Authorship, ghost-science, access to data and control of the pharmaceutical scientific literature: Who stands behind the word?

Dr Aubrey Blumsohn MSc , MB BCh, PhD, MRCPath

Reprinted from The American Association for the the Advancement of Science:
Professional Ethics Reports Volume XIX (3) Summer 2006

Aubrey Blumsohn is a pathologist and osteoporosis specialist, and previously Senior Lecturer in Metabolic Bone Medicine at the University of Sheffield in the United Kingdom . He was suspended from his academic post in September 2005 after discussing concerns about research integrity outside of his institution.

"the integrity of a body of literature is itself our society's ultimate temporal forum for negotiating life and death, suffering and wellness... the medical well-being of the society it serves is dependent on the question of who stands behind the word" – Fr Mark Gruber, 1999 (cited in ref 1)

The past two years has seen widespread commentary about the integrity of pharmaceutical medicine (2-13). The suggested remedy is that pharmaceutical companies must be divorced from direct involvement in researching clinical aspects of their own drugs ( 3 , 7 ). We are heading, like the Titanic, towards an iceberg of enormous size.

Pharmaceutical companies sell products under the banner of science and medicine. However their raison d'être is to make money. If industry gets involved in science, it has to balance genuine hypothesis testing and transparency against commercial interests, bureaucracy of drug regulation, and the financial consequences of dishonesty. This is not in itself a criticism – it is a simple fact.

Universities exist for a different reason:- to add to human knowledge and to disseminate that knowledge through publication and teaching. Subtle compromises (2) have allowed the pharmaceutical industry to develop an extraordinary stranglehold over the scientific process, academic discourse, regulatory safeguards and common sense (8,9,10,11,14). It is hard to see how safeguards for dispassionate scientific discourse can be sustained when medicine flagrantly disregards them.

Ghosts in the machine

The pharmaceutical industry is accused of overturning the usual safeguards of science. The most fundamental of these safeguards is the accountability of authors (1). Readers of legitimate science expect that stated authors are truly the authors, that they vouch for the work and that they would be able to defend their findings if challenged. They expect that authors have seen and scrutinized raw data, and would be able to provide that data if asked. That it is necessary to write this indicates how much we have lost.

Industry has been inclined to use universities to give tainted science a veneer of respectability, while denying the very basis of that respectability. “Ghost-writing” has been repeatedly criticized. However professional "medical writers" may sometimes have a legitimate role if clearly acknowledged. By emphasizing the “writing” aspect we divert attention from the far more important problem – that of “ghost-science” of which “ghost-writing” is only a part.

International standards were adopted by many scientific journal editors following embarrassing disclosures. These standards (15) reassert the obvious - that authors should state in writing that they have full control of all primary data, controlled the decision to publish, and will supply raw data upon request.

The usual definitions of scientific misconduct do not apply to pharmaceutical research. In February 2006 Gerald Schatten was accused of research misconduct (16). His crime was to have co-authored a stem-cell publication with the discredited Dr. Hwang Woo Suk while shirking the “responsibilities of verifying the data”. Schatten might have been irked to discover that at the same time, Procter and Gamble Pharmaceuticals (P&G) declared to the media that it was “standard industry practice” (17) to deny authors access to raw data in drug studies.

Lessons from Sheffield

In 2002 I signed a research agreement with P&G in collaboration with another academic, Professor Richard Eastell. The consequences of my disagreement with the company and with my collaborator have been widely discussed in the media. (17,18 ) and some original documents have been disclosed on a blog (19).

In spring 2006 The Journal of Bone and Mineral Research (JBMR, 20 ) placed an undated "Statement of Concern" on its website. The statement relates to one of three intended P&G publications (21) about change in bone turnover and fractures in patients taking the osteoporosis drug Actonel. The other two publications (one based on an extended set of the same data and another based on new data) have only been published in abstract form because I declined, as first author to sign journal declarations while being refused access by the company to randomization and event codes (17,18,19).

The research involved an important secondary endpoint in the key randomized trials used to gain regulatory approval for Actonel (annual sales ~$1 billion). P&G repeatedly refused to provide data codes to academic “collaborators”. This breached the terms of its contract with the University. Data were required by the academics to verify scientific reports, statistical analyses, meeting abstracts, and draft publications "ghost written" in their names. Over time, increasing information emerged to suggest that the data analysis and data presentation had been incorrect and misleading, but underlying data were still not disclosed.

The first of the three intended publications was submitted by my collaborator to the Lancet in 2002 and upon rejection was published in JBMR (21). The Lancet prescribes that an author must "state that he or she had full access to all the data in the study”, and "at any time up to 5 years after publication authors may be asked to provide the raw data". JBMR has similar guidelines. Academics at Sheffield would not have been able to provide data if asked (and were indeed not able to) - because they never had them.

Various statements made by P&G officials in their defense are illuminating (17,18,19). They claimed that “ we don't need to ask an independent person to analyse the data just to make a few people happy ” (the independent person being the intended first author). They claimed that by supplying authors with data “ industry loses the opportunity to demonstrate its ability to be a true partner in scientific endeavours ” (17). They suggested (19) that refusal to supply data to authors was in accordance with “PhRMA guidelines” (PhRMA is the main US lobby group that represents pharmaceutical manufacturers). They defended their actions in the press (17) by saying that it is “ standard industry practice ” to limit authors access to data, and that “ occasionally the researcher is given temporary and limited access to data to perform the analyses directly ”. In “legal” correspondence P&G attempted to redefine the meaning of “access to data” suggesting that showing an author company outputs or statistical interpretations somehow constitutes access to data.

The Bill of Rights and disclosure of data

In response to media scrutiny P&G produced a new “Bill of Rights” governing its relationships with academics in February 2006 (17). The bill stated that “ research authors will define and control the content and direction of any publication resulting from their work ” and will have “ final authority ” over all publication content. It stated that, although P&G would retain ownership of data researchers will “ own the analysis and conclusions ” and will be “ in no way restricted ” from publishing their findings. It says that researchers will have “ full access to all relevant data to confirm the accuracy of statements and conclusions ”. This statement simply stipulates the rights and obligations academics have always had, as well as the conditions for publication in any respectable journal. Nevertheless, it is a step forward.

In April 2006, after a three-year delay, P&G supplied me and Eastell with the data codes underlying the three intended publications. These data, as well as many documents and dozens of tape recordings confirm that the conclusions of the three publications were not in accordance with the data. Discrepancies were obvious. For example, in all three manuscripts, the x-axis of a critical graph was scaled so that about 40% of the data would not have appeared within the scale of the graph. A key conclusion of all three papers was that there was plateau at a commercially convenient point in the response relationship for the drug -- a matter of practical clinical relevance (18 explains how this would have benefited P&G). The data provided no credible evidence to support this conclusion in any of the three publications (17,22).

The problem of the regulators

The Sheffield dispute was discussed in the UK parliament in December 2005 and was transmitted by the Health Minister to the UK drugs regulator (the Medicine and Healthcare Products Regulatory Agency, MHRA) for “investigation.” The MHRA is itself accused of failing to examine or to secure raw data in drug licensing applications, simply accepting the word of industry with blind faith ( 6 , 14 ). Since this was precisely the problem in Sheffield, its disinclination to investigate was hardly surprising.

No investigation (or at least anything fitting that definition) took place. The MHRA failed to produce any report, declined to accept any documentary evidence ( 17 , MHRA response to FOI request #06/115), stated that the matter was of “low priority” ( 17 ), and that the agency does not have any procedure for investigating research misconduct (MHRA response to FOI #06/188). Further, it claimed that the drug regulator has no remit, nor any necessary obligation to be interested in the integrity of the scientific literature about drugs ( 17 , MHRA FOI #06/188) unless related to licensing (and collected using documentation appropriate for licensing). It even argued that it is “illegal” for a scientist to have data pertaining to information written in his name without the consent of the company “owning” that data ( 17 ,MHRA FOI #06/115). It refused to compare data it was sent from Sheffield with the original data it should have received and examined as part of the licensing process for Actonel. Initially, this refusal was on the basis that it would be “too much work” (MHRA FOI #06/059). Later, it admitted that that it had not in fact seen or retained raw data prior to approving the drug (MHRA FOI #05/404). With governments setting the standard for scientific conduct, it is hardly surprising that independent science has encountered such difficulties.

The problem of academics who don't acquiesce

There have been many cases where academics have refused to acquiesce. A dispute arose between James Kahn of UC San Francisco and Immune Response Corp. over effectiveness of an AIDS vaccine in a multi-center trial. The company objected to publication of the analysis of data (which was incomplete since the company refused to supply the rest to the researchers). When UCSF researchers refused to interpret the data more favourably, the company threatened legal action. The study was published with incomplete data ( 23 ). The company maintained that because it paid for the trial, it somehow owned the data and therefore the mode of presentation.

Many other cases happen beneath the radar. A recent example involves the drug Famciclovir used to treat herpes. A 1997 study comparing Famciclovir with its main competitor was funded by the manufacturer (then Smithkline Beecham). Study findings were not beneficial to the sponsor. They were published only a few months ago after a nine year delay, with a disclaimer that the authors were denied raw data and were forced to accept the company's own partial summary of findings ( 24 ). Scientists may disagree about the presentation of data. There can however be no legitimate debate when that data are not available for scrutiny even to the authors. Should we prescribe a drug knowing about such “missing” data? What of the patients who volunteered to take part in these studies?

Other disputes have related not so much to access to data but the right of authors to publish or speak about what they believe to be true. These include the celebrated cases of Nancy Olivieri ( 9 , 11 ), Betty Dong ( 25 ), David Healy ( 9 ), David Kern ( 9 , 26 ) among others. These instances should be discussed and analyzed so they are not repeated.

Where to from here?

The ethical challenge in pharmaceutical medicine is to use available data in the best possible way.

Data is derived from human participants who subject themselves to risk in the public interest. They have the right to know that the data derived from their assumption of risk are used properly. When data are closed to scrutiny even by the supposed authors of research, this cannot constitute an appropriate or ethical use of that data. Patients have to be involved in solving the problem.

We need to address the way in which drug regulators are currently operating. Recent reports about these agencies should provide a wake-up call ( 6 , 13 , 14 ). The stated intention of the FDA to support a preemption rule ( 27 ) is of grave concern. This would disallow lawsuits against drug makers if a drug is FDA approved, even apparently in the case of scientific fraud or the withholding of information by a company.

The problems of medicine could not happen without the complicity of medical journals ( 12 ), most of which receive substantial advertising and “reprint” income from industry. Anyone interested in the functioning of journals might wish to peruse my collated correspondence with the editor of The Journal of Bone and Mineral Research that I have placed online ( 28 ). Initially polite correspondence became confused as I encountered the endless distortion of reality that is part and parcel of pharmaceutical science. Other journals, such as PLoS Medicine, have made appropriate comments about their role ( 29 ), and we should therefore make sure such publications receive the respect they deserve.

Some continue to press for public right of access to anonymized raw clinical data supporting the licensing of medicines ( 30 ). Reasonable as this might seem, such proposals are unthinkable when currently neither academic authors fronting publications nor the regulators have seen that data. Issues of supposed “ownership” of data are irrelevant to data access. If the industry wishes to sell its products under the banner of science, it has to accept the rules of science.

Most importantly, as academics we need to reassert the importance of data and the meaning of authorship. We also need to assert “old fashioned” ideas of academic freedom, our right to speak the truth as we see it, and to allow that truth to be subjected to open debate.

In the words of George Orwell (1984) "Freedom is the freedom to say that two plus two make four. If that is granted, all else follows."


1. Sheikh,A. ( 2000). Publication ethics and the research assessment exercise: reflections on the troubled question of authorship. J. Med. Ethics 2000: (26):422-6. <>

2. Dieppe ,P. Complicity theory: an explanation for the `coxib problem'? J.R.Soc.Med. 2006:(99):273-4

3. Godley,F. Can We Tame the Monster? BMJ 2006: 333 <>

4. Mintzberg,H. Patent nonsense: Evidence tells of an industry out of social control. CMAJ 2006:174(4) <>

5. Chalmers,I. From optimism to disillusion about commitment to transparency in the medico-industrial complex. J.R.Soc.Med 2006:(99):337–41 <>

6. Healy,D. Did regulators fail over selective serotonin reuptake inhibitors? BMJ 2006:(333):92-5

7. Steinman,M.A. et al. Narrative Review: The Promotion of Gabapentin: An Analysis of Internal Industry Documents Ann. Intern. Med. 2006:(145):284-93 <>

8. Angell,M. The Truth About The Drug Companies: How They Deceive Us and What To Do About It. London: Random House,2005 (ISBN 0375508465)

9. Washburn,J. University, Inc: The Corporate Corruption of Higher Education. Basic Books, 2005 (ISBN 0465090516)

10. Krimsky,S. Science in the Private Interest: Has the Lure of Profits Corrupted Biomedical Research? Rowman,2004 (ISBN 0742543714)

11. Kassirer,J. On the Take: How Medicine's Complicity with Big Business Can Endanger Your Health. Oxford University Press,2004 (ISBN 0195176847)

12. Smith,R. Medical Journals Are an Extension of the Marketing Arm of Pharmaceutical Companies. PLoS Med 2005:2(5): e138 <>

13. Graham,D. COX-2 Inhibitors, Other NSAIDs, and Cardiovascular Risk. The Seduction of Common Sense. JAMA 2006: 296

14. House of Commons Health Committee. The Influence of the Pharmaceutical Industry. Fourth Report of Session, 2005 <>

15. International Committee of Medical Journal Editors (ICMJE) <>, World Association of Medical Editors (WAME) <>

16. Wade,N. University Panel Faults Cloning Co-Author New York Times 2006, Feb 11

17. Collated media reports Blumsohn case: <> and <>

18. Washburn,J. Rent-a-Researcher: Did a British university sell out to Procter & Gamble? Slate Washington Post, Newsweek Interactive, 2005, 22 Dec <>


20. Journal of Bone and Mineral Research <>

21. Eastell,R., et al. Relationship of early changes in bone resorption to the reduction in fracture risk with risedronate. J. Bone. Miner. Res. 2003:18:1051-6

22. Bland,M. Report for the BBC, 2006.

23. Kahn,J.O., et al. Evaluation of HIV-1 immunogen, an immunologic modifier, administered to patients infected with HIV having 300 to 549 x 10(6)/L CD4 cell counts: A randomized controlled trial. JAMA .2000:284:2193-202.

24. Wald A, Selke S, Warren T et al. Comparative efficacy of famciclovir and valacyclovir for suppression of recurrent genital herpes and viral shedding. Sex. Transmitted. Dis. 2006:33: 529-33.

25. Rennie,D. "Thyroid storm" JAMA 1997:277: 1238-43

26. Kern,D.G.. “The Unexpected Result of an Investigation of an Outbreak of Occupational Lung Disease.” Int. J. Occup. Environ. Health 1998:4(1):19-32

27. Avorn,J., Shrank,W. Highlights and a Hidden Hazard: The FDA's New Labeling Regulations. NEJM 2006: 354(23):2409-2411.

28. Correspondence with JBMR <>

29. The PLoS Medicine Editors: PLoS Medicine and the Pharmaceutical Industry. PLoS Med 2006:3:e329

30. Abraham,J. et al. Rethinking transparency and accountability in medicines regulation in the United Kingdom . BMJ 1999:318:46-7