Kwan et al. POTS Following COVID-19 Vaccination vs. Infection: Perspective Manipulation via Baseline Difference Subterfuge and Mixed-Exposure Confounding?
Maybe I'm wrong. Maybe it's an honest mistake. Either way, if these are real issues with this study, it's fatal.
Everywhere I look, are examples of data cookery and manipulation to warp public perception on vaccine safety and efficacy. Frankly, it’s exhausting in its predictability, and I’m tired of seeing junk science. So please forgive the tone of suspicion of intent.
Representativeness is such an important part of epidemiological and clinical studies that students in my classes in the Analytics Track at IPAK-EDU learn in many classes, like “How to Read & Interpret a Scientific Study”, “The Math of Vaccine Science”, “Applied Biostatistics”, and “Principles of Epidemiology” that for a study to be valid, the results have to be generalizable to the population of interest. If the sample group, or cohort, being studied is not representative of the general population about which an inference is sought, there is one or more bias in the data, and making policy or establishing medical standards of care based on biased data is hazardous.
The most recent of these is the manipulation of data on the rates of POTS (positional orthostatic tachycardia syndrome) in a comparison of the vaccinated and the infected is actually an oversight in interpretation based on the lack of representativeness leading to baseline differences in rates of POTS.
The study, entitled “Apparent risks of postural orthostatic tachycardia syndrome diagnoses after COVID-19 vaccination and SARS-Cov-2 Infection”, published Dec 12 in Nature Cardiovascular Research, concluded that the while the risk of POTS is higher following COVID-19 vaccination than before vaccination, and higher following infection than before infection, the risk of higher in people who have been infected compared to those who have been vaccinated.
This would seem to support the narrative to get vaccinated to avoid the relative risk increase in the infected.
There are some serious issues with the study (those as cynical I am will not surprised):
Did they make the correct comparison? We do not know how many of the “infected” were vaccinated prior to being infected, or, for that matter, after being infected. Indeed, reading the study I cannot tell if the two cohorts actually overlap to some degree. A subset analysis might show that the increased risk in the infected might be much higher in patients who were ever-vaccinated (any dose exposure) and infected (any order) compared to patients who were never-vaccinated and infected. If the risk is isolated to those persons who had both exposures, pathogenic priming may be at play and the correct comparison would be “never-vaccinated and infected” compared to “vaccinated and never-infected”. The data are large enough to allow meaningfully powered subgroup analyses. If the comparison I described is not the comparison conducted by the study, the study is fatally flawed.
Did they fail to account for non-representativeness (baseline differences)? The study reported to have found roughly 5x adjusted greater risk of POTS in infected persons compared to vaccinated persons failed to account for pre-exposure differences in the rate of POTS before exposure, which was already 5.5x greater in the vaccinated prior to vaccination compared to the infected prior to the infected. The paper does not mention the difference between the pre-exposure risk in the two cohorts, and this is a second fatal flaw that must be addressed.
Credit where credit is due. Dr. Mobeen Syed does a great job at 10:00 min breaking the pre-existing baseline different problem down and I thank him for bringing this second problem to our attention.
Did you learn something new?
Consider Buying Me a Coffee!
PS: Coffee w/Dr. Jack is being scheduled. A couple of hours of idle chitchat in an unstructured zoom before Jan 31. You can bring a friend!
It's the same story all over again. Every time they claim to be comparing some phenomenon following the vaccine vs following infection, they muddy the waters so much that it's impossible to tell what's what.
And then they use their false comparison (which always seems to show worse effects after infection) to insist on even MORE vaccination.
Looks like anyone that had covid and vaccination within 90 days of each other was excluded from any analysis. Can't see why they would do that.
Studies that defines any kind of risk and control periods (i.e. 90 days after exposure and 90 days before exposure in this case) tend to be highly sensitive to those definitions. Studies can totally flip or conceal depending on what you pick.