Traditionally, Clinical trials have either relied on the Fisher approach of randomness, assumed almost complete similarity to an already approved device, or assumed no prior knowledge or understanding about the clinical application. However, medical device development often involves considerable engineering modeling work that can provide benefit in reducing uncertainties and providing insights that can serve as an effective starting point in clinical trials. In collaboration with the FDA through the MDIC (Medical Device Innovation Consortium), and building on the FDA Guidance 2010 on Bayesian Clinical Trials, engineers and biostatisticians have been able to structure the engineering modeling work on a medical devices so as to provide a basis for an informative prior for Bayesian Clinical Trials. Incorporation of engineering models as prior knowledge in a Bayesian clinical trial design can decrease sample size and trial length while handling type I and type II errors. This approach has been successfully applied in the clinical trials for medical devices for defibrillation.
Tarek Haddad currently leads Medtronic’s AI Research as Senior Director; during his 17 years at Medtronic, he has been recognized as a Bakken Fellow and Technical Fellow, and also as a DRM/ DFSS Master Black Belt. He received his master’s degree in Biostatistics from the University of Minnesota.
Adam Himes has worked as a mechanical engineer who embraces statistical engineering in dealing with engineering modeling challenges. During his 17 years at Medtronic, he has been recognized as a Senior Distinguished Engineer, Bakken Fellow, Technical Fellow and as a Medtronic DRM / DFSS Black Belt. Adam received his MS and BS in Mech E from the University of Minnesota and Michigan Technological Institute respectively.
Book now