Could somebody explain these results from 51,000 employees at the Cleveland Clinic? They followed employees after the bivalent vaccine came out. Figure 2 compares the cumulative incidence of covid by the number of vaccine doses previously received.
The lowest incidence was with zero doses, and the incidence monotonically increased with the number of doses.
DB2
From Wikipedia, the free encyclopedia
Jump to navigationJump to search
medrXiv 
Type of site Distribution of preliminary medical research
Available in English
Owner Cold Spring Harbor Laboratory
URL medrxiv.org
Commercial No
Launched June 2019; 3 years ago
Current status Online
medRxiv (pronounced “med-archive”) is an Internet site distributing unpublished eprints about health sciences.[1][2][3][4] It distributes complete but unpublished manuscripts in the areas of medicine, clinical research, and related health sciences without charge to the reader. Such manuscripts have yet to undergo peer review and the site notes that preliminary status and that the manuscripts should not be considered for clinical application, nor relied upon for news reporting as established information.[5]
In January 2022, there were over 10,000 preprints released on medRxiv, which is a 50% increase compared to January 2020.
The site was founded in 2019 by John Inglis and Richard Sever of Cold Spring Harbor Laboratory (CSHL), Theodora Bloom and Claire Rawlinson of BMJ (the medical publisher), and Joseph Ross and Harlan Krumholz of Yale University. The server is owned and operated by CSHL.
medRxiv, and its sister site, bioRxiv, have been major sources for the dissemination of research on COVID-19.[6][7]
Since February, 2020 medRxiv indexed in PubMed.[8]
bioRxiv (pronounced “bio-archive”[1][2]) is an open access preprint repository for the biological sciences co-founded by John Inglis and Richard Sever in November 2013.[3][4] It is hosted by the Cold Spring Harbor Laboratory (CSHL).[5]
As preprints, papers hosted on bioRxiv are not peer-reviewed, but undergo basic screening and checked against plagiarism. However, peer reviews from other sources may be posted alongside preprints. Moreover, readers may post comments.
It has been measured that two thirds of the papers posted in bioRxiv are later published in peer-reviewed journals.[6] A service called Rxivist parses the metadata of preprints from bioRxiv, and combines it with data from Twitter allowing to detect trending preprints.[7]
MedRxiv, and its sister site, bioRxiv, have been major sources for the dissemination of COVID-19 research.[8][9]
In mid-December, the CDC put out new data from two studies that suggest the bivalent boosters offer significant protection against illness and hospitalization. In the first study, 798 patients ages 65 and up received one dose of a bivalent booster in addition to at least two doses of the original, monovalent vaccines. Those who received the bivalent vaccine were 84% less likely to be hospitalized for COVID-19 than those who were unvaccinated and 73% less likely to be hospitalized than those who received two or more doses of the monovalent vaccine.
This study was conducted from Sept. 8 through Nov. 30, during a period of Omicron subvariant BA.5 or BQ.1.1 predominance.
In the second study, adults 18 and older who received a bivalent booster were 57% less likely to seek care at an emergency department or urgent care clinic compared to those who were unvaccinated. And those with the bivalent booster were 38% less likely to seek such care than those who received monovalent vaccination only with the last dose five to seven months earlier and 45% less likely than those who had monovalent vaccination only with the last dose 11 months earlier or more.
The study took place from Sept. 13 through Nov. 18, when Omicron subvariant BA.5 predominated and additional Omicron sublineages emerged.
The information from both studies came from comparisons of databases the CDC uses to track real-world effectiveness of vaccines. Neither were randomized controlled trials. Because the original monovalent booster is no longer available, the studies could not compare the bivalent and monovalent vaccines head-to-head during the same period of time.
This Yale comment is telling us why to your question.
Meaning the doses on your charts fig 1 and fig 2 were different weeks for each dose. Different periods of time. Currently the other vaccines are not even on the market or used in practice.
The reason three doses or more would see a greater number of infections is the contagiousness of Omicron.
What the article actually says:
Results Among 51011 employees, 20689 (41%) had had a previous documented episode of COVID-19, and 42064 (83%) had received at least two doses of a COVID-19 vaccine. COVID-19 occurred in 2452 (5%) during the study. Risk of COVID-19 increased with time since the most recent prior COVID-19 episode and with the number of vaccine doses previously received. In multivariable analysis, the bivalent vaccinated state was independently associated with lower risk of COVID-19 (HR, .70; 95% C.I., .61-.80), leading to an estimated vaccine effectiveness (VE) of 30% (95% CI, 20-39%). Compared to last exposure to SARS-CoV-2 within 90 days, last exposure 6-9 months previously was associated with twice the risk of COVID-19, and last exposure 9-12 months previously with 3.5 times the risk.
Conclusions The bivalent COVID-19 vaccine given to working-aged adults afforded modest protection overall against COVID-19, while the virus strains dominant in the community were those represented in the vaccine.
Summary Among 51011 working-aged Cleveland Clinic employees, the bivalent COVID-19 vaccine booster was 30% effective in preventing infection, during the time when the virus strains dominant in the community were represented in the vaccine.
So, you have better resistance to being infected at all, vs having had covid,
Steve
All of the subjects working during the 90-day period were working with the omicron in circulation.
Figure 1 shows that the more recent a prior infection was, the lower the incidence of infection during the 90-day study. That makes sense to me.
However, Figure 2 doesn’t make sense to me. The authors write:
“The risk of COVID-19 also varied by the number of COVID-19 vaccine doses previously received. The higher the number of vaccines previously received, the higher the risk of contracting COVID-19”
This seems quite counterintuitive. Why would the incidence of covid during the study amongst those who had three shots be 3-4x higher than with subjects who had zero shots?
DB2
Still puzzled
Before you spend too much time a-wondering, remind yourself that this is a pre print. Basically a test by the authors to see if it’s likedly to pass muster if/when presented to an actual journal with editorial oversight (the pay-to-publish /predatory journals don’t count)…or an effort to get eyeballs on their publication as quick as possible.
The Covid era has seen a big rise in shoddy research reaching publication level … and consequently the general public via press release…only to be retracted in near record time (you rarely hear about that unless you follow RetractionWatch)
Understood, although at the same time pre-prints are quite common in this covid-era. As Leap noted, two-thirds of the papers posted in bioRxiv were later published in peer-reviewed journals. I also note (justified or not) that all of the authors are researchers at the Cleveland Clinic, not Wossamotta U.
DB2
​​If you want to understand what the authors concluded, it’s easiest to look at their stated conclusions.
Conclusions
The bivalent COVID-19 vaccine given to working-aged adults afforded modest protection overall against COVID-19, while the virus strains dominant in the community were those represented in the vaccine.
Summary
Among 51011 working-aged Cleveland Clinic employees, the bivalent COVID-19 vaccine booster was 30% effective in preventing infection, during the time when the virus strains dominant in the community were represented in the vaccine.
Trying to come to your own conclusions by looking at the raw data is somewhere between very hard and impossible. That’s because you need to make adjustments for various risk factors. Even if you know how, they haven’t provided the underlying raw data you need to make the adjustments.
This is not a blind study where participants are randomly assigned to get various numbers of doses. Individuals who are at higher risk of contracting COVID are the ones you would expect to receive a higher number of doses. If you are older, overweight, work providing frontline medical care, or for any other reason are at higher risk, you are more likely to have had more shots. Part of this is self-selection. Higher risk individuals will tend to want their next dose as soon as it’s available to them. Part of this is rationing. When vaccines first became available, they were offered to higher risks individuals first, with eligibility based on risk level decreasing over time. I’m sure there are other reasons to expect non-randomness in vaccine count.
Because the data is clearly non-blinded and likely non-random, you need to make adjustments to the data. To do so, you need access to the full database which is not provided so we can’t do it. They have access to it and provided their conclusions. Those conclusions don’t contradict my approximate world view that vaccines decrease COVID risk, so I wouldn’t look further.
I admit that if I were looking for holes in the study, the one you point out would pop out at me. But for the reasons I mentioned, it would just be a yellow flag. I would wait for it to be published and see what their fellow scientists had to say. Peer-review is nice because people who are smarter than me look at their methodology and try to poke holes in it. But that hasn’t happened yet.
That probably is not what the method was, stated or meant to state.
snippet on the method
Methods Employees of Cleveland Clinic in employment on the day the bivalent COVID-19 vaccine first became available to employees, were included. (my comment as in Omicron was part of the group) The cumulative incidence of COVID-19 was examined over the following weeks. Protection provided by vaccination (analyzed as a time-dependent covariate) was evaluated using Cox proportional hazards regression. The analysis was adjusted for the pandemic phase when the last prior COVID-19 episode occurred, and the number of prior vaccine doses received.
Veenee,
Your point is very well said, the heart of it being Omicron is much more contagious. The maths do not work for this method of comparison.
DB2,
The conclusion is based on an apples to oranges comparison. It does not hold up. Any scientist will come up with conclusions that wont hold up. Writing them down does not change that.
Dr Bob,
30% effective means 70% ineffective which is why they said it has modest protection. Apparently even if you have two or three vaccinations, the protection is not like the old vaccines where if you got the smallpox vaccine you would never get it. This vaccine is temporary and there are numerous published studies saying that.
Oh, I am Cleveland Clinic trained btw…doc
That needs some qualifications such as fewer or lesser symptoms with the vaccine, Covid is less risky to seniors with the vaccine and long Covid may be less frequent with the vaccines.
True…but irrelevant.
The thing to remember (or be aware of if you didn’t know in the first place) …publication is not validation Working backwards, even papers from august research institutions and published in high impact journals can’t be assumed to be “settled science”. Indeed, a good many of these fail the reproducibility test for a whole variety of reasons. So, allowing that a claim of a given percentage of preprints eventually make it into a peer reviewed publication is accurate, even that doesn’t validate a pre print in any way.
As noted upstream, there isn’t enough raw data in this little “taster” for even someone with expertise in the area to make an informed opinion so, like I said, not worth fretting over … unless you want to,of course.
There’s a reason for that … it’s the virus, not the vaccine.
If variola had shared the same characteristics as SARS CoV2, the “old vaccine” wouldn’t have eradicated Smallpox
Not me, although I do confess a weakness for counter-intuitive results.
DB2
Indeed…it all depends on what “effective” is supposed to mean, right? Without that caveat doesn’t seem to me that 30% effective (or 70% ineffective, depending upon what floats yer boat) means very much.
For sure, the vaccine doesn’t create a sterilising immunity…but neither does coming down with Covid. Unlike, say, Smallpox, both vaccine and disease.
This was discussed pretty extensively in the early days of vaccine rollout on TWiV. I don’t think anyone but the ill informed hang onto the notion that a vaccine for this virus should be considered ineffective just because there’s not a guaranteed elimination of the virus/disease.
As many people also seem to have.
However, if you feel drawn to counter intuitive results for the counter intuitive appeal, it’s also good to remember Carl Sagan’s observation that extraordinary claims require extraordinary evidence.
In this instance, you basically only have the authors’ suggestion that the results of their research are counterintuitive minus a detailed analysis of the results themselves or the usual oversight from peer review to confirm that, right. No way to tell for sure if they really have stumbled across something new and newsworthy…or are just plain wrong.
And one-third were not published. How do you tell one from the other before they are peer-reviewed?
Hint - you probably can’t. Unless you have the qualifications to be a peer reviewer on the particular topic in the paper. In which case you still don’t know for sure, but might have an idea of where weaknesses might lie.
–Peter
So, just for the heck of it, I did a Goggle to see which other vaccines also fail to produce sterilising immunity and maybe why … and if it’s even necessary to have an impact on a pandemic or any other outbreak of a viral disease. Here’s but one article…
Fair enough.
I was teething on counter intuitive results. Dad’s favorite topic whenever I asked a question as a teen.