Guest blog by Kim Klausner, Tobacco Digital Library Manager, University of California, San Francisco, USA
The Drug Industry Document Archive (http://dida.library.ucsf.edu) has added 241 documents about the marketing of Seroquel from the files of AstraZeneca. These documents show how the company obscured the potential for patient weight gain and diabetes from physicians and regulatory bodies and how they balanced the desire for sales with the need for scientific rigor. You can find these documents by entering “ddu:2010*” without the quotation marks in the query box.
As usual, the value of these documents is found by examining the corpus as a whole. I have, however, chosen a few documents that illustrate to me, a non-scientist, how a company deviates from what I understand as standard scientific methods.
http://dida.library.ucsf.edu/tid/tgv09b10: “[Do] not to discuss details surrounding trial 41 with any external customers” who include investigators.
http://dida.library.ucsf.edu/tid/giv09b10: How to spin and de-emphasize weight gain.
http://dida.library.ucsf.edu/tid/gmv09b10: AstraZeneca Director of Clinical Research asks for information about BristolMyersSquibb’s trials in exchange for sexual favors.
http://dida.library.ucsf.edu/tid/hiv09b10: Suggests data mining of Study 50 focusing on efficacy by measuring psychotic relapse.
http://dida.library.ucsf.edu/tid/mov09b10: Suggests that AZ-supported research has suspect results.
http://dida.library.ucsf.edu/tid/nhv09b10: Discusses how much to disclose about Study 15.
http://dida.library.ucsf.edu/tid/nnv09b10: Suggests pooling data to achieve certain efficacy claims.
http://dida.library.ucsf.edu/tid/piv09b10: “all off-label slides are financed outside of commercial for obvious legal reasons”
http://dida.library.ucsf.edu/tid/pjv09b10: “R&D is no longer responsible for Seroquel research – it is now the responsibility of Sales and Marketing.”
http://dida.library.ucsf.edu/tid/dpv09b10: Discusses giving limited data to an investigator on Trial 31 because he is perceived to be a “Lily advocate.” “I don’t want to irritate him nor give him the impression that we are hiding data.”
http://dida.library.ucsf.edu/tid/wjv09b10: Proposal for a study on the diabetogenic and hyperlipidemia side effects of quetiapine; concern about the potential risk and damage if the study had negative results.
http://dida.library.ucsf.edu/tid/ajv09b10: Questions why “limited” would remain as a qualifier to weight gain in the Core Data Sheet even though SERM [Safety Evaluation and Review Meeting] had decided otherwise.
http://dida.library.ucsf.edu/tid/cov09b10: Email re the request of Dr. Ghaemi, who was asked to be an author on the Trial 104 manuscript, for the study’s raw data. Suggests dropping Dr. Ghaemi as an author.
http://dida.library.ucsf.edu/tid/mpv09b10: Contemplates Study 125 endpoints based on potential adverse effects on US sales.
We’d like to thank Kris Hundley of the St. Petersburg Times and Ed Blizzard of Blizzard, McCarthy and Nabers for facilitating the addition of these documents to DIDA.
We’d like to add thousands more pharma documents to DIDA to increase the amount of data available publically about practices that degrade the way drugs. One of the major obstacles is that documents sometimes do not come with certain necessary pieces of indexing information (metadata) such as date, title, author and document type. Creating this metadata is costly because it must be done manually.
We’d like to try a ‘crowd sourcing’ solution to this problem, drawing upon the labor of many people to do the work. We’re hoping to create a password-protected, web-based mechanism that people could log into. They’d read/skim a document and enter specified information into a form. We’d make sure the entries are accurate and once a critical mass of documents was indexed we’d add them to DIDA.
We’re thinking that the following groups of people might sign on to help:
- students in classes in medical anthropology, sociology, history of medicine, public policy, etc., as part of a class assignment or for extra credit
- journalists who want to view documents before they are “public”
- members of the public who want to contribute to something important, possibly retirees
We have to find funding for programming this software and we have to recruit a “crowd” to do the indexing. If you have any thoughts about how we can make this happen, please email Kim Klausner at firstname.lastname@example.org.