Abstract:
|
Under some circumstances, a pre-market application of medical devices may be supported by a single arm study where clinical endpoint is compared with a performance goal (PG). Occasionally, performance is expected to be different between two clinically defined subgroups of patients, and an approach based on a weighted PG is adopted. In this article, we discuss considerations of current practice where weights need to be pre-specified and fixed. We also propose an approach which relaxes the fix weights requirement.
|