Friday, February 15, 2013

How to Fix Your Clinical Trial

Don't like your study results?  Just drop the patients who inconveniently don't benefit from treatment.  That's what these authors have documented as occurring at Pfizer during the trials of gabapentin for off-label treatment of migraine prophylaxis, bipolar disorder, and neuropathic pain.

As part of legal proceedings against Pfizer, internal gabapentin study documents became available for public review.  These authors collected these internal study documents and attempted to correlate the study methods and enrollment between the internal documents and subsequent journal publications.  What these authors found were important irregularities between study protocols and published results.

Specifically, the authors identified at least two studies that randomized patients, but then failed to report their enrollment in subsequent publications.  In addition, even when patients were reported – the intention-to-treat analysis was altered to further exclude additional patients.  They also noted missing pre-specified safety analysis in nearly all publications, despite their present in original study protocols.

Clinicaltrials.gov and other transparency campaigns are steps in the right direction – but, clearly, those steps can be only of limited effectiveness if this sort of unethical results massaging remains rampant.

"Differences in Reporting of Analyses in Internal Company Documents Versus Published Trial Reports: Comparisons in Industry-Sponsored Trials in Off-Label Uses of Gabapentin"

No comments:

Post a Comment

Comments on posts > 10 days old will be moderated; blame spammers.