TL20: Statistical Design and Analysis Issues for Cardiovascular Medical Device Studies
*Gary Lynn Kamer, FDA/CDRH/OSB/DBS 


The purpose of this roundtable will be to discuss both participant-requested study design and analysis issues/options as well as those suggested by the organizers. While protecting confidentiality, examples may be given. The advantages and disadvantages of competing designs or analyses may be discussed. Examples of some possible topics are i) the comparison of adjustment techniques involving propensity scores and covariate-regressive methods of adjustment and ii) how to do tipping-point analysis for Kaplan-Meier analyses involving either comparison to a performance goal or a comparison between two subject cohorts. Since statisticians can usually find disagreements over even the most simple design and analysis issues, this roundtable should prove to be quite lively.