malpractice in medicine Flashcards
name some ways trials can play design games
by choosing an obscure comparator
comparator with wrong dose
choosing outcomes that inflate effects
avoiding active comparators
name some ways trials can play analysis games
making plans up as go along
torture the data using lots of different tests
use subgroups to show significance
using percentage relative change
testing the wrong things
avoiding intention to treat
changing your primary outcomes to the one that looks best
ignoring baseline differences
stopping the trial early when have positive results
name some ways trials can play write up games
mechanism smoke screen misinterpreting stuff with equivalence avoid mentioning side effects inflating scales using positive language (framing bias) selective reporting bias
name some ways trials can play publishing games
dont/delay publish neg studies
publishing in obscure journals
salami publishing
duplicate publishing