Interesting preprint, but it has issues.
1. They added in a control group during their revision, but there's relatively little information about them; they barely get mentioned in the paper.
2. The "control group" isn't actually "people who've never been infected" - it's...
1. They added in a control group during their revision, but there's relatively little information about them; they barely get mentioned in the paper.
2. The "control group" isn't actually "people who've never been infected" - it's...
https://twitter.com/epi_michael/status/1483143937359495172
"people who were invited because they got a negative result at the time the other subjects got a positive result". It's unclear if anything was done to ascertain a lack of infection beyond this (possible question on the questionnaire?). There is no discussion of serology.
3. Most single-dose individuals were infected than vaccinated, while most two-dose individuals were vaccinated then infected. A poor showing for single-dose individuals mainly just implies that "vaccination doesn't cure Long COVID", and doesn't say much about prevention.
4. The double vaccinated cohort is older than unvaccinated, with 1/3rd over age 60, vs. 11% unvaccinated and 17,6% single-vaccinated. Long COVID is more common in the young. It's also slightly more male (Long COVID is more common in women).
5. A hospitalization rate among double vaccinated only slightly below that of the unvaccinated (7,1% vs. 9,1%) in the pre-Omicron era, when hospitalization protection should have been an order of magnitude greater among the double-vaccinated, also emphasizes the demographic...
differences between groups.
6. The double vaccinated group was infected much later than the single- and unvaccinated groups (median= 114,5d vs. 348d and 246,5d, respectively) , and thus likely faced different strains.
7. While they surveyed many symptoms, they only report...
6. The double vaccinated group was infected much later than the single- and unvaccinated groups (median= 114,5d vs. 348d and 246,5d, respectively) , and thus likely faced different strains.
7. While they surveyed many symptoms, they only report...
them for unvaccinated, single vaccinated, and double vaccinated - again, the addition of a control group feels like an afterthought.
8. They do a regression analysis only for the "most common symptoms". No! This is a common mistake. "Common symptoms" that are *common in the...
8. They do a regression analysis only for the "most common symptoms". No! This is a common mistake. "Common symptoms" that are *common in the...
general public as well* lead to the results getting washed out by background noise. The focus should be on symptoms that are *rare in the general public* but common post-COVID.
9. The study finds that Long COVID symptoms tend to be confined to the elderly - contradicting most..
9. The study finds that Long COVID symptoms tend to be confined to the elderly - contradicting most..
studies.
Altogether, an interesting preprint, but with some problems, and at the very least needing a lot of rework before publishing (at least with respect to the seemingly hastily-added control group, of which almost nothing is known, yet on which the conclusions hinge).
Altogether, an interesting preprint, but with some problems, and at the very least needing a lot of rework before publishing (at least with respect to the seemingly hastily-added control group, of which almost nothing is known, yet on which the conclusions hinge).
• • •
Missing some Tweet in this thread? You can try to
force a refresh
