1) Here's a quick example of how the federal government is censoring the best scientific research. It's not just cuts to ongoing research.
It's new grant submissions too...
2) In January, I re-submitted a promising Covid/cancer grant to a non-federal funder. Hundreds of pages. Hundreds of hours of work. The best proposal I've submitted as a scientist.
Out of curiosity, I used Sean Mullen's Scan Assist tool to see how many banned words it had...
3) The proposal had 1,750 banned words. No big deal -- they're non-federal.
BUT I had planned to submit a smaller version to NIH this month as a "back up." Impossible!
It's not a matter of using a thesaurus or the find/replace command. The grant is on *Covid*...
4) Covid is a banned word and -- because grants will be checked with natural language models -- a de facto banned *topic*.
Nobody in their right mind would submit a grant requiring tens or hundreds of hours on a work on a banned topic because...
5) Even if somehow it snuck through and got funded, it could get defunded at any moment when the topic became apparent.
Anything Covid, health disparities, environmental, etc. is DOA from a federal funding perspective.
BUT I wanted to submit something this cycle...
6) So I brainstormed what I could submit for the June 2025 deadlines. The least controversial idea was basically #17 on my list.
So, instead of the federal government supporting the best science or the 2nd best science or 3rd best science...
7) The federal government has banned so many words *and topics* that they are basically left with the 17th best science.
So, I wrote my proposal. It's cool stuff, old stuff I've been working on for the past 18 years, since a PhD student.
I'd rather work on the best science...
8) or the 2nd best science, etc. But it's a good study and would help a lot of people.
Each draft of each document, I used Scan Assist to verify that I didn't have any banned words.
9) but the funny thing was that even upon researching the least controversial 17th best idea on my list, banned words kept popping up.
A lot of these words could be used in a controversial context, but generally are not controversial. A few examples are...
10) "promote" (like promote better care), "barrier" (like overcome barriers to better care), "political" (like someone having an undergrad poli sci major 30 years ago on their biosketch), "bias" (like statistical measurement error), "continuum" (like levels of symptoms)...
11) It was really striking how many words pop up even in a non-controversial application.
My goal was to submit an application with ZERO banned words...
12) And today I submitted a federal grant application with ZERO banned words.
BUT this process censors topics (covid, health disparities, environment, LGBTQ health, etc.).
AND many apps will not be funded because they accidentally include banned words...
13) You can find Dr. Sean Mullen's Scan Assist tool here:
The PMC website includes an international directory of websites with COVlD wastewater monitoring. It is more up to date than the directories of the EU and WHO.
Let's review what's happening in Europe...
1/
Data in #Austria show a rising COVlD wave. The x axis (bottom) has infrequent labels, but the data shown go through October 8th.
An estimated 1 in 81 people are actively infectious during the ongoing 11th wave.
The "shutdown" has created a blackout at the state level.
Transmission is half that of the peak one month ago, and we anticipate a relative national "lull" in early-to-mid November, albeit still at dangerous levels.
Our model uses a combination of CDC and Biobot data, so we are able to estimate national statistics despite the CDC data going offline. On the map, note that Puerto Rico continues to update; they use a CDC-style system but were dropped by the CDC long ago. For full methodology, review the technical appendix on the website.
PMC COVlD Update, Week of Oct 6, 2025 (U.S.)
🧵2/9
#DuringCOVID is today. We estimate >600,000 new daily infections. This is about half the peak on September 6.
Notice current levels are similar to the estimated peaks of the first 3 waves.
PMC COVlD Update, Week of Oct 6, 2025 (U.S.)
🧵3/9
Weekly estimates:
🔹4.5 million infections
🔹>200,000 resulting long-term health conditions
🔹>1,300 resulting excess deaths
Key points in my letter to the pharmacy boards. 🧵1/7
Georgia law indicates that the pharmacy board is to follow ACIP. They do not dictate further nuance. Georgia continues to require prescriptions, going against the spirit of the law, ACIP, and 47 other states.
🧵2/7
Louisiana law tells the pharmacy board to follow ACIP. ACIP says do not require a prescription, and 47 other states agree.
The Louisiana pharmacy board continues to require a prescription.
On the back end of this unprecedented 11th wave, "times they are a-changin." In particular, COVlD levels are shifting north and east.
Notice that few states are in the highest and lowest categories. Much of the south and west have considerable transmission post-peak. Many places in the north and east are seeing steady or increasing transmission after relatively lower levels.
Overall, levels are lower than the past few weeks, but transmission remains considerable. Those relying on anecdata (friends, coworkers, and family infected) may increasingly realize we are in a wave.
We estimate nearly 750,000 new daily infections nationwide, meaning approximately 1 in 66 people or 1.5% are actively infectious.
These estimates are derived by linking wastewater levels to IHME true case estimates using methodology commonly employed worldwide, detailed on the website, noted in a pre-print. Many publications in leading medical journals link wastewater data to key metrics that matter, noted in the online technical appendix.
In this week's report, we note adding North Dakota and Puerto Rico to the heat map in support of health equity. We have been imputing ND levels since the launch of PMC 3.0 using data from neighboring states. PR continues to report qualitative levels using the CDC format but is not longer included on the CDC website.
1/8 🧵
PMC COVlD Update, Sep 29, 2025 (U.S.)
State-level prevalence estimates, AL to MS. The levels use CDC labels, which tend to have an optimistic portrayal of risk. For example, CO is listed at "low" (by our estimate 1.5% infectious).
#MaskUp at 1.5% if having lapsed.
2/8 🧵
PMC COVlD Update, Sep 29, 2025 (U.S.)
State-level prevalence estimates, Missouri to Wyoming.
New York had *huge* retroactive upward corrections, and is now "High," as many residents hypothesized.
Note, Puerto Rico only provides CDC qualitative levels, so no data.