search

microCOVID / microCOVID

Estimating the COVID risk of ordinary activities
https://www.microCOVID.org
MIT License
268 stars 54 forks source link

Update to reflect BA.2 transmissibility #1433

Open sameerjain0123 opened 2 years ago

sameerjain0123 commented 2 years ago

MicroCOVID.com indicates that it was last updated on 1/4/22 for the Omicron variant, but the update section doesn't list an updates for the BA.2 subvariant or subsequent subvariants.

A 4/22/22 AMA article explains that BA.2 appears to be much more transmissible than the original Omicron variant, stating:

"Dr. Crum noted that most experts “say it’s 30% to 60% more transmissible, so it’s a risk—especially if it gets into a nonimmune population—to be able to skyrocket in terms of the number of cases."

Here's the article's bio of Dr. Crum:

"AMA member Nancy Crum, MD, an infectious disease specialist at Avita Health System in Galion, Ohio, discusses what patients need to know about the BA.2 Omicron subvariant. Dr. Crum is also an alternate delegate in the AMA House of Delegates for the Infectious Diseases Society of America."

This article is available here: https://www.ama-assn.org/delivering-care/public-health/what-doctors-wish-patients-knew-about-ba2-omicron-subvariant.

(HT: My friend Alice)

sameerjain0123 commented 2 years ago

Additional info supporting the assertion above, from a 4/26/22 article:

"BA.2 SUBVARIANTS STILL ON THE RISE U.S. health officials have estimated that BA.2 is roughly 40% more transmissible than the original omicron variant, BA.1; and that a related strain called BA.2.12.1 is about 25% more transmissible than BA.2, which would make it about 75% more contagious than BA.2. The U.S. Centers for Disease Control and Prevention in a weekly update Tuesday said the two subvariants combined for 97% of U.S. virus cases last week, up from 93% the prior week. Of sampled cases nationwide, 68% were BA.2 and 29% were BA.2.12.1, compared to 74% and 19% one week earlier, respectively. For the CDC region that comprises California, Arizona, Nevada, Hawaii and Pacific island territories, BA.2 made up 82% and the more contagious BA.2.12.1 comprised 15% last week, shifting from 85% and 9%, respectively, the previous week.

The precise prevalence of the two subvariants in California is not yet clear because the CDPH data dashboard for variants does not distinguish between BA.1, BA.2 and BA.2.12.1. All three subvariants remain lumped together as “omicron,” which the state last week said made up 99% of recent specimens sequenced for variants. One Sacramento-area testing network that does monitor for BA.2, the Healthy Davis Together project in Yolo County, reported BA.2 comprised 93% of positive cases for the week ending April 16, with the remaining 7% BA.1. BA.2 has increased from 83% on April 9 and 68% on April 2 in Yolo County. The rate of Healthy Davis Together tests returning positive increased by just over double between April 2 and April 16, according to the project website.

Healthy Davis Together data does not yet distinguish BA.2 from BA.2.12.1. Experts are still working to understand BA.2 and BA.2.12.1, especially the latter, and how they may influence factors like vaccine efficacy and immune protection from previous infection. Health officials across the U.S. have said in recent weeks that it does not appear that BA.2 causes more severe illness than the original omicron variant."

Read more at: https://www.sacbee.com/news/coronavirus/article260767907.html#storylink=cpy

sameerjain0123 commented 2 years ago

Any update on this, or heuristics on how to use MicroCOVID in the meantime? To put it bluntly, it seems like if anyone is still using MicroCOVID at all, without these adjustments it is underestimating risk to a dangerous degree.

Some ideas if it is too labor-intensive to update the actual code for the MicroCOVID tool:

  1. Write a note (e.g. in blog or whitepaper) on how you'd suggest adjusting for the latest variants (and ideally post a link to this in the updates section on the top right of the main page.
  2. If that's still too labor-intensive, at least write a note to that effect telling users that the latest variants are way more transmissible and the current numbers are a vast underestimate.

If there's anything I can do to help with this as someone who doesn't code and has some but not particularly strong ability to read papers carefully (I could probably excerpt them usefully, but would not be able to analyze their methods in the way that say, Zvi has in their LessWrong posts), let me know! If throwing money at the problem would help, let me know that too and I am happy to contribute what I can and/or help raise it.

apiology commented 1 year ago

Thanks for the write-up, @sameerjain0123! I too would welcome a PR that updates our model using updated numbers for this variant, especially if it cites additional research done since you passed this along and documents the update.

I share your fears about general usage and the model falling out of date - please see my post here for my thoughts and a request for help from the community:

If the technical aspect of this is the blocker, I can help with that, just let me know and I can pair with someone on that part of the process.