Skip to main content

Comparison of Safety Signaling Methods for Survival Outcomes to Control for Confounding in the Mini-Sentinel Distributed Database – Phase 2

    Basic Details
    Date Posted
    Status
    Complete
    Description

    This project builds on work done by a previous group that focused on "Safety Signaling Methods for Survival Outcomes to Control for Confounding in the Mini-Sentinel Distributed Database." The new work focuses on developing new survival methods that control for confounding using propensity scores. In rare event settings it is often not feasible to adjust for all known confounders directly, but a summary score such as a propensity score is necessary to reduce the dimensionality of the data. Currently implemented in Sentinel, there is a method that uses propensity scores via exposure matching and uses a stratified Cox Proportional Hazards model propensity score matched sets serving as the stratification variable. This can have particular limitations in the rare event setting since information is relegated into a large number of very small strata, and only those strata with outcomes contribute information to the model. The project assessed and developed other approaches that use propensity scores, such as stratification and direct adjustment that use the whole cohort (instead of the subset of informative matched sets).  

    Information
    Data Source(s)
    Mini-Sentinel Distributed Database (MSDD)
    Workgroup Leader(s)

    Andrea J. Cook, PhD; Biostatistics Unit, Kaiser Permanente Washington Health Research Institute and Department of Biostatistics, University of Washington, Seattle, WA

    Workgroup Member(s)

    Susan Gruber, PhD; Department of Population Medicine, Harvard Pilgrim Health Care Institute and Harvard Medical School, Boston, MA

    Susan R. Heckbert, MD, PhD; Department of Epidemiology, University of Washington, Seattle, WA

    Rima Izem, PhD; Ram C. Tiwari, PhD; Rongmei Zhang, PhD; Office of Biostatistics, Center for Drug Evaluation and Research, U.S. Food and Drug Administration, Silver Spring, MD

    Jennifer C. Nelson, PhD; Biostatistics Unit, Kaiser Permanente Washington Health Research Institute and Department of Biostatistics, University of Washington, Seattle, WA. 

    Michael Nguyen, MD; Office of Surveillance and Epidemiology, Center for Drug Evaluation and Research, U.S. Food and Drug Administration, Silver Spring, MD

    Robert D. Wellman, MS; Biostatistics Unit, Kaiser Permanente Washington Health Research Institute, Seattle, WA