Online Program

Combining data to study utilization and effectiveness of medical devices

*Laura Hatfield, Harvard 
Daniel B Kramer, Beth Israel Deaconess Medical Center 
Sharon Lise Normand, Harvard University 
Frederic S Resnic, Lahey Clinic Medical Center 
David Strauss, US Food and Drug Administration 

Keywords: medical devices, missing data, observational studies, meta-analysis

After receiving regulatory approval, medical devices are often deployed in more diverse populations, health care settings, and diseases than those on which the pre-market approval studies were based. Post-market surveillance ensures that safety and effectiveness estimates are updated to reflect this expanded use. A central challenge is assembling a data set that a) adequately represents the population of interest, b) contains sufficient clinical detail to define an eligible cohort and control for important confounders, and c) follows patients long enough to assess relevant outcomes. We focus on the setting of device-based treatment for heart failure: cardiac resynchronization therapy (CRT), in which a left ventricular pacing lead is combined with an implantable cardioverter-defibrillator (CRT-D) system. Pre-approval randomized controlled trials indicated differences in device efficacy in men and women; additional studies in the post-market setting are needed. We combine detailed patient-level data from multiple clinical trials with population-level administrative data to refine treatment effect estimates for men and women. This work is supported by contract DHHS/FDA-22320172C from the Center for Devices and Radiological Health, USA Food and Drug Administration.