factual

Slicing a randomized trial produces spurious subgroups

If you keep slicing and dicing a randomized trial into subgroups, you will find spurious effects—as when patients enrolled in May showed a survival advantage but those enrolled in June did not—demonstrating that post-hoc subgroup analysis manufactures meaningless findings.

factualpending

Speaker

Vinay Prasad

Evidence Quote

“you keep slicing and dicing a randomized trial, you're going to find these spurious things.”

Source

Vinay Prasad on Pharmaceuticals, the FDA, and the Death of Duty— EconTalk
Created: 6/13/2026, 3:32:57 AM

Your notes

Loading notes...