I know my treatment doesn't work across a large population but what about subgroups?! Ah yes. Subgroups. Subgroups are absolutely a thing but how do we go about defining them? Do we just track everyone we treat then look back to see who responded and who didn't? Or is it more complicated than that? Find out on this dynamite episode of Nobel proportions!

Evaluation of Evidence of Statistical Support and Corroboration of Subgroup Claims in Randomized Clinical Trials.
Wallach JD, Sullivan PG, Trepanowski JF, Sainani KL, Steyerberg EW, Ioannidis JP.
JAMA Intern Med. 2017 Feb 13. doi: 10.1001/jamainternmed.2016.9125. [Epub ahead of print]

Due to copyright laws, unless the article is open source we cannot legally post the PDF on the website for the world to download at will. That said, if you are having difficulty obtaining an article, contact us.

Music for PT Inquest:
"The Science of Selling Yourself Short" by Less Than Jake
Used by Permission

102 Subgrouping Is Really Hard

PT Inquest

102 Subgrouping Is Really Hard