The authors of this study no doubt spent hundreds of hours of their lives on this - all of that time was completely wasted.
As you would expect just from the headline, this study suffers from numerous methodological problems and biases. I'd be ashamed to have my name associated with it. A few examples:
โ1. The authors compare patients who had a healthcare visit and tested positive for COVID to patients who had a healthcare visit and tested negative for COVID.
โFine, but people do not go to the doctor and get a PCR test for no reason. The "control" group that tested negative for COVID was sick enough, or exposed enough, to warrant a medical visit and a test. This means they really don't have a control group. If not COVID, those people will have other infections that can also cause long lasting symptoms. Thus, the study is not comparing the effects of COVID to a healthy baseline. It is comparing Long COVID to "Long Flu", etc. This makes the excess risk of COVID look less significant.
โ2. โPCR tests are not perfect, and their accuracy depends heavily on timing.
โThe "test-negative" group will contain false negatives due to people who tested too early or too late. If these false negatives go on to develop Long COVID, it dilutes the difference between the two groups.
โ3. And finally the worst of all: billing codes. This methodological flaw is well known to anyone with even a passing interest in this space. It should be disqualifying from ever doing research on Long COVID. Long COVID symptoms (brain fog, PEM, etc) do not easily fit into billing codes, and many doctors don't understand them, so they don't get coded at all.
โ
The authors conclude that patients see a "return to baseline health within approximately one year" due to lack of billing codes. This doesn't mean that they returned to health - only that they stopped seeking treatment. Which makes sense, since there are no effective treatments for Long COVID, and the ones that do exist aren't found within the traditional medical system.
Stanford University researchers analyzed 14.4 billion claims from 244.7 million US patients and found long COVID may be far less common and shorter lasting than thought, with most excess risks fading within 1 year after infection.
medrxiv.org/content/10.64898โฆ