Thread by @Bob_Wachter

1/ Covid (@UCSF) Chronicles, Day 58

Grand Rounds day; it's here: bit.ly/2WW49bH

Three segments: a) two @UCSF ICU docs returning from NYC; b) community-based studies in SF’s Mission District & rural Bolinas; and c) a deep dive into antibody testing. Summaries below.

2/ First, the local update; all good. @ucsfhospitals: 10 pts, 5 on vents. Cont downtrend (Fig L). No new deaths.

San Francisco data terrific: new cases (top R) & hospitalizations (bottom R) falling fast. Over last 5 days, average of 14 new cases/day in SF and zero new deaths.

3/ Grand Rounds: Today, 2 of our ICU MDs, @mpeter34 & Michelle Yu, back from NYC. Most surprising thing? @ 6:15: Peters: it was more chaotic & more overwhelming than I expected (even though he expected the worst). “Overwhelming of the infrastructure… was really shocking.”

4/ Peters @ 8:00: “When I got there I realized I didn’t know the right thing to do.” (Covid is so different than usual ARDS). “They were all the sickest pts I’ve ever taken care of.” Wildly short-staffed, 1 RN for 7 pts (vs ~1:1 @ucsfhospitals). Scared they'd make bad errors.

5/ Yu @ 16:40: What really impressed me was the “local heroes”, “the janitorial staff to the mid-levels to the nurses... they were absolutely incredibly and pulled together as a team.” @ 17:10: Impressed by how helpful proning was for Covid patients with respiratory failure.

6/ Yu @ 18:25: Plenty of empiric anticoagulation – “I have lots of mixed feelings... Almost every MD had a case of intracranial hemorrhage…(on anticoagulants).” Bad side effects, but she also had healthy young pt on a blood thinner who developed a massive clot. Very tough calls.

7/ Peters @ 19:15: “I learned humility ….anyone who is very confident that they know the right thing to do in these scenarios doesn’t appreciate how complicated these decisions are.” “There was a lot of death… it’s a great way to crush your ego.”

So grateful to our NYC team.

8/ Diane Havlir on the Mission & Bolinas studies: @ 28:20: The good news is that many people with Covid don’t have severe course. So they studied 2 communities to better understand the state of the epidemic and build strategies to control it.

9/ Havlir @ 31:00: "Reach, Test, and Respond" strategy. The key was a great partnership. “This study was done for the community, with the community.” Recruited & trained 450 medical and community volunteers, close collaboration with @SF_DPH, Mayor @LondonBreed, & local community.

10/ @Jon_Jacobo, a Mission-based activist, led the community task force. @ 34:00: We knew Covid would impact Latino community, but didn’t know how much. “Many of us in the community could tell there was something going on.” Had to overcome tendency to distrust large institutions.

11/ @ 35:50: “What I can tell you is that there was a true partnership with @UCSF. Always diverting to us as the experts in our realm, community & the engagement side of it.” Particularly important since people were fearful about being tested (re: privacy, undocumented concerns).

12/ @ 39:00: In 4 days, the team tested 4,160 people (29% of SF’s overall testing). @ 40:30, sobering results: 2.1% PCR positive overall. Of those who worked in Mission neighborhood, 6.1% positive. Among PCR+, 90% said they could not work at home (vs. 57% of overall population).

13/ @ 42:00: Tested 981 Caucasians living in same neighborhood: ZERO were positive. 44% of community was Latino, yet 95% of positives were Latino. Most positives in low-wage workers unable to shelter-in-place. Common workplaces: food & beverage industries, janitorial services

14/ @ 43:20: Now-familiar finding: 53% of people who tested positive had no symptoms (not clear what fraction had symptoms later). When people tested positive, they were offered a “community wellness” intervention: PCP access, care packages for quarantine, education. Impressive.

15/ @ 52:10: @bgreenhouse1 described companion Bolinas study, testing entire rural town an hr north of SF. 1,847 people tested over 4 days. 13% believed they had Covid (~10-20% had symptoms). Despite this, ZERO positives in this town; vivid disparity with Mission District.

16/ Jacobo @ 58:45: “was a bit of shock” on seeing Mission results. “Poor folks are…@ higher risk than those who have privilege to stay home.” Policy response: SF making tests available to essential workers. And new policy to provide $ s to people w/ Covid so they can stay home.

17/ Havlir @ 1:02:45: Lessons from the HIV epidemic in SF: Having a partnership between the community, academics, & public health organizations is critical to making progress in Covid. "It's the most effective and efficient way to combat an epidemic.”

18/ Last segment on antibodies. Michael Busch, @VitalantNorCal @ 1:09:30: He presents some early results on the time course for antibodies to turn positive. @ 1:11:45: he and others launching a large national study of antibody prevalence in blood donors. Will be very instructive.

19/ @ 1:15:15: How to explain high antibody % in controversial @Stanford-Santa Clara study (since Bay Area data shows v. low %): “I believe it was all false positive results. They were using a rapid test with a 1.5% false positive rate, and they detected a 1.5% reactivity rate.”

20/ @ 1:33:00: Given problems w/ antibody tests, wise to consider combo approach: point of care test, then follow-up w/ a confirmatory test. Per Greenhouse, in their antibody tests in zero-prevalence Bolinas, they'll do a confirmatory test in follow-up to any positive result.

21/ 1:16:50: Busch: “There’s never been a transfusion transmission case of any respiratory virus.” That's a relief. No concern re: blood supply being safe, and @US_FDA recommends against viral screening of blood (though he notes @Stanford blood bank is doing it).

22/ @ 1:20:00: Alex Marson @MarsonLab describes motivation to study antibody test performance. Notes that results are generally reported as pos or neg, but (especially w/ colorimetric output) are often more ambiguous. Note (on R) sensitivity goes up as time after Covid goes by.

23/ @ 1:25:05: Marson: High specificity is key – need to try to avoid false positives, particularly in testing low-prevalence populations. Question of whether antibodies are protective: Answer: no way to tell from lab, will require following antibody positive patients over time.

24/ 1:39:20: I asked Busch & Marson if they'd predict that people will be immune after Covid. Answer: yes, likely protected from re-infection for 1-2 yrs. “Immunity will wane,” & need to be boosted w/ vaccine. Reinfections are possible, but they will tend to be milder infections.

25/ That’s it. If you can spare 100 min (great weekend viewing!) it's worth watching the whole thing. Once again, you can find it here: bit.ly/2WW49bH

Back tomorrow with tweets. Next Grand Rounds next Thursday. Stay well...