Unlike static websites, the blogging paradigm allows a great deal of collaboration and interchange between people (witness the rallying of support for Bob Fiddaman over the legal threats he received from GSK). Investigative journalism through Blogs has becoming increasingly important. It is fair to say that the majority of concerns raised about the ENHANCE trial of Vytorin have been brought to light by investigative bloggers, Forbes magazine and the Wall Street Journal. The immediacy of blogs (and news reports) makes it hard to develop a body of systematic evidence, or to get keep historical perspecive. A Wiki provides a different model which allows both collaboration and systematic collection of information. I have therefore started the Scientific Misconduct Wiki to run in parallel with this blog. Hopefully this will form the seed of a peer reviewed online journal devoted to Scientific Integrity. For the moment, it will serve as a non-collaborative repository for reports and academic analysis of the integrity scandal involving Procter and Gamble and the drug Actonel (including the serious implications of the failed but rather sad attempts at coverup and delay by "regulators"). I will add to the collection in parallel with the blog postings. Currently, the Wiki contains all of the Collated News Reports about the Actonel story to set a framework and timescale for discussion of the science and principles.
Most recent in the news listings is this paper by Professor Martin Bland on the statistical re-analysis and attempted reconstruction of the first of the three implicated P&G publications. (Bland, M. 2007. "Risedronate, the BBC and me". Significance - Royal Statistical Society:@ 4:4:175-178). The reanalysis of the second of the three publications is here. Here is the raw data eventually provided by Procter and Gamble in April 2006 [Link] that formed the supposed basis of three P&G publications (two only in draft) as well as many scientific meeting abstracts. Under the circumstances, P&G should have allowed an open statistical analysis, but they have refused to allow such open scrutiny. The data is therefore encrypted (see "Procter & Gamble - Let's take the high road"). Nevertheless, please join the people who downloaded it previously.
This refusal of openness is unexpected, because P&G no longer have much to hide. P&G and "authors" (of publication 1) have of course admitted to all of the key "errors" in the first publication. The problems in the second and third publications were similar, but even more overt. Authors have also admitted to having been less than truthful over access to data - the central problem. Of ongoing concern is the refusal by P&G to provide confounding variables to authors, and more perversely the refusal to provide the wording of consent forms that patients would have signed.
If anyone would like to assist as editor of a peer-reviewed Wiki-based Scientific Integrity Journal, please let me know.
Earlier|Later|Main Page
