Extreme restriction design as a method for reducing confounding by indication in pharmacoepidemiologic research.
Matthew H SecrestRobert W PlattColin R DormuthDan ChateauLaura TargownikRui NieCarla M DoyleSophie Dell'AnielloKristian B FilionPublished in: Pharmacoepidemiology and drug safety (2019)
Restriction to a single indication for treatment may reduce confounding by indication in studies conducted in distributed data networks and other large databases.