Hi Eric, Leigh-Ann, all,
I’m not sure I agree.
While unpublished data are used in some
reviews, and what Ahmed suggests would certainly help reduce bias in the interpretation of these results, you would
also need to be able to demonstrate that there was no bias in your search
strategy, and I think that it would be very hard to do so in this case. There
are two issues - the company involved may be much more likely to send things
your way that portray them in a positive light – for understandable
reasons, but ones which may conflict with your own goals of conducting an
impartial systematic review. Second, would you have been equally likely to
get hold of comparable info about products made by other companies? Your
search strategy should demonstrate as far as possible that (a) you would have
been just as likely to get this sort of information about the client’s
product, as about other companies’ medical devices, and (b) that the
information you would ‘find’ through the search strategy would be unbiased
in terms of how it portrays the device – other things being equal, you’d
be just as likely to come across info on the negative features of the device.
Any ideas for how to get around this?
Best wishes,
Catherine
From: Evidence based
health (EBH) [mailto:
Sent: 21 July 2010 18:22
To:
Subject: Re: question re using
unpublished, industry supplied data in systematic reviews
Hi Eric and Leigh-Ann,
What you are describing
is the grey literature; that semi-black, semi-white area that comprises of
material not readily available from the published literature. There are many
forms of grey literature and unpublished data from the files of the
manufacturer. It is both important and vital to include this information
because this information may not be available from any other source outside the
company files (unless it was published). This will diminish the probability of
publication bias and add merit to your systematic review. Cochrane systematic
reviews, health technology assessments, comparative effectiveness reviews, etc.
all use unpublished data, including from industry. The important this is that
you transparent where you got the information and to keep yourself sheltered
from any potential risk of bias towards or against the companies involved.
Example, when performing comparative effectiveness reviews for AHRQ, we are not
allowed to have any contact with the drug or devise manufacturers. This is done
for us by a third independent party to shield us from any possible notion of
bias. But in the end, the companies involved are asked to provide what is known
as a Scientific Information Packet (SIP) containing anything from monographs to
unpublished, or in-press articles, they would like us to consider.
Hope this helps.
Best wishes,
Ahmed
_______________________________________
Ahmed M. Abou-Setta, MD,
PhD
Postdoctoral Fellow/
Project Coordinator,
Principal Evidence-based
Medicine Consultant, PharmArchitecture Limited (
Member, Editorial
Advisory Board, The Open Medical Devices Journal
Member,
Member, Menstrual
Disorders & Subfertility Subgroup, Cochrane Collaboration
Aberhart Centre One,
Room 8412
Tel:
(780) 492-6248
Fax:
(780) 407-6435
E-mail:
[log in to unmask]
Website:
http://www.ualberta.ca/ARCHE/
From: Evidence based
health (EBH) [mailto:
Sent: July 21, 2010 11:04 AM
To:
Subject: question re using
unpublished, industry supplied data in systematic reviews
Hi,
Question posted below on behalf of a colleague who would appreciate
your advice on this.
Thanks for your help. Cheers,
Leigh-Ann
From: Eric Schuur
[mailto:[log in to unmask]]
Sent: Wednesday, July 21, 2010
9:29 AM
To:
[log in to unmask]
Subject: [AMWA.EDITING-WRITING]
Systematic Literature Review Content
I
am working on a systematic literature review covering clinical studies of a
class of medical devices. The client in this case manufactures one of
these devices and wants to use this review as part of a regulatory
filing. The client is asking if we think it would be appropriate to add
some information about safety features of the device to the clinical literature
review report.
Note
that this information is being given to us by the client and is not in the
literature. My guess is that the client wants to emphasize the safety of
their device since it is not yet on the market.
I
have my opinion about this, but don’t have data to support it. Does
any one of you have a feeling about whether including this sponsor-provided
info is appropriate in a systematic literature review?
Many
thanks for your feedback!
Eric
Eric
R. Schuur, Ph.D.
Medical
and Scientific Communications
VMWA
LLC
(650)
224-4178
Leigh-Ann Topfer
Coordinator, Information Services
Health Technology & Policy Unit
3021 Research Transition Facility
8308-114 St
phone: (780) 492-9079
fax: (780) 248-1546
e-mail: [log in to unmask]