This is a thread about the @stanford research group that has led the Santa Clara/Los Angeles @lapublichealth serology studies. I want to highlight the importance of content expertise and biases of personal beliefs.
Who was involved in these studies and how were they performed?
Links to pre-prints:
medrxiv.org/content/10.110…
The second seemed to be leaked on @RedState than removed.
dropbox.com/s/pk7nu9gz0vhg…
Links to pre-prints:
medrxiv.org/content/10.110…
The second seemed to be leaked on @RedState than removed.
dropbox.com/s/pk7nu9gz0vhg…
The Santa Clara study was led by corresponding author Eran Bendavid and Jay Bhattacharya from @StanfordMed .
Dr. Bendavid is an MD in infectious diseases with master's in health services research. His prior work was in global child health (not pandemic related)
Dr. Bendavid is an MD in infectious diseases with master's in health services research. His prior work was in global child health (not pandemic related)
Dr. Bhattacharya has a PhD in economics and MD with no clinical post-doc training. His research has been on health policy, medical expenditures and HSR-related work.
On March 24, both Bendavid and Bhattacharya co-authored an op-ed in the WSJ “Is the Coronavirus as Deadly as They Say?’
wsj.com/articles/is-th…”
wsj.com/articles/is-th…”
Dr. Bhattacharya has gone to remain active with interviews with @SkyNews and @TuckerCarlson as well as a number of Podcast/@YouTube interviews:
These are all strong opinion pieces meant to minimize COVID19 cost to public health and based on little known facts during a time of great uncertainty.
John Ioannides is 2nd to last author on the Santa Clara study. He is a highly cited physician-writer known for meta-research. Highlighting biases in the literature and reporting inconsistencies, and citation practices. I've enjoyed his early writings, predominately criticisms.
JPAI has had some contrarian takes recently on the call to not dichotomize p-value interpretations led by @Lester_Domes (Personally, I've disagreed with JPAI's takes on this).
More recently, he made waves early in the pandemic with this opinion piece. “A fiasco in the making? As the coronavirus pandemic takes hold, we are making decisions without reliable data”
statnews.com/2020/03/17/a-f…
statnews.com/2020/03/17/a-f…
As of April 17 he continues to make claims that COVID19 is no worse than the flu. It appears the group he leads has funded a series of videos to make these points: 3 featuring him and one of Knut Wittkowski
All these strong opinion pieces are based on little factual certainty but emphasize that the COVID19 pandemic is overblown.
The U.S. has about 45,000 deaths in less than 2 months. Hospitals are strained but none of their arguments have back-tracked.
The U.S. has about 45,000 deaths in less than 2 months. Hospitals are strained but none of their arguments have back-tracked.
For the LA County Preprint that was leaked. Neeraj Sood of @USC is the lead author. He is an economist trained at @RANDCorporation . His prior research was on health policy and U.S. health economics. He's never done pandemic-related or clinical researcher.
There is a master's graduate student in epidemiology, 4 medical students (one with an epi masters), one @lapublichealth official with a masters of public health.
There are no PhD level bio-statisticians or epidemiologist on the study. No MD with clinical trial experience.
There's some strange corporate authors. Two that seem to be related to drug testing for athletes. I believe this relates to the USC and Stanford news reports of testing professional athletes to re-open the season.
The last thing I will point out that is that lack of content expertise has led to egregious errors made in the study design and estimations of county wide prevalence. I won't go into detail, but they are best highlighted by a number content experts on Twitter links to follow:
@ChristosArgyrop highlights incorrect and unusual decision on estimating test performance and confidence intervals.
@wfithian corresponded with authors about their statistical assumptions and was concerned with the responses and errors he's already detected.
Epidemiologist @nataliexdean has a number of threads on these studies and excellent insight into appropriate methods.
I will end with this. The last month has seen a few vocal academic faculty at @Stanford make bold claims outside of their expertise. They also have gone on to steer two large studies with inappropriate methods. The media has given the microphone to the wrong scientists.
.@lapublichealth should be partnering with established epidemiologists, virology pathologists, and other content experts in survey study design.
It is dangerous to twist public health messages for personal beliefs not grounded in scientific facts. I believe the recent reports are either incompetently assembled or intentionally misconstrued to advocate for lower measurable risk.
@threadreaderapp unroll
