Healthy Skepticism Library item: 7081

Warning: This library includes all items relevant to health product marketing that we are aware of regardless of quality. Often we do not agree with all or part of the contents.

Publication type: Journal Article

Gogol M.
Cochrane reviews v industry supported meta-analyses: ten questions to assess bias in medical research.
BMJ 2006 Oct 28; 333:(7574):916-7
http://www.bmj.com/cgi/content/full/333/7574/916-b

Abstract:

EDITOR-Jørgensen et al show that much published research may be biased.1 The following 10 questions, the result of my experience as a hospital physician, may be a practical way to assess bias.

Is the research conducted by or together with the sponsor?

Is there open access to the complete study design?

Is there a declaration of competing interests?

Are members of the publishing committee receiving money or other benefits from the sponsor (grants, consulting fees, lecture fees, other) and/or reporting stock holding?

Are members of the publishing committee employees and/or stock holders of the sponsor?

Are employees of the sponsor involved in any way in the data management? (If question 5 is judged yes or unknown tick the same here.)

Are employees of the sponsor involved in any way on judgment committees, i.e. clinical endpoint committee? (If question 5 is judged yes or unknown tick the same here.)

Was the study monitoring partly or totally done by the sponsor?

Was the data management partly or totally done by the sponsor or a for-profit organisation?

Does the abstract address critically the limitations of the study?

Answer each question “yes,” “unknown,” or “no.” Count two “unknown” as one “yes.” The possibility of bias arises with the number of questions you judged yes or unknown. Be careful if you have ticked yes 50% or more of the time and discuss the paper with your colleagues.

Keywords:
Bias (Epidemiology) Biomedical Research* Drug Industry* Meta-Analysis Pharmaceutical Preparations* Review Literature*