Postmarketing Pharmacoepidemiologic Safety Studies Using Electronic Health Care Databases: Matching Control Subjects to Resemble Treated Subjects

• Activity Number: 43
• Type: Contributed
• Date/Time: Sunday, August 3, 2014 : 2:00 PM to 3:50 PM
• Sponsor: Section on Statistics in Epidemiology
• Author(s): Violeta Hennessey*+ and Fei Xue and Haijun Ma and Trevor McMullan and Paul Petraro
• Companies: Amgen and Amgen and Amgen and Amgen and Amgen
• Keywords: Propensity score matching ; matching on key confounders

Abstract

An increasing number of postmarketing pharmacoepidemiologic safety studies are being conducted using electronic healthcare databases. Studies that involve electronic healthcare databases are typically complex and require matching control subjects to resemble treated subjects. In this paper we evaluate different matching methods for a postmarketing observational study that requires medical chart review for confirmation of a case. Because of the high cost of medical chart review and the relatively small sample size of treated subjects and extremely large sample size of control, only a sample of controlled subjects will be matched with the treated subjects.