This is a big step forward in improving the efficiency of clinical trials of drugs and biologics, and a big day for @US_FDA which I've been dreaming of for decades : https://www.fda.gov/news-events/press-announcements/fda-issues-guidance-modernizing-statistical-methods-clinical-trials #bayes #RCT #clinicaltrial #pharma
R to @f2harrell: P.S. Solved the pretty-printing problem with CotEditor by telling the print dialog to use standard color theme instead of the on-screen dark theme I use.
Computing tool of the day: I use Visual Studio Code Editor as my general purpose editor. But it's an overkill for almost everything I do. Enter CotEditor for Mac which so far is my dream editor (except for printing syntax-highlighted code). https://coteditor.com
I want to see more "outcome explained variation" plots like this to put genetic variants, biomarkers, protein signatures, and fMRI markers in the proper predictive context.
Unfortunate but the whole text (except the reference list) is actually shown in the preview. Can also read the Datamethods thread which also links to freely available resources such as this one: https://academic.oup.com/jrssig/article/22/3/40/8099023
I am thankful that we have a knowledgeable crusader against the silliness of non-comparative randomized trials, in @PavlosMsaouel #Statistics #ClinicalTrial
𝗖𝗹𝗶𝗻𝗶𝗰𝗮𝗹 𝗣𝗿𝗶𝗺𝗮𝗰𝘆
Despite the hype, perfectly free and widely available biomarkers—age and stage—remain the primary anchors of effect on log survival time (lower = worse) @f2harrell
This was a cool read of a new way to think of endpoints, where ordinal outcomes are incorporated to "boost" power. Question: what blocks the FDA from adopting such stuff? What can we do to make this type of research enter practice and to speed its validation?