|
Abstract:
|
Hemostasis evaluations are commonly used to assess clinical conditions in trauma, cardiovascular surgery and cardiology procedures to assess hemorrhage or thrombosis conditions before, during and following the procedure. Clotting time, clot stiffness/firmness, and platelet function are parameters to be evaluated when the new device is seeking a substantial equivalence to an already marketed device. Although these parameters are semi-quantitative, if the units are different between the investigational and the predicate devices due to using different techniques to measure these parameters, regression analyses are not an appropriate method to use. In this talk, whether using Passing-Bablok regression analysis with rank-based transferred data may be the possible analysis choice will be reviewed with a simulated data set. Some statistical challenges and issues in study design, and sample size calculation will be discussed as well.
|
