@RWJE_BA @stephensenn @f2harrell @yudapearl @dylanarmbruste3 @elmir1omerovic @drjohnm @kaulcsmc @GreggWStone The term "Reference Class Forcasting" emphasizes that the reference class must be defined by valid measures & the disease variability & bas
@doc_BLocke @BenjaminAbella @kurzmc @NickJohnsonMD @sarahmperman I don’t have access to the PDF, but sounds like they applied a multivariate risk-based model (https://t.co/myx0ul9bEd). Did they assess HTE?
@mikejohansenmd 4/ And a bit longer discussion here: https://t.co/7J6vdDr1CT
See this, for example, w/ @ProfHayward et al: https://t.co/zeTC8wEAR1 9/
@drjohnm This article that includes John Ioannidis as an author introduced me to this very interesting concept. https://t.co/FgiIzmnrXO
It depends on scale!! 😄
@Lester_Domes @f2harrell @boback @AnnalsofIM @ProfHayward Doug was also an engaged and enthusiastic coauthor on some of our earlier papers which framed HTE as scale-dependent: https://t.co/BotyJteOfm https://t.co/onBblGktgE
RT @msh_manu: HTE arises from key domains 1.Risk of the primary study outcome; 2.Competing risk; 3.Risk of treatment-related harm; 4. Direc…
RT @msh_manu: HTE arises from key domains 1.Risk of the primary study outcome; 2.Competing risk; 3.Risk of treatment-related harm; 4. Direc…
@WalkeyAllan @f_g_zampieri @ADAlthousePhD @MaartenvSmeden I see your point, but not sure in the case you adjust for the baseline, as proposed. I think HTE has diff. definitions and sometimes we are meaning diff. concepts. Also, I think HTE should be assess
@mikejohansen2 I'd rather have all primary analyses (reasonable priors based on theory &/or past empirical work) in main paper (0-2 HTE analyses [leave continuous vars continuous + quadratic]) & 1-3 adverse effects, & all other exploratory anal
Enjoyed reviewing article https://t.co/C1CMZVXReQ in Stats meeting today. Wasn’t quite sure of implications for practice though. Can see benefit for validated prediction models but endorsement to use internally-developed models #2 seems questionable.
RT @BerlinStroke: During our lunchtime seminar, we discussed treatment heterogeneity in clinical trials and this paper came up. https://t.…
During our lunchtime seminar, we discussed treatment heterogeneity in clinical trials and this paper came up. https://t.co/7exeiscFC2
@karimbrohi Not true. Pre-specified, adequately powered HTE w/ reasonable priors = just as legit as main effects assessments. Problem is that most HTE analyses lacks those requisites. https://t.co/5ajHwCD4ni
@f2harrell @medevidenceblog 3/ I do think RCTs should summarize variation in base risk & in ARR in RCTs, & check BaseRisk*Rx interaction https://t.co/5ajHwCD4ni
Proposal for multivariate risk-based HTE analysis described in journal Trials (Open Access) #pctGR http://t.co/3CeZpA9Ovy
Proposal for multivariate risk-based HTE analysis described in journal Trials (Open Access) #pctGR http://t.co/3CeZpA9Ovy
“@ChristosArgyrop: A non-fatal stroke that leaves one drooling is worse than a fatal one. And NNTs can mislead http://t.co/rqOzxDM6lo” QFT
“@ChristosArgyrop: A non-fatal stroke that leaves one drooling is worse than a fatal one. And NNTs can mislead http://t.co/rqOzxDM6lo” QFT
@kidney_boy @MtnMD A non-fatal stroke that leaves one drooling is worse than a fatal one. And NNTs can mislead http://t.co/nPPvVeNyjy
Heterogeneity of treatment effects in subgroup analysis of RCTs. This is a proposal that appears reasonable http://t.co/2C650fDmd6
Heterogeneity of treatment effects in subgroup analysis of RCTs. This is a proposal that appears reasonable http://t.co/2C650fDmd6
Heterogeneity of treatment effects in subgroup analysis of RCTs. This is a proposal that appears reasonable http://t.co/2C650fDmd6