An evolutionary analysis based on the genome sequences of coronavirus samples taken from patients in the Seattle area suggests that the number of infections doubles roughly every six days, which translates into hundreds of infections over the course of the past six weeks.
So far, 18 cases have been confirmed in Western Washington, including 14 in King County and four in Snohomish County, north of Seattle. As of today, five patients have died — four in King County and one in Snohomish County.
But the analysis laid out in a series of tweets from Trevor Bedford, a researcher at Seattle's Fred Hutchinson Cancer Research Center who specializes in the study of viral dynamics, concludes that many more people are likely to have picked up the infection from the first patient in the U.S. to be diagnosed with the respiratory disease caused by the virus. Some probably passed along the virus even if they didn't know they were infected — a phenomenon known as "cryptic transmission."
Depending on how the computer modeling is tweaked, as many as 1,500 people may have picked up the virus through the transmission chain that began with the patient known as WA1, who traveled from the Chinese city of Wuhan to Snohomish County in mid-January.
"There will be more in the whole state," Bedford wrote. He said he suspected that the Seattle area's current coronavirus situation is similar to what the situation was in Wuhan around Jan. 1, when the spread of the infection was beginning to pick up steam. "Three weeks later, Wuhan had thousands of infections and was put in large-scale lockdown," Bedford wrote today in a blog post that supplemented his tweets.
Bedford's conclusions are based on a close comparison of viral genome sequences from WA1 and another Snohomish County patient known as WA2, leading to an assessment of where they fit on the broader evolutionary tree for the virus.
The two sequences are similar, but patterns of variation in the genetic code can indicate how much that code has changed in the course of transmission.
The virus from WA1 was sampled on Jan. 19, and the virus from WA2 was sampled on Feb. 28, The viruses' genetic codes were sequenced by the research team behind the Seattle Flu Study and shared publicly to the worldwide GISAID database for pathogenic viruses. That allowed Bedford to reconstruct how the coronavirus evolutionary tree spread out over the course of those six weeks.
In today's tweetstorm, Bedford said WA1's case appears to have been the start of a transmission chain leading to WA2. "This suggests that the case WA1 infected someone who was missed by surveillance due to mild symptoms, and a transmission chain was initiated at this point" in mid-January, he wrote.
The transmission chain that went through WA2 wasn't picked up, probably due to the fact that until last week, monitoring efforts were focused only on sick people who were traveling directly from China or who were known to be in direct contact with a known case.
"This lack of testing was a critical error, and allowed an outbreak in Snohomish County and surroundings to grow to a sizable problem before it was even detected," Bedford wrote in today's blog post.
Bedford emphasized that his analysis, conducted in partnership with epidemiologist Mike Famulare of the Institute for Disease Modeling, was still preliminary. We've reached out to Fred Hutch for more information about the analysis.
The preliminary conclusions emphasize the importance of taking steps to reduce the spread of the virus: washing hands often, making an effort to avoid touching your face, staying home if you're sick, and avoiding close contact with sick people.
Here's today's full series of tweets from Bedford:
A follow up to Saturday's analysis of undetected #COVID19 transmission in Washington State. https://t.co/8wWxtiotE3 1/11
— Trevor Bedford (@trvrb) March 2, 2020
As before, the genomic data support a linkage between the case WA1 from Jan 19 and the case WA2 from Feb 24, both from Snohomish County. 2/11
— Trevor Bedford (@trvrb) March 2, 2020
This suggests that the case WA1 infected someone who was missed by surveillance due to mild symptoms and a transmission chain was initiated at this point in mid-Jan. 3/11
— Trevor Bedford (@trvrb) March 2, 2020
Because the case definition from mid-Jan until Feb 27 required direct travel from China, any of these subsequent cases in the transmission chain would not have been tested for #SARSCoV2. 4/11
— Trevor Bedford (@trvrb) March 2, 2020
Assuming an introduction into Snohomish County on Jan 15, Mike Famulare @IDMOD_ORG simulated onward transmission using consensus epidemiological parameters curated by the @MIDAS_Network (https://t.co/p3MwdZjm6e). 5/11
— Trevor Bedford (@trvrb) March 2, 2020
Running these simulations forward with a mean doubling time of 6.1 days, we estimate the number of current infections in this transmission chain on March 1 to be 570 with an 90% uncertainty interval of between 80 and 1500. 6/11 pic.twitter.com/Vm99N3x61s
— Trevor Bedford (@trvrb) March 2, 2020
In addition to this "bottom-up" approach, Mike used a "top-down" approach in which we accounted for the single positive screening result out of the number of samples screened to date along with census information. 7/11
— Trevor Bedford (@trvrb) March 2, 2020
This approach leads to a similar estimate of the number of current infections at 330 with a 90% uncertainty interval of between 20 and 1500 infections. 8/11
— Trevor Bedford (@trvrb) March 2, 2020
We believe this particular transmission chain will have a foci in Snohomish County. We're working as fast as possible to understand extent of spread in the greater Seattle area. 9/11
— Trevor Bedford (@trvrb) March 2, 2020
Please consider this to be a preliminary analysis. We are actively working on this and will continue to update our numbers. 10/11
— Trevor Bedford (@trvrb) March 2, 2020
We are hugely thankful to @SnoHD, @KCPubHealth and @WADeptHealth for their support and clear thinking for state and local public health. They are doing a fantastic job. 11/11
— Trevor Bedford (@trvrb) March 2, 2020
via https://ift.tt/2PHJy8c
No comments:
Post a Comment