Statistical approaches to group sequential monitoring of postmarket safety surveillance data: current state of the art for use in the Mini-Sentinel pilot.

摘要: Purpose This manuscript describes the current statistical methodology available for active postmarket surveillance of pre-specified safety outcomes using a prospective incident user concurrent control cohort design with existing electronic healthcare data. Methods Motivation setting is provided Food and Drug Administration’s Mini-Sentinel Pilot as an example. Four sequential monitoring methods are presented including Lan–Demets error spending approach, matched likelihood ratio test statistic approach binomial MaxSPRT special case, conditional sampling procedure stratification, generalized estimating equation regression permutation. Information on assumptions, limitations, advantages each provided, how method defines boundaries, what used, robust it to settings rare events or frequent testing. Results A hypothetical example approaches could be applied data comparing medical product interest, drug A, drug, B, providing type information one would have such drugs. Summary We described state outcomes. describe limitations while acknowledging areas future development. Copyright © 2012 John Wiley & Sons, Ltd. key words—incident cohort; observational study; postmarket; group monitoring; signal refinement;