Very excited to announce that California has formed a brand partnership with a terrifying new COVID strain. The “California variant” has now received a dedicated headline in the LA Times, the San Francisco Chronicle, the New York Times, the Washington Post, and DEADLINE, BABY!!!
All of these articles mention the fact that this new variant might be more transmissible and also might be more resistant to available vaccines — which, if true, would be at least in the running for the most devastating single development in human history.
I read all the articles about this variant, completely lost my shit, and then looked into the data a little more. Now I am an expert and will calmly share the knowledge I’ve accumulated.
Let’s start with the reasons this strain might be something to worry about, followed by the reasons it might not be.
WHY THE NEW VARIANT MIGHT BE BAD
1. It’s showing up more often in genomic testing samples. On Sunday, the California Department of Public Health put out a press release reporting that the California labs doing genomic sequencing of COVID had observed a higher frequency of this one variant in their most recent data. The variant has the cute name of L452R. All of the recent articles written about L452R are going off of this one press release.
The big headline is this: in the labs’ sequencing data from November 22nd to December 3rd, only 3.8% of their samples showed the variant. But in the next batch, from December 14th to January 3rd, the variant appeared in 25.2% of the samples.
That is a huge increase, over a very short time. Even faster than the growth of the more-famous UK variant. (In a second we’ll get into why this data could be misleading and definitely shouldn’t have been published without context in so many major media outlets.)
(Update 1/19 7:02pm: A new NYT article reports that the California variant has also been detected at a higher frequency in LA. This article describes the variant differently — it’s now referred to as CAL.20C, and it’s said to have originated in California in July rather than Denmark in March. No word as to why this information conflicts from yesterday’s.)
2. Some researchers say it might be more transmissible. The California COVID variant has three mutations to the spike protein. The spike protein is obsessed with mutating. The UK variant, the South African variant, and the Brazilian variant all have spike protein mutations.
These mutations can make the spike protein “stickier,” which can make it more transmissible. For the UK variant, this definitely appears to be true — it first appeared in the UK in September and already shows up in about 60% of cases in London.
3. Some researchers have expressed concern that vaccines might be less effective against it. The Moderna and Pfizer vaccines train your body to detect and destroy the spike protein. Because of the mutations to this variant, there’s apparently a chance the vaccine will be less effective at detecting it — although this doesn’t appear to be the case for the UK variant, which has a different spike protein mutation.
Concerns about vaccine effectiveness weren’t raised directly in the California DPH press release, but Dr. Charles Chiu, one of the scientists in charge of the genomic sequencing, did a lot of press and appears to have mentioned vaccine effectiveness in every interview he did. Some researchers have previously expressed concern about L452R’s resistance to monoclonal antibodies, including in this paper, but that doesn’t mean it’s more vaccine-resistant.
WHY THE NEW VARIANT MIGHT NOT BE ANYTHING
1. The strain is not actually “new” at all. This seems important? L254R, in combination with the other mutations of this variant, was first detected in Denmark last March. It’s been in California since May. Despite this variant not being new in any way whatsoever, the word “new” was used in the headlines for both the LA Times and San Francisco Chronicle. Wow!
(Update 1/19 7:02pm: Again, new NYT reporting suggesting that the variant originated in California in July conflicts with this information.)
Researchers have detected an increased share of cases in the last few weeks, which we’ll get into in a second. But help me with this: if this variant is really more transmissible, why would it only suddenly start to show up more frequently now? After eight months?
The fact that L254R is not at all new also means that the entirety of the Pfizer and Moderna vaccine trials took place while the variant was out there in the world. If the vaccines were really less effective against it, wouldn’t that have shown up in the data? The same data that showed these vaccines are “among the best ever created”?
How can a variant that has been around from the beginning of the pandemic suddenly be more transmissible and more vaccine-resistant now? It’s the same variant! I would genuinely love someone to explain this to me.
2. The “increased share of new cases” is debatable at best. Remember how the press release showed that prevalence of the variant jumped from 3.8% to 25.2% in just a couple weeks? You wouldn’t know this from reading some of the coverage, but the sample size was very, very, very small.
Genomic sequencing is expensive and hard. They don’t do it on a lot of samples! The first batch of samples in this comparison came from 185 COVID cases, and the second batch was from 147. The number of L452R-positive samples jumped from 7 to 37.
Here, look:
That is a big increase, but this is really not a large enough data set to suggest that this new variant has suddenly grown in prevalence in California, at all. The fact that so many media outlets reported the increased percentages without noting the tiny sample size is really not good.
(Update 1/19 7:02pm: There’s no word on the sample size of the genome-sequencing study in Los Angeles. The report the NYT article refers to has not been published yet.)
The samples weren’t even randomly gathered from around the state. Look at the bottom of the graphic: there’s a list of counties, including LA County, but “more than 90% of samples are from Northern California.” Where did the sample with the variant come from? We don’t know! Except for this, from the press release: at least some of the L452R-positive samples were from a very large outbreak at Kaiser Permanente San Jose hospital. That outbreak, officials now believe, was set off by an employee running around in this air-powered inflatable Christmas tree costume that blasted the virus all over the building:
That one outbreak has led to 89 cases. How many of the samples in the new genomic sequencing data came from the Kaiser San Jose outbreak? We don’t know. But the press release says a “large number” of sequenced samples have come from Santa Clara County, where the hospital is located.
If a lot of the samples where L452R was detected came from that one outbreak, that wouldn’t really suggest that the variant is becoming more prevalent across California. It might mean that one person who was infected with the variant put on an air-powered Christmas tree costume and ran through a hospital.
3. It’s not really a “California variant.” California just does a lot more sequencing than other states, so the variant has shown up more often here. Again, L452R was first detected in Denmark ten months ago. There is still time to rename it “the Dane Strain.”
4. Some of the reporting is completely wrong. Here’s the San Francisco Chronicle reporting that the new strain “accounts for roughly a quarter of the cases in California.”
That is a hugely untrue statement that I’m positive Dr. Chiu didn’t make. The variant appeared in a quarter of 147 new cases that were 90% from Northern California. It does not account for a quarter of the cases in the state! Don’t publish stuff like that about a pandemic!
It feels kind of irresponsible, when California already has so many people declining the vaccine, to report that a new COVID variant that could be more transmissible and vaccine-resistant is exploding across the state when all of that — the newness, the transmissibility, the prevalence, and the vaccine-resistance — is either barely indicated or just wrong. The absolute last thing we need is even fewer people choosing to get the vaccine. Doesn’t this need to be a consideration for journalists right now? You don’t have to write an article about this if you aren’t sure what it means!
To borrow a phrase from all of the worst Twitter threads about COVID: I am not an epidemiologist. But nobody who wrote about this variant is either.
(Update 1/19 7:02pm: All in all, the new information in the NYT article — different from yesterday’s reporting — seems to indicates a possibly more transmissible virus that may be partially responsible for the recent surge in LA. If it originated in California in July, it’s been here for less time than previously reported, and vaccine trials would have had less exposure to it. There’s no new information available with regard to vaccine resistance.)
(Update 1/25 8:22am: At least one infectious disease expert has looked at the preprint from the Cedars-Sinai study cited in the NYT article and he is not impressed).
