FDA Clinical Requirements

MEDICAL DEVICE 21 CFR PART 820 (The Regulations)
1

Attached is some information I got second hand, but I thought it
might elicit more conversation from the group.

“SOCRA and FDA’s Chicago District Office co-sponsored a conference
entitled, “FDA Clinical Requirements, Regulations, Compliance and
GCP” on May 16-17. Patricia Beers Block, Special Assistant to the
Director, Good Clinical Practice Program with FDA spoke on Part 11
and the recently issued guidance. During her presentation, Ms.
Block emphasized that it was no longer sufficient for sponsors to
provide clinical investigators a CD containing eCRF data at the
conclusion of a study. Instead, sponsors, according to Ms. Block,
are to ensure that investigators retain a contemporaneous copy of
eCRF data at the trial site. She stated that one means to achieve
this is by use of a `daemon program’ that writes directly to the
investigator’s hard drive, or to some other location at the clinical
trial site. Her concern as stated was that data could be changed by
the sponsor or another third party. Clarification was sought by an
audience member, and Ms. Block reiterated this position.”

Based on this, I believe the second paragraph of section C. is
geared to those situations where the source data is not generated at
the clinical site, which is supported by the 3 examples given. The
FDA is not looking for data to be retained on a patient’s PDA/eDiary
nor for a copy of source data to be retained in an escrow system
when an investigator is entering observations into his/her local
database. This recommendation is designed to ensure that the
clinical site retains the source data or a “certified copy” of the
source data. If data is being recorded directly to a sponsor or
perhaps a CRO’s computerized system, they are expecting that
a “contemporaneous copy” be provided back to the investigator.

My questions for FDA would be:

  1. As Greg originally posted, how contemporaneous does the copy
    pushed back to the investigator have to be? Assuming we validate
    the system to show that data cannot be altered at the sponsor site,
    is a nightly transfer to the investigator sufficient or must the
    copy be made truly in parallel with the original recording of the
    data?
  2. Is the fact that we validate the transfer process enough of a
    verification that copies are accurate and complete?
  3. What is the flexibility around situations where data is hosted
    and the investigator can access it remotely? Hosting could be at a
    third-party not involved with the study, at a CRO, or the sponsor
    site.

Opinions? Has anyone seen this contemporaneous copying in practice
in the clinical domain?