We can now track daily TaqPath tests from Yale New Haven Hospital, and of the pos tests with ORF target < 30 CT, report the fraction likely Omicron (SGTF) and Delta (non-SGTF).
From tests on Dec 13, *23/214 (9.7%)* had a SGTF
Explanation & caveats in 🧵
Our data represent a fraction of the samples tested by YNHH, and the catchment is primarily New Haven and Fairfield Counties. So they don’t represent all of Connecticut. I’ve heard from others tracking SGTFs elsewhere in the state that they are not seeing a high % yet. (2/11)
Not all non-SGTFs are Delta and not all SGTFs are Omicron (see BA.2). But of the 184 sequences on GISAID from the US collected after 11/25 with the spike 69/70 del that causes SGTF:
We haven’t sequenced all of our SGTFs, but so far they’ve all turned out to be Omicron. Some probably wont be though.
Also, our daily denominators (100-200 +samples w/ CTs <30) are fairly small and individual data points can be noisy.
(4/11)
I expect that Omicron frequencies are very different by location, and what we are seeing in parts of Connecticut is probably behind the growth in NYC and other big cities, but probably ahead of many other parts of the country not as connected to major travel/urban hubs. (5/11)
Interestingly, our trends are similar to what @pavitrarc reported from @UWVirology - <1% through early December then a rapid increase starting around Dec 6th. (6/11)
And the trends that we are seeing with the YNHH Connecticut data are similar to the SGTF % from Denmark. It looks like we are about 1 week behind their rates, which seems about right. (8/11)
Keeping the caveats above in mind, I expect SGTFs (mostly Omicron) from the YNHH testing catchment area (mostly New Haven & Fairfield Counties, CT) to surpass 50% within 1-2 weeks.
That’s *if* the current growth rate at 0.26/day (0.1-0.3) remains the same. (9/11)
The rapid doubling time of ~3.8 days estimated from the YNHH SGTF data above is consistent with early estimates from Europe, but this may decrease as explained by @richardneher. (10/11)
I am hoping that by presenting these early findings that individuals and public health agencies will start to take more precautions. Lets not put more stress on our already over-stressed healthcare system when this should be a time to enjoy our friends and family. (11/11)
• • •
Missing some Tweet in this thread? You can try to
force a refresh
S-gene target failure (SGTF) data (a proxy for omicron) from yesterday took a big leap from 12% to 30%, which may end up being a bit of an outlier (noise w/ small n). Regardless, the % keeps rising fast, and we predict that Omicron will >50% in this population within days. (2/5)
The above data does not represent all of Connecticut – only parts of Fairfield and New Haven Counties. Other areas may be lagging behind, though at this rate Omicron could reach dominance (>50%) across the state by the end of the month. (3/5)
(2/6) The CDC N1 probe mismatch in Omicron is a C28311T mutation. Of the 91 B.1.1.529 sequences on GISAID:
- 48 have a T (mismatch)
- 19 have a C (exact match)
- 26 have no coverage (NNNs)
(3/6) All of the Omicron sequences with a C at 28,311 (exact match to N1 probe) are from the CERI-KRISP group, so maybe @Tuliodna can provide some info if these are potentially a sequencing artifact (which we've seen with Delta) or if this site is variable within the lineage
Like Alpha (B.1.1.7), B.1.1.529 has the spike 69-70 deletion that causes a S-gene target failure (SGTF) when using the ThermoFisher TaqPath COVID-19 assay. Because most other variants didn't have the deletion, this was helpful for tracking Alpha (1/8)
Likewise, reporting SGTFs will be extremely helpful again for tracking the spread & growth of B.1.1.529, like the amazing folks in South Africa are already demonstrating (2/8)
While the ThermoFisher TaqPath COVID-19 assay is very common, not everyone is using it; and while sequencing is required to confirm B.1.1.529 cases, slow turn-around times can delay its tracking. Here is where other PCR screening assays can be helpful (3/8)
The Delta sublineage leaderboard remained virtually unchanged this week (boring is good), though things could get very interesting through the holidays. (2/4)
Unfortunately, we are going into the holidays with transmission and infections estimated to be rising in Connecticut (and across the US). (3/4)
Much of the discussion is still about AY.4.2, which is still slowly 📈 in the UK, which they have designated as a 'variant under investigation' (VUI). (2/6)