Why did the authors perform a meta‐analysis of studies with primary endpoints they consider clinically unimportant?