Regulatory Open Forum

This message was posted by a user wishing to remain anonymous

What is the proper approach to incorporate ad hoc or post hoc analysis not defined in either the protocol or the SAP? Can this now be added to the SAP after unblinding and should this analysis be added to the CSR as well or simply leave the protocol, SAP and CSR aligned as planned pre-DBL and unblinding.

Post hoc analysis means an analysis after the experiment has been completed.  It is usually frowned upon by FDA reviewers. All analyses should be included in the statistical analysis plan. There are times when developing the statistical analysis plan that the writers may not include certain subgroup analyses. An examination of these post hoc may indicate significance. One can perform post hoc analysis and include in the clinical study report. If this is the only thing that shows significance it is unlikely that the FDA will accept the outcome and point to a new clinical study. This very much depends upon the nature of the drug or device, etc. Recently we have had occasion to examine the summary basis of approval of several drugs; two of which had post hoc analyses. The FDA accepted one and denied the other.

------------------------------
Robert Schiff PhD, RAC, CQA, FRAPS
CEO
Schiff & Company, Inc.
1120 Bloomfield Ave., Suite 103
West Caldwell, NJ 07006
rschiff13@aol.com
973-568-3361
------------------------------

Original Message:
Sent: 30-Jul-2019 10:27
From: Anonymous Member
Subject: SAP and CSR post unblinding

This message was posted by a user wishing to remain anonymous

What is the proper approach to incorporate ad hoc or post hoc analysis not defined in either the protocol or the SAP? Can this now be added to the SAP after unblinding and should this analysis be added to the CSR as well or simply leave the protocol, SAP and CSR aligned as planned pre-DBL and unblinding.