Today's update on SF/@UCSF numbers. Bottom line –huge increase in cases & test positivity. Hospitalizations still pretty benign (though early signs of an uptick). Most remarkable stat is our asymptomatic test positivity rate (ATPR) @UCSFHospitals – now ~7%, or ~1-in-14.(1/16)
You'll recall that ATPR is fraction of pts getting care (hospitalization, procedure) @UCSFHospitals who we test for Covid routinely, even tho they have no symptoms. I use this as a rough approximation of the prevalence of Covid in Bay Area people who have no Covid symptoms.(2/16)
The ATPR was as low as 1/500 (0.2%) a few mths ago. Today, our 7d ATPR is 6.9%, and last 3 days it's ~8% (going up fast).
If its 8%, it means 1-in-12 people in SF who feel fine would test positive for Covid.(Not all would be in their infectious period, but most would be.)(3/16)
Does that mean that we should treat everyone – even fully vaxxed – as if they might have Covid? With those odds, that wouldn't be wrong. Does it mean I'm going to wear a mask in my home w/ family? No – I know they're all vaccinated x 3 shots & being careful. But still...(4/16)
...the (scary) math is what it is, & Omicron's breathtaking infectivity is changing the odds in ways that are hard to grok.
You might be wondering whether all these asymptomatic people testing positive are unvaxxed. Here in SF, we don't see many unvaxxed people, and...(5/16) any case, our data gurus have done breakdown (Fig). Shows that ATPR is in fact higher in people where we can't confirm vax status (presumed unvaxxed), but not THAT much higher: 6.3% vaxxed (booster status unknown), vs. 8% unvaxxed. So vax offers only mild reassurance.(6/16)
Let's turn to SF case & test + rates – the theme is the same. (We don't get to the good news until we look at hospitalizations, but even there we now see a few dark-ish clouds.)
In SF, the 7d avg=305 cases/d but last few days were in ~800-900/day, up from ~50/d last month.(7/16)
So even in highly vaxxed, still mask-y SF, we're beginning to see a tsunami of cases, mostly breakthroughs (expected, since >80% of people vaxxed). And SF's test positivity curve has a similar vertical shape (Fig), with a 7d avg of 5.2% but more like 10% for last few days.(8/16)
Today, a reporter asked what ATPR is in other cities. Answer: I don't know; I've not seen anyone else report it. But if @UCSF is ~8% in Covid-light, vax-heavy SF (w/ a ~10% overall test + rate), one can only imagine what it is in NYC, where the city's test + rate is ~20%.(9/16)
Bummed? Me too.
So let's turn to the still-good-ish news re: hospitalizations. SF still sees no uptick in hospitalizations (Fig) – just 34 in entire city (a lag; data as of 12/24). Further evidence that the average Omicron case is far milder than Delta. Thank goodness! (10/16)
As I said last wk, this can't possibly hold: with this massive a surge in cases, some bump in hospitalizations has to follow (even if Omicron is substantially milder in a highly vaxxed city). We're seeing earliest signs of it @UCSFHospitals – the first hints will be...(11/16)
... in non-ICU hospitalizations, since ICU stays typically come later. And in fact we do see a new signal: our non-ICU ("floor") admits have risen from 10-12 last week to 18 today, a subtle but I think meaningful rise.(12/16)
Breaking this down by vax status (Fig) shows hospital cases in "fully" (2 or 3 shots) vaxxed are up only a little stable; biggest uptick is in unvaxxed. Note also that virtually all the pts on ventilators for last few mths have been unvaxxed, a pattern I expect to persist.(13/16)
How to process all of this? Omicron continues to be astoundingly infectious. We now need to assume that nearly everyone might be infected... and infectious. But the evidence grows that Omi is a milder illness in most people, particularly the vaccinated. Personally, I'm...(14/16)
...reacting by being more cautious – it feels like the right thing to do, for me and my contacts. So no indoor dining, N95s when indoors except small 3x vaxxed groups. Even this: given data on infections via lingering aerosols, I now keep mask on entering empty elevators.(15/16)
But these data will also, understandably, cause a shift in thinking, since consequences of Covid are now less dire (particularly if vaxxed!) & the behaviors needed to avoid infection are growing more challenging and – as we're all seeing in friends/family – less reliable.(16/end)

• • •

Missing some Tweet in this thread? You can try to force a refresh

Keep Current with Bob Wachter

Bob Wachter Profile picture

Stay in touch and get notified when new unrolls are available from this author!

Read all threads

This Thread may be Removed Anytime!


Twitter may remove this content at anytime! Save it as PDF for later use!

Try unrolling a thread yourself!

how to unroll video
  1. Follow @ThreadReaderApp to mention us!

  2. From a Twitter thread mention us with a keyword "unroll"
@threadreaderapp unroll

Practice here first or read more on our help page!

More from @Bob_Wachter

31 Dec
A brief San Francisco update. As always, SF is a key test case since it's so highly vaxxed (81% of people w/ ≥ 2 shots) & most still follow public health recs (incl. masking). Nevertheless, we see skyrocketing cases; hospitalizations more modest but starting to tick up.(1/12)
Let's look @UCSFHospitals first. Test positivity rates continue to climb, though if you squint you might see a hint of a plateau. As always, my key # is the asymptomatic test pos. rate, now 7d avg of 8.27% – implying 1 in 12 people in SF without Covid symptoms have Covid.(2/12) Image
With that prevalence, you have to assume that everybody you come in contact with might have Covid.(Cities experiencing greater surges probably have even higher asymptomatic %.)
This is why I haven't been too critical of the CDC's 5d isolation guideline – the risk of...(3/12)
Read 12 tweets
29 Dec
My latest tweets have mostly been bad news, which saddens me, particularly during holiday season.
Today I’ll take you to my Happy Place, with some thoughts on why we could be in good shape – and maybe even great shape – in 6-8 weeks.

A 🧵(1/24)
Let’s start by agreeing that the current state is awful, and likely to get worse. Case rates are skyrocketing (Fig L), hospitalizations are going up fast (but not nearly as fast as cases; R), it feels like everybody's infected or recently exposed, & there are shortages of…(2/24) ImageImage
…key tools, including testing (both PCR & antigen) & therapies (mostly Sotrovimab [the monoclonal Ab that works vs. Omicron] & Paxlovid, the Pfizer oral antiviral with outstanding efficacy in preventing serious illness in outpatients with Covid).(3/24)…
Read 24 tweets
28 Dec
There are so many Omicron #'s floating around, it’s overwhelming. In today’s 🧵,I’ll try to simplify things with a couple of rules-of-thumb that I’ve found useful. I’ll be interested in hearing if this helps or confuses more – it might hinge on whether you’re a chess fan.(1/25)
Chess? We’ll get to that in a bit. For several of the big picture Omicron issues, my go-to is not a chess board, but rather the Number 2.
Why 2? Because many of Omicron’s key variables are either just about twice, or approximately half, of what they were with Delta.(2/25)
Two key variables for which “2” is a useful rule-of-thumb:
a) How much more infectious is Omicron than Delta? About twice. (And Delta was about twice as infectious as Alpha was.)
b) How much less serious (measured by hospitalizations) is Omicron than Delta? About half. (3/25)
Read 25 tweets
24 Dec
Covid (@UCSF) Chronicles, Day 647 (!)
I haven’t done a San Francisco update for a while. Now seems like a good time, since SF – the most vaxxed big city in the U.S. – may offer lessons for the nation about how an Omicron surge is influenced by high vax (& masking) rates.(1/25)
Bottom line is that some normally-scary numbers (surging cases & test positivity %) haven’t led to big jumps in hospitalizations. The big question: is it still too early to tell, or is this the hoped-for case–hospitalization dissociation at play? I’m going with the latter.(2/25)
Let’s start w/ SF before focusing @UCSFHospitals. Clearly SF is seeing a big surge in cases: ~250 cases per day (7-day daily average is 161; this lags due to the steep uptick), up from avg of 45/d in late Nov. (Figure). So about a 5-fold rise in recent weeks. Concerning. (3/25)
Read 27 tweets
22 Dec
This thing has rapidly become the world’s exasperating good news/bad news story. If your head isn’t spinning, you’re not paying attention.

A 🧵on my take on both the good news & the bad, with an emphasis on the things that have changed in the past few days... or minutes.(1/25)
Bad news: Omicron has exploded in the U.S., weeks before we thought it would. As recently as 2 weeks ago, most experts thought this would be a January issue, not a December one. The rapid uptick nationally, particularly in cities like NYC, Miami, & Houston, is jaw-dropping.(2/25)
Even here in SF, the nation’s most vaccinated (and likely boosted) city – and a city where masking is still the norm – we’re now seeing the familiar northward-facing curve (Figure). It’s now clear that no place in the U.S. will be spared a direct hit from Omicron.(3/25)
Read 25 tweets
21 Dec
FINALLY, real data on Omicron's severity – from So. Africa & Australia, summarized in threads from @jburnmurdoch & @andrewlilley_au:

Bottom line from these preliminary studies: looks like Omi's per case hospitalization rate ...(1/4)
... is ⬇50-70% compared w/ Delta's. But once in hospital, studies find similar risk of "severe disease" (eg, ICU admit). Impact on death rates? Too soon to say.
Findings are good news overall (equal severity would be awful), though need more data to be sure – the stakes...(2/4)
are just too high. The studies don't tell us if the lower case-hospitalization rate is due to a ⬇in the inherent virulence of Omicron, vs. the impact of immunity from vaccines or past Covid.
Your friendly reminder that we'll likely see case rates that rise by more... (3/4)
Read 4 tweets

Did Thread Reader help you today?

Support us! We are indie developers!

This site is made by just two indie developers on a laptop doing marketing, support and development! Read more about the story.

Become a Premium Member ($3/month or $30/year) and get exclusive features!

Become Premium

Too expensive? Make a small donation by buying us coffee ($5) or help with server cost ($10)

Donate via Paypal

Or Donate anonymously using crypto!


0xfe58350B80634f60Fa6Dc149a72b4DFbc17D341E copy


3ATGMxNzCUFzxpMCHL5sWSt4DVtS8UqXpi copy

Thank you for your support!

Follow Us on Twitter!