BiMo Clarification Email - Cervarix, May 23, 2007
From: Sharon.W.Shapowal@gsk.com [mailto:Sharon.W.Shapowal@gsk.com]
Sent: Wednesday, May 23, 2007 11:21 AM
To: Sullivan, Helen M
Subject: request of 21 May 2007
Dear Helen: My colleagues in biometrics are considering the request of 21 May 2007 and offer the following draft proposal (they will discuss further tomorrow, as one of our colleagues is out of the office today). Yet, to help us in resolving the issue with "DATE" field, I'm sending this ahead for your BiMo team to consider.
Thank you for advising. I am pasting the original request below (blue text), for comparison.
Kind regards, Sharon
S.W.Shapowal Director, Adult Vaccines
"We request that you submit a table which includes a complete list of all protocol deviations reported for Study HPV-008. We request that the table include the following information for each deviation:
Type of Deviation
Code for deviation
Study Cohort(s) impacted by the deviation (i.e., safety, and/or immunogenecity, and/or efficacy)"
I have looked at the question from FDA and to Section 6.2.3 [HPV-008 clinical study report] and have the following proposal:
We provide two additional datasets (in XPT format) with the following information:
1) A dataset ELIM_DES which describes the elimination codes that are already in the database in a dataset WELIM (which FDA already has). This dataset would contain:
PID (subject number),
ELIMCODE (elimination code),
â—� ELIM_DES (description of code as in clinical report), â—� ELIM_COH (cohorts for which this code is used).
Note that PID, CENTER and ELIMCODE are also in WELIM and we could simply describe on one page the code and the cohorts for which this code is used => no need to have this XPT file. With this dataset they have an explanation of all the subjects that were eliminated from the different analyses.
Question: How should DATE be entered/interpreted, since this is different for each code, e.g., for a vaccine prohibited by the protocol we should give the date of concomitant vaccination and the date of study vaccination, but for a subject whose blood sample is out of interval, we should give both the date of vaccination and the date of blood sample, ... Proposal: Propose to not include the date in this dataset (Alternatively, perhaps give an explanation per code which dates are used and where they can be found)
2) A dataset NO_EXCL, which contains the subjects that had a protocol deviation but were not excluded (e.g., vaccination in dominant arm, longer needle). These would be the subjects listed in Section 188.8.131.52 of the interim report. This XPT file would be needed, because the information on, e.g., a longer needle, is not contained in the database.
Note: Although these datasets do not contain group information, they still have to be provided by the external stat (through the firewall committee), because there is a code (2060, infection with HPV-16 or 18) that only the external statistician has access to.
"EMF <fda.hhs.gov>" made the following annotations.
This message was sent by Glaxo Wellcome across the Internet in encrypted format and was successfully decrypted, unless otherwise noted. Glaxo Wellcome ==============================================================================