@moog77 RCTs are not the be all and end all The effect size determined from them equals that of observational studies examining the same intervention and if poorly designed are worse than a good observational study. https://t.co/XVfdTQoPCm
4,808 followers
3,996 followers
People need to think through this paper harder. 23 of 34 meta-analyses thought the observational results sorta matched the RCT results. Where matched means sig direction. So 66% of areas. Versus 50% for a coin flip. For just areas with lots of RCTs. https
3,977 followers
@NehrilO @SophieGuillot30 @Dianepetitemam @karmarais @PhilippeRadault @BanounHelene @SabatierJeanMa1 @MartinZ_uncut @drpguerin @LMucchielli @SoloBucketpond @PavanVincent Je suis prêt à baisser mes standards, mais pas jusqu'au fond de l'abime dans lequel se
395 followers
@yoshkapundrick @CollignonPeter Also, you should always link to the latest version of a cochrane review: https://t.co/AT5lnAtx8P