Logo

Stanford Center for Biomedical Informatics Research Colloquia - Shared screen with speaker view
jonathan chen
26:41
Example questions I'd like to raise by the end of the discussion:
jonathan chen
30:19
- “High” fatality rate from WHO that you kind of take down as overestimate, isn’t this just a straw man because that was known to be a crude fatality rate vs. you’re trying to estimate infection fatality rate (which is known to be a different number, but fair that you’re trying to figure out what that other number is)○ Infection rate is predictably underestimated, but could death count be mis-estimated as well? E.g., apparent Covid19 related death noted on autopsy from death back in February in Santa Clara County, before anyone even realized could test for it (and general limited availability of testing, especially early on)(largely being addressed now)
jonathan chen
38:55
Removed him and changing chat box to only panelists so such things don't go to every one
jonathan chen
39:37
- Comparing to not much worse than influenza, but is that a good thing? Influenza kills 30-40K per year even with vaccines and partial herd immunity.- Seroprevalence estimates <5% currently infected, implication is also that there remains a massive reservoir of uninfected people that contagion can spread to.- If 0.1% of US population died, that would be 300K people, doesn’t seem good, and we’re already past 80K (way more than what Ioannidis tried to early estimate at close to 10K). 80K is already more than Vietnam War, Korean War and way, way more than terrorism, gun violence, etc. but we went made drastic changes to society after 9/11 (with fair caveat that mass lockdown/shutdown is even more extreme impacts)
jonathan chen
44:30
Agreed
jonathan chen
01:10:09
Thoughts I've had, but many attendees have similar theme to discuss: - medRxiv has a banner that specifically indicates non-peer reviewed studies are for scientific discourse but should NOT be discussed with lay press/media yet. During active pandemic where timeliness of information is important, how to balance these factors?As scientists, you can be very nuanced and precise with your words and conclusions (I have observed this in some of Jay’s interviews), but then I respectively see that those with an agenda will scoop up the statement towards their own conclusions, so how to manage communication of information when you can be nuanced, but readers/viewers may propagate something you didn’t really say
Nigam Shah
01:16:41
I encourage everyone to think about:
Nigam Shah
01:16:46
(1) directional correctness of the results(2) the process for scientific debate about the methods (3) media coverage and political weaponization of findings(4) should scientific debate spill over into media and politics
Nigam Shah
01:21:18
@Ben — the directional correctness is referring to the question they asked, which is that how can be IFR this high? Given their results, it is clear that the number of infected people in > than what the case count tell us.
Nigam Shah
01:21:39
The debate is that larger by how much. That is the methods argument.