One in a million will not be the final estimate for the unconfirmed J&J events and I think we should be cautious in citing it. The incidence will surely be rare, but @C_R_Watson and I wrote about biases in estimating mortality last Feb, and similar issues are in play here 1/
@C_R_Watson First, the numerator. Now that awareness has been raised, more cases may be reported. 2/…
Second, the denominator. All of the cases in US have been in women ages 18-48. It is reasonable to ask whether the denominator should reflect that. The expert committee will consider this. 3/
Third, time. FDA said it was 1-2 weeks to onset. People who have received doses within that time would not normally be considered in the denominator either. On Mar 26 there were 2.7M J&J doses administered. (Not exactly 2 wks but closest date I could find.) 4/
Again, these events are under investigation and we will learn more from the experts. If confirmed, the events will still be rare. But one in a million is unlikely to be the final tally. 5/5

• • •

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

Keep Current with Caitlin Rivers, PhD

Caitlin Rivers, PhD 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 @cmyeaton

29 Mar
Cases are on the rise in many states. How worried am I? I am concerned, but not as much as I would have been 6 months ago with these trends. Short thread. 1/
Many states are doing quite well. I would like to see cases fall below 20 cases per 100,000 per day as a first goal and 10 as a second goal (and then the lower the better). By that measure, we are doing ok: 35 states are at or below 20 and 12 are below 10. 2/
Yet some states are resurging. MI and NJ are at ~50 cases per 100,000 per day and hosp rising too. Risk in those states is high, and leaders should intervene by closing high risk settings and accelerating vaccination. Fed govt could help by sending extra vax coming online. 3/
Read 7 tweets
4 Feb
So, how are we doing with covid? Nationally, reported incidence has fallen from around 76 cases per 100,000 population per day to about 43. Better! But not yet good. Thresholds are contentious, but I think 20 as a first goal and 10 as a second goal are reasonable to start. 1/
Nationally, those thresholds correspond to approx. 66,000 and 33,000 daily cases, respectively. Right now, we’re at ~140,000 cases reported each day on average, so we need to more than halve incidence to get to Goal 1. 2/
What do 20 and 10 cases per 100,000 per day get you, practically? We will still be masking and distancing, for sure. In VT, where reported incidence is currently about 21, in a group of 10 people there is about a 10% chance someone has covid (caveat ahead) 3/
Read 7 tweets
1 Feb
Fantastic development. This test is already authorized for home use without a prescription, including in children and people without symptoms. Increasing supply critical to allow people to easily access tests.
The test was authorized in Dec but supplies are limited. The Biden Admin is investing in expanding manufacturing capacity. It will take a while (months, probably) for that to result in more supply available to consumers, but we'll need tests for a long time so still useful.
The ~$30 price point is a little high to keep a few in the bathroom cabinet, but it will be a nice option. Looks like the Federal government will keep some of the supply, will be interesting to see how they are deployed.…
Read 4 tweets
21 Jan
I could not be more thrilled that the Biden-Harris Administration has announced an intention to create a National Center for Epidemic Forecasting and Outbreak Analytics @dylanbgeorge 1/…
@dylanbgeorge This idea has a long history, with dozens of brilliant scientists, policymakers and innovators behind the wheel, including many who roll up their sleeves every day to make sure our leaders have what they need to fight outbreaks effectively @dylanbgeorge 2/…
But they shouldn’t have to self-organize or do this on a volunteer basis. We need a standing capability that can advance the science by creating and improve epidemic models and outbreak analytics… @dylanbgeorge 3/…
Read 6 tweets
18 Jan
There are around 30 big counties, with a combined population of over 12 million, where upwards of 40% of the population has been infected. It's stunning how hard the pandemic has hit part of the US.

Estimates from…
% total infected is a modeled number that tries to estimate missed cases. You can read more here:… @youyanggu
Many of these counties have large institutions like prisons and jails, universities, and manufacturing or food production facilities. The NYT keeps a useful list of outbreaks in group settings.…
Read 4 tweets
4 Sep 20
I’m volunteering w/ the Covid-19 Symptom Data Challenge. Colleagues at CMU and UMD have been surveying people (e.g. on Facebook) whether they or people they know have covid-like symptoms. The goal is to explore how this data can be useful for tracking the virus. 1/
The CMU team has some analyses exploring uses for the data. They found, for example, an association between covid-like illness and covid incidence at the state level in July. 2/ Image
This seems promising at the county level as well, but there is a lot more exploration of the data to be done. What else can we do with this data, and how can it be useful to inform public health? That’s what the challenge is about – data discovery for public health. 3/
Read 5 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 Become our Patreon

Thank you for your support!

Follow Us on Twitter!