here's a note on medcape https://www.medscape.com/viewarticle/935058?src=WNL_trdalrt_200807_MSCPEDIT&uac=161908AX&impID=2493666&faf=1
that suggests that
a) no randomized controlled tests show HCQ efficacy, the only ones done show no efficacy
b) therefore HCQ has no efficacy. QED
but not quite. let us remember, absence of evidence is not equal to evidence of absence. there are huge holes in the logic. and the note is quite easy to rebut, which i hope to do below.
basic issue: this doctor keeps asserting that xyz RCT trials he quotes are good: "The groups were well balanced because this was a randomized trial." exactly what does that mean? how do you know this was well-balanced? at what stage of the disease was HCQ administered? mind you, it HAS to be given very early for it to be effective according to all. if you give it to later-stage patients there will clearly be no effect.
first issue: there is a serious problem in ALL RCTs: you control for certain factors. what are those factors? age, gender, certain co-morbidities? how do you know THESE are the right factors and the only right factors? what about race? nationality? locality? climate? vaccination history? diet? general fitness? height? weight? allergies? chest infections? level of wealth? were these people already hospitalized, ie. late stage?
the RCT proponents run these trials on a bunch of hospitalized white people from one locality, and then announce that the results are accurate for ALL people in ALL countries at ALL times. why? on what basis do you extrapolate so wildly?
the factors that were controlled for in these 5 studies may have NOTHING whatsoever to do with the underlying causes. it is nothing but statistical correlation, not necessarily causation at all. you might as well control for astrological sign or date of the week you were born on, or height, and i'm sure you can find spurious correlations there as well.
here's my general concern about the whole idea of RCTs, which have been elevated to the status of some shibboleth. i think basically RCTs are totally bogus and in particular abhijit's work is particularly meaningless.
second issue: there is a serious ethical problem with RCTs. if you find that a drug is actually effective, then will you stop the trial because you are unethically denying the placebo users the drug?
third issue: the 5 quoted studies plus the preprint are using far greater amounts of HCQ than recommended in india, which is 200mg/day.
study 1: n=150. 1200mg first day followed by 1200 on 3 subsequent days, then 800. no placebo. hospitalized
study 2: n=821. 1400 first day/600 subsequent days. placebo. not hospitalized? (says asymptomatic)
study 3: n=497. same as above. no placebo. non-hospitalized
study 4: n=93. 800/400. no placebo. non-hospitalized
study 5: n=667. 800/800. no placebo. hospitalized
preprint study 6: n=4716. 1600/800. placebo. hospitalized
these studies are all over the place, with overdoses of HCQ being used, some with placebos and some without, some hospitalized and some non-hospitalized. how do you draw any consistent conclusions from these studies?
the medscape note is vacuous.
--
sent from xiaomi redmi note 5, so please excuse brevity and typos
No comments:
Post a Comment