Eh, this limitation mentioned in the study seems pretty huge to me:

> Finally, the study only assessed conditions thought to be attributable to COVID-19 or post-COVID illness, which might have biased RRs away from the null. For example, clinicians might have been more likely to document possible post-COVID conditions among case-patients. In addition, because several conditions examined are also risk factors for moderate to severe COVID-19, it is possible that case-patients were more likely to have had an existing condition that was not documented in their EHR during the year preceding their COVID-19 diagnosis, resulting in overestimated risk for this group.

It also seems very possible that people with COVID diagnoses would be more likely to be on edge about their health / long COVID type symptoms. It also would have been nice if they differentiated things that can actually be tested for vs. things that can’t.

I’m not saying there isn’t an effect. I had bad cases of COVID twice. The first time I didn’t have any apparent effects afterwards other than my sense of taste taking a while to come back. After the second time - Delta - I definitely had some after effects.

We just need to be careful about claiming how many people get them. People are placeboing themselves into actually having symptoms (or thinking they do), and it’s hard to design studies to get around that.

For sure, that overall number could vary a lot, and this really needs further breakdown that distinguishes between more minor, annoying effects versus fairly life altering ones (like new diabetes or significant kidney damage). There was another one about breakthrough outcomes that did distinguish a difference in risk reduction for various severe and non-severe post-acute effects (the TL;DR being, vaccine breakthroughs are less severe but still far from harmless):

https://threadreaderapp.com/thread/1529477481127759872.html

Many of these are amenable to direct observation though. Research has already had many objective findings such as lung fibrosis, kidney function damage, abnormal liver function tests, shrunken brain matter (UK biobank study), etc. So the overall number might be debateable, but this is almost certainly going to have a significant impact on the number of healthy workers, not to mention the personal tragedies behind those statistics.

Risk is the product of probability and severity. The OP is talking about severity being corrected upwards, while the GP is talking about probability increasing too.