose of this exercise is to minimize tragedy, not feed data points to bored data scientists to make cool presentations.

    1. Of course we can know if distancing is flattening the curve. ‘Noise at the left end…’ Say what?

  4. Numbers, and the lack of numbers, say a lot. But “the numbers” only tell part of the story. For example, we are told that the current coronavirus is much worse than seasonal flu, that seasonal flu does not “overwhelm” our hospitals and health care system. A careful look at that assumption reveals something else.

    https://off-guardian.org/2020/04/02/coronavirus-fact-check-1-flu-doesnt-overwhelm-our-hospitals/

    I don’t believe this is an argument for going out and partying, or lifting shelter-in-place. But on a day when we are told that the President of the United States only yesterday found out about a critical national shortage of supplies, we should not be quick to abandon what’s left of our capacity for critical thinking. The breakdown of the public (and private) health system in this country is nothing new, which many in public health have argued repeatedly for decades. Yes, I believe Covid-19 is a real threat. But the bigger threat is the politics that created the mess in the first place and continues to do so daily.

    1. We are told the current coronavirus is a novel virus. Experts are seasoned at dealing with the seasonal flu.

      All indicators show the novel virus to be deadlier than the common virus. A rate of 2X deadlier would be staggering. We won’t know an accurate estimated percentage for some time, mostly due to lack of testing capabilities.

      A virus with an equal rate to the seasonal flu would ‘overwhelm’ hospitals. Many people are able to avoid the common flu with a vaccination, not so with the novel virus.

      I’d be careful associating shelter-in-place with critical thinking, or comparing a pandemic threat to a partisan threat.

      1. I agree. You’ve described the fundamental differences between covid and the flu that result in more potential covid cases and each covid case requiring more medical resources than flu

        No one has immunity to covid, many(dur to vaccine) have (at least some)immunity to the flu. That means the possible number of covid cases is much larger than flu. And because of the lack of immunity, covid patients need to be isolated. That makes medical treatment more complicated than flu. And because it’s new, medical treatment for covid is less developed, so doctors assume the most aggressive treatments, requiring more resources. And finally covid appears to be deadlier than flu. So treatment takes longer, using more resources.

  5. Clearly, SFDPH’s limited release of covid metrics hampers any kind of predictive analysis of how effective the city’s early & proactive measures were in flattening the curve. Yet based solely on the reported morbidity rate alone, which in other regions has risen exponentially in just the last week, it seems remarkable that the death toll is still < 10 in a city of our size and density. Dr. Scott’s explanation of the city’s rationale in withholding “favorable” data confirms my own intuitive sense that public health officials fear a premature relaxation of physical distancing practices at the first sign that the rate of spread is slowing. While that position may have some validity from a herding perspective, wouldn’t full transparency also have the opposite effect as well —by incentivizing the sacrifices we are all making for the collective good? I for one would feel gratified to know that all the precautions taken so far are actually having a salubrious effect in halting the spread, making me more likely to redouble those efforts going forward. From a broader policy perspective, it seems imperative that San Francisco share its full range of data with other communities across the nation as a “best practices” model. The city’s “Stay the F- Home!” order of March 16 was issued only 4 or 5 days earlier than other localities, but given the transmission rate of the virus, that put us ahead by weeks, and potentially spared hundreds of lives here in SF. Yet until there is widespread testing and transparency, we are still whistling in the dark.

  6. Hey Joe… Exactly what is the mayor of SF doing these days. I see mayors and governors across the country getting in front of cameras and journos answering questions and dispensing info. SF Mayor? Ghosted…. I mean nothing. All I saw last week is how to get your hair done. Really? DPH not returning emails to press… No bueno. Would love some insight into what is really happening in the Civic Center. PS__Kepp up the good work you do.

    1. Agree Breed needs to be out there more. She or DPH needs to answer why they’re withholding data.

  7. Totally agree with this article – it has left people (like myself) tracking the data on their own which is just ridiculous. Open sourcing data to the public seems essential! It seems especially useful in San Francisco, a city that likely more computer scientists and engineers per capita than anywhere else in the nation.

    If this data was open sourced I am sure that clever folks would not only be able to model spread, but also quickly build maps with hotspots within San Francisco, or even build out contact tracing infrastructure.

    Why is this not being done!?

  8. CA per county (including SF) hospitalization (with ICU partition) for both COVID-19 patients and those showing symptoms:

    https://public.tableau.com/profile/ca.open.data#!/vizhome/COVID-19PublicDashboard/Covid-19Hospitals

    I can’t speak to how frequently this is updated, nor does any page in the board include time stamps. An archive would also be nice.

    I read about this in the following article:

    https://www.mercurynews.com/2020/04/01/coronavirus-first-look-at-hospitalization-data/

    While SFDPH is not reporting these metrics (crucial given the state’s testing short fall to date) on their website, the information is in the public domain

    1. Thanks, Tim! It’s so nice to see that the data is available. I am supportive of our mayor’s actions on covid-19 and particularly pleased that our city had the courage to be one of the first areas to implement shelter measures, and that they were further upgraded last week. That is the action that really counted and is helping us stay under the curve unlike so many other places.
      We have a president who is giving many press conferences with data and references – many of which have been mistaken or false .. and consequently the country at large has not understood the epidemic and has not implemented timely measures – and is, overall, on a worse trajectory than our city-county and state.
      As for the lack of testing available, that is also a federal issue. There still simply are not enough tests so the doctors are having to only use them when their results will affect the treatment. Sad but true – this from doctors I know ..

  9. I have, ever since this pandemic started, been looking for better day-to-day data. People are not going to want to social distance unless they see the curve flatten. DPH is, in San Francisco, a huge bureaucracy and HIPAA only complicates it. I think the Mayor needs to see to it that DPH kicks those thorough stats out. We are adults. And without thorough data, we are being treated like children.

  10. I found a mysterious number in the latest health order from San Francisco’s Department of Public Health, ORDER OF THE HEALTH OFFICER No. C19-07b. Item 9 states in part:

    This Order is also issued in light of the existence, as of March 29, 2020, of 848 cases
    of COVID-19 in the County, as well as at least 2,092 confirmed cases (up from 258
    confirmed cases on March 15, 2020, just before the Prior Shelter Order) and at least
    51 deaths (up from three deaths on March 15, 2020) in the seven Bay Area
    jurisdictions jointly issuing this Order, including a significant and increasing
    number of suspected cases of community transmission and likely further significant
    increases in transmission. 

    Where does the 848 number of cases in the county come from? On the Department of Health website, the count for today, April 3rd, is 497, considerably less than 848.

      1. Thanks, Andrew.

        That was my suspicion, too, and I appreciate you doing the legwork.

        JE

  11. Thank you for this article! I have tried to find out how to get more data from SF, with no success. Most recently, I was told to call 3-1-1.

    1. Sus — 

      Looks like it has updated, but San Francisco doesn’t list the last date of an update — which is pretty basic — so confusion like this can arise.

      Thanks for reading,

      JE

  12. This is the best reporting on what SF has not told us.
    I’ll be looking for your updates as, hopefully, more information is released.
    Thank you, Joe! You’re the best!

  13. How would we go about compelling SF to release this information? Who should we bother and what is the most effective way to bother them? I am sure they are doing what they’re doing with the best of intentions and I am generally really impressed by SF’s response to this emergency, but on the subject of public data they are making a poor decision and we should really be doing our best to insist that they correct that mistake.

Leave a comment

Your email address will not be published. Required fields are marked *