1a/ 🚩UPDATE: Last week, I posted a critique of the CDC/MMWR study: cdc.gov/mmwr/volumes/7…. One of my points (4b) highlighted the misleading claim that “9 in 10” hosp. children were unvaxed. I made an inquiry to the MMWR, and got a response. 🧵👇 Image
1b/ Bottom line – they present technically true statistics taken out of context, and produce simplified graphics that are easily amplified in social media, without scrutiny. However, we NEED to scrutinize!
2a/ The claim, centered in their “graphic” summary states that 9 in 10 children who were hospitalized with confirmed C-19 illness, is based on their review of 349 records, of which 301 were unvaccinated (86.2%). So crudely, yes – 9 of 10 children in their data set was unvaxed. Image
2b/ HOWEVER, in study period of12/19/2021 – 2/28/2022, the vast majority of 5-11yos were unvaxed. Auths claim vax rate was about 32% at END of study, but it was ~13% at start. Assuming linearity, the avg. vax rate in the study period was about 18% (about 8 in 10 were unvaxed). Image
2c/ Accounting for approximations, statistical variation etc., the 9 in 10 in the MMWR sample is virtually NO DIFFERENT than the 8-9/10 proportion of children who were unvaccinated in the community, during the study time!
2d/ The auths could have provided some comparison to the % unvaccinated in the community. However, they found it more convenient to report the crude “9 in 10” statistic, instead.
3a/ So, I inquired in an email to the author, on the context and validity of the claim - because science requires DIALOGUE: Image
3b/ And their response confirmed my critique, that “we did not correct for baseline vaccination rates in the catchment areas when examining the proportion of vaccinated vs. unvaccinated children during the early month of Omicron predominance.” (!) Image
3c/ They further write that this “9 in 10” stat is solely “DESCRIPTIVE” – a simple stat(=301/349), rather than INFERENTIAL which allows generalizations about a hypothesis. That the CDC/MMWR CHOSE to publish this misleading stat as the CENTRAL POINT of the infographic is telling Image
3d/ The message they want to promote is at the bottom of the infographic: “ Protect all eligible children by keeping their vaccinations up to date.” The misleading “9 in 10” stat was clearly placed to gain support of the unwitting on social media.
4a/ In fact, while we are on the topic, the other two infographic claims are also misleading.
4b/ They claim that “3 in 10” children hosp. with C19 had NO underlying conditions. The STAT is true over all children in the study, but its simply NOT reported in terms of vax status (Table 1). Yet, the graphic implies vax prevents hosp. in kids w/o underlying conditions. Image
4c/ The same graphic claims “2 in 10” required ICU care. Again, this is true across both unvaxed and vaxed kids. But the supplementary table shows that ICU admission in both unvaxed and vaxed were statistically similar. Image
4d/ SO, ALL 3 CLAIMS ON THEIR GRAPHIC MISLEAD into believing this study strongly supports need to vax 5-11 yos to prevent hosp. Instead, it shows relative risk reduction is modest, absolute risk reduction is minimal, and there is no identifiable difference in outcomes w/ vax.
5a/ The @cdc @CDCDirector needs to be held to a higher standard than media outlets. Data should be presented in an accurate and honest manner, regardless of how it reflects on the desired narrative. I truly hope they can change their ways, it is important for public health!
@cdc @CDCDirector 5b/ And at least, I do appreciate that the MMWR got back to my inquiry – dialogue is the first step in a healthy debate. That is a positive.
@cdc @CDCDirector 6/ Thank you!

• • •

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

Keep Current with Mahesh Shenai, MD

Mahesh Shenai, MD 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!

PDF

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 @mahesh_shenai

Apr 22
1a/ CDC/MMWR cdc.gov/mmwr/volumes/7… report just published, claiming C19/Omi hosp. in 5-11yo were 2x more likely in the unvaxed, and 9/10 hosp. were unvaxed. With MMWR, toplines & graphics are selected for a predetermined narrative. Appraisal & perspective required. 🧵👇
1b/ I will appraise the most critical, related to their claims on vaccination.
[1] That hospitalizations were 2x more likely in the UV, and
[2] 9/10 hospitalizations were in the UV.

(UV=unvaxed, V=vax)
1c/ For purposes of this thread, let us assume that the methodology is sound. In reality, there are a lot of methodological issues, such as classification errors, selection biases, and other issues. But for brevity, we will assume the numbers to be true.
Read 24 tweets
Jan 20
1/ 🚩🚨🚩cdc.gov/mmwr/volumes/7… Surprising paper in MMWR: C19 Cases and Hospitalizations in vaxed AND previous infected. Explicitly concludes “By early October, persons who survived a previous infection had LOWER case rates than persons who were vaccinated alone!” Finally! 🧵👇
2a/ Paper reports on populations in California and NY, in 1.1 M total cases. They divided the population into four groups:

No prior infection/Unvaxed (NPI/UV)
No prior infection/Vaxed (NPI/V)
Prior infection/Unvaxed (PI/UV)
Prior infection/Vaxed (PI/V)
2b/ This is a longitudinal cohort study. The “prior” C19 infxn had to take place prior to 3/1/21, and vaccination prior to 5/16/21. They then measured new C19 cases, hospitalizations weekly from May30-Nov 20. They then performed adjustments and calculated hazard ratios.
Read 31 tweets
Jan 19
assets.publishing.service.gov.uk/government/upl… UK HSA briefing, suggesting additional vax and booster doses on top of prior infection (PI) adds protection w/Omicron. The reporting on this study (and interpretation by some on Twitter) is one-sided. Here are my thoughts- 🧵👇
2a/ The UK HSA has been a leader in COVID data dissemination, particularly in the context of Omicron and variants of concern. They quickly aggregate the data and studies, and put out weekly analysis, which asks the pertinent questions. Their findings are credible.
2b/ The data comes from the “SIREN” cohort, which consists of UK health care workers, whose serological statuses are known. The study includes symptomatic AND asymptomatic infections, which is different from other VE studies, which usually look at SYMPTOMATIC illness.
Read 16 tweets
Jan 11
1a/ medrxiv.org/content/10.110… Recent pre-print by Kaiser group, on Moderna vax effectiveness in Delta & Omi. Finds VE reduced with Omi to 30%, but “boost” can improve to 63% overall, but with minimal effect in immunocompromised. Study is Moderna sponsored, and should raise ?? 🧵👇
1b/ The analysis appears well-designed and well-conducted. The results are generally not favorable for Moderna, but since its industry-funded, there is some “spin” control going on.
1c/ The results of this study are in line with other reports from Canada, UK, Denmark and Scotland. Basically, two dose mRNA vax wanes to near 0% or LESS by 6 months, but a 3rd dose can restore VE to ~40-60%, but this also wanes.
Read 31 tweets
Jan 7
1/ medrxiv.org/content/10.110… Another solid preprint study by Qatari group re: protection of PRIOR INFECTION (PI) to Omicron & other variants. Overall, PI protection reduced with Om, but extremely good vs. hosp/death, similar to vax. AND, PI+Vax does not improve protection. 🧵👇
2/ The study finds 54% protection against symptomatic infection, 87.8% against hosp/death.

PI protection vs. Omicron infxn is 56% (90% Alpha, 85% Beta, 92% Delta).

PI protection vs. Omicron hosp/death: 87.8% (69.4% Alpha, 88% Beta, 100% Delta)
3/ The protection by PI vs. Omicron infection is similar to vax protection in other studies from Canada, UK, Scotland, Denmark, and South Africa. See below.

PI and Vax+Boost remain similar in their protection against Omicron variant.
Read 11 tweets
Dec 2, 2021
1/ 🚩🚩medrxiv.org/content/10.110… New pre-print study out of UK SIREN group looking at durability of vax & prev. infxn (PI) from 3/20 - 9/21. Conclusion: Greater protection by PI vs. vax, vax wanes faster than PI, but PI may confer increased but (modest) benefit. 🧵below
2a/ This is an exceptional author group, “SIREN” , that led some of the earlier studies on NI, which found an 84% risk reduction (RR) by PI, lasting at least 7 months. Link: pubmed.ncbi.nlm.nih.gov/33844963/
2b/ This new study looks specifically at health-care workers, with known serological status, and routine asymptomatic testing, which overcomes some of the main critiques of other observational studies.
Read 25 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

Don't want to be a Premium member but still want to support us?

Make a small donation by buying us coffee ($5) or help with server cost ($10)

Donate via Paypal

Or Donate anonymously using crypto!

Ethereum

0xfe58350B80634f60Fa6Dc149a72b4DFbc17D341E copy

Bitcoin

3ATGMxNzCUFzxpMCHL5sWSt4DVtS8UqXpi copy

Thank you for your support!

Follow Us on Twitter!

:(