Friday, September 25, 2009

The Reappearance of a Ghost of Seasons Past

About a year after we started Health Care Renewal, in late 2005, we wrote multiple posts about the complex and unfortunate case of Dr Aubrey Blumsohn's attempts to keep a research project honest. The early posts were here, here, here, and here. In this post, we summarized the case thus:

  • Dr Aubrey Blumsohn, a senior lecturer at Sheffield University, and Professor Richard Eastell performed a research project on the effects of the drug risedronate (Actonel, made by Procter & Gamble Pharmaceuticals [P&G]) under a contract between P&G and the University.
  • Although the research contract designated Blumsohn and Eastell as "Investigators" under whose direction the project would be carried out, Blumsohn was not given access to the original data collected by the project.
  • Despite numerous requests, (like this one), P&G refused access to this data repeatedly.
    Blumsohn was concerned that he and Eastell could be accused of scientific fraud if they continued to make presentations and write articles and abstracts without access to the data which they were supposedly writing about.
  • Blumsohn became suspicious that some of the analyses done by P&G could be misleading, especially related to a graph shown to him that omitted 40% of patient data.
  • Blumsohn objected to P&G arranging for papers and abstracts to be written by a professional writer, but with Blumsohn listed as first author. Blumsohn was concerned that such ghost-written documents were mainly meant to convey "key messages" in support of P&G's commercial interests.
  • Eastell warned Blumsohn not to aggravate P&G, because the company was providing a grant to the University which "is a good source of income."
  • After repeated failed attempt to get the data, Blumsohn complained to numerous officials at Sheffield University, including Eastell, medical school Dean Tony Weetman, University Vice-Chancellor Robert Boucher, and the Head of the University's Department of Human Resources, Ms R Valerio.
  • Still unable to get the data, he spoke with news reporters about his case. At this point, Sheffield suspended him, but then offered him a severance agreement if he signed a contract binding him not to make any detrimental or derogatory statements about the University and its leaders.

So the case involved suppression and manipulation of research, ghost-writing, institutional conflicts of interest, and attempts to silence a whistle blower. It provides lessons about the downsides of letting commercial firms sponsor and hence control human research designed to evaluate the products or services they sell; and of academic medicine becoming dependent on research money from such firms for such research. Although Health Care Renewal, being US based, most often writes about such issues in the US, this case is a reminder that they are global. (Note that we posted more about this case in 2006, here, here and here, but since then it has not gotten much public attention.)

Last weekend the (UK) Guardian returned to it:

The Guardian has learned that one of Britain's leading bone specialists is facing disciplinary action over accusations that he was involved in 'ghost writing'.

The General Medical Council will call Professor Richard Eastell in front of a fitness to practice committee. Eastell, a bone expert at Sheffield University, has admitted he allowed his name to go forward as first author of a study on an osteoporosis drug even though he did not have access to all the data on which the study's conclusions were based. An employee of Proctor and Gamble, the US company making Actonel, was the only author who had all the figures.

Experts believe the practice is widespread in Britain.

In a letter published in the Journal of Bone and Mineral Research, which carried the original study, he stated: 'In the original paper one of the authors, a statistician working for P&G, Ian Barton, had full access to all the data.' The authors had full access to all the analyses of the data that they requested, he said – but those analyses were carried out by the company.

The letter, published in 2007, also acknowledged flaws in the study. A later independent analysis of the data 'identified some errors and poor practice', he wrote. The study was designed to show the strengths of Actonel which was in fierce competition with a rival bone-strengthening drug called Fosamax, made by Merck.

Eastell's paper concerned a study carried out on behalf of Proctor and Gamble, comparing the bone density of women prescribed Actonel with others who were not. Only the company knew which women were on the drug and which were taking something else.

Eastell's colleague, Dr Aubrey Blumsohn, wanted the codes which would say which of the patients who suffered fractures had been on the drug. The company refused. Blumsohn took his concerns to Eastell, but in a conversation which Blumsohn says he taped , Eastell said he was concerned that persistent requests might damage the relationship they had with the company. Eastell is said to have told him: 'The only thing that we have to watch all the time is our relationship with P&G. Because … we have the big Sheffield Centre Grant [from P&G] which is a good source of income, we have got to really watch it.' .

So, after four years, this case has generated an official hearing of sorts. The hearing is obviously late, and seemingly will only be devoted to only one aspect of this case (ghost writing). However, at least our friends in the UK are doing something. I cannot recall a single case that resulted in any serious consideration of imposing negative consequences on anyone who was accused of suppressing research, manipulating research, endorsing ghost-writing, or intimidating a whistle-blower. In fact, many of the more troubling cases have never resulted in any sort of public discussion either at the institutions at which they occurred, or at any organization with relevant regulatory, or even just moral authority. So the GMC hearing is at least a step forward. Two cheers for the British GMC, and none for US universities, academic medical centers, professional societies, and government regulators.

(If anyone can remind me of a case in which there was a public discussion at the relevant institution, or some public consideration of the case by a regulatory agency, professional society, or some group with moral authority, please remind me of it, and I would be happy to post about it.)


MedInformaticsMD said...

But wait...

According to Wyeth's senior vice president of global medical affairs Dr. Joseph Camardo, medical opinion "evolves" and ghostwritten papers reflect scientific understanding when written. The authors alone also control the content and writing of the papers, though some do get "outside help" paid for by the pharma company.

And Wyeth's lawyer Stephen Urbanczyk acknowledged thatghostwriting is a marketing effort. But he said ghostwritten papers are "fair, balanced, and scientific."

So, what's the problem with ghostwriting?

We should have all papers ghostwritten, giving scientists more time to do more important things, like filling out their computerized lab notebooks, their management by objective updates, and their 360 degree evaluation papers from HR.


Anonymous said...

Not sure whether you mean any case or any of the pharma or health situations. In recent memory, Purdue disciplined Teleyharken (sp?). Florida punished Twitchell. Vermont and Poehlman. Kammer and Wake Forest. Sultan and Harvard.

How are those state medical boards doing in disciplining doctors?

The problem of self-regulation among the professions, while distressing, is common.

Roy M. Poses MD said...

Anonymous, I meant the comments to apply only to health care situations. Are any of your examples above in such situations?

Anonymous said...

The last three are all health care situations.

Anonymous said...

I should add that a short visit here-- show that regulatory agencies inthe US are certainly not ignoring research integrity.

Roy M. Poses MD said...

Anonymous, I confess to being too telegraphic in my writing sometimes, based on the assumption that most readers are familiar with the context of this blog. Maybe that is not a good assumption.

By "research manipulation" I did not mean garden variety research fabrication by an investigator meant to enhance the career of the researcher. Instead, we have used the term as short-hand for decisions made about the study question, research design, data collection, data analysis, or data dissemination made to further vested interests (usually the commercial, sometimes the ideological interests of a sponsor) other than merely to advance the career of the researcher. Such manipulation may involve collaboration among sponsors and investigators, and may at times fit the Transparency International definition of corruption: abuse of entrusted power for private gain.

I certainly know that there have been cases of individual researchers erasing, changing, or fabricating data to make the results more striking, and that when this has been discovered in the course of US government sponsored research, some researchers have been punished. But this is different, particularly in that in these cases the researchers often have acted alone, and counter to the interests of the sponsor.

Another way to make the distinction: we mainly discuss research manipulation done to favor the sponsor's interests, sometimes by the sponsor directly, rather than manipulation done by an individual researcher, often counter to the sponsor's interest. The distinction is important since in the former case, the manipulation may be more difficult to detect, and as I first noted, much less likely to cause negative consequences for the responsible parties.