I agree that forest plots do a good job of showing what they purport to. The main issue I see is that people tend to over-interpret subset results when this analysis is meant to be more "hypothesis-generating" for individual subgroups.
Comments
Log in with your Bluesky account to leave a comment
It's true that forest plots often show results for small subgroups. However, as you know, when analyzing trial results for a rarer lung cancer like ROS1+, data from small subgroups is often all we have.
Agree. I'm a believer in using the best you've got ("in the land of the blind, the one-eyed man is king"), as long as we interpret data in small subgroups with appropriate caution that the signals are provisional & not definitive answers.
Comments