Transcript FDA-Indmeddev05

Other Issues for Statisticians in
Medical Devices
Gregory Campbell, Ph.D.
Director, Division of Biostatistics
Center for Devices and Radiological Health, FDA
FDA-Industry Workshop
Washington DC
September 16, 2005
1
Sessions with Strong Device Relevance
•
•
•
•
•
•
Bayesian Trial Design Thurs. AM
Pharmacogenomics Thurs. PM
PostMarket Adverse Events Fri. AM
Non-inferiority Studies Fri. AM
Statist. Eval. of Diagnostic Products Fri PM
Statist. Methods for Nonrandomized and Subverted
Trials Fri. PM
2
Combination Products
• Combination drug-therapeutic device
– Drug-eluting stents
– Drug pumps
• Combination drug-diagnostic device
– Devices that diagnose and then deliver therapy
• Separate drug, separate diagnostic device
– Pharmacogenomics Co-development document
3
Communication with CDRH
• How to request a statistical meeting
• Who to contact?
• When? Early with pre-IDE and pre-PMA
meetings
• Agreement and Determination Meetings
4
Least Burdensome Provision and
Levels of Evidence
• At first sight one might think that an historically
controlled study is less burdensome than a
randomized clinical trial. However, if the
historical control cannot be demonstrated to be
comparable with the data from the current study,
then the study may be not controlled (so no
comparison is possible) and another study may
need to be launched.
5
National and International
• PhARMA and DIA versus Advamed and MDMA
• International Conference on Harmonization (ICH) versus
Global Harmonization Task Force (GHTF)
• Organizations with an international presence (DIA vs.
Advamed) as well as standards organizations: ISO,
CLSI (NCCLS), ASTM, AAMI
• Few statisticians who are in government regulatory
positions outside US (Japan PMDA, UK, Germany) all of
whom serve both drugs and devices
6
Other Statistical Venues
• Joint Statistical Meetings
– 9 device sessions in Minneapolis orchestrated by a committee
from industry and CDRH and an organizational meeting
• Society for Clinical Trials and ENAR
• Standards organizations
• Other professional meetings where statistical topics arise
• Help to plan future meetings
7
8