Why all randomised controlled trials produce biased resultshttps://buff.ly/2qR8gGO
-
-
Probably best if I wait for full reply so all can be given in proper context but here’s the preview...
-
In the vast majority of RCT’s for medical interventions, this is an impossible condition. I am guessing that zero of the 10 trials he cited could have done this. Why? Because we can’t just freeze a bunch of patients and then thaw them out & randomize all on the same day.
-
-
Those probabilities are decidedly not the same. Anyone with a reasonable understanding of probabilities should know this.
-
I find that passage difficult to parse. Is he saying that Pr(54 die | Treat) is to Pr(64 die | Control) as Pr(54 Heads | Coin #1) is to Pr(64 Heads | Coin #2)? Really baffling passage.
-
It’s an awful passage, and shows a poor understanding of basic probabilities. The probability of “getting 10 more heads than tails” in 624 flips is very different than the probability that there is no difference between tx given 54 deaths in one group and 64 in other group
-
Stuff like this is why it’s hard to take any of this article seriously, BTW. It’s not that RCT’s are perfect and should never be questioned; it’s that the person who wrote this chose a bunch of curious and/or outright wrong examples to illustrate his point
- 2 more replies
New conversation -
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.