Using observational data from registries to compare treatments: The fallacy of omnimetrics

Abstract
No abstract available