LC FP
Fri, Apr-18-08, 17:08
I don't like to copy articles from Medscape but this is just too important to not be seen. The articles referred to are in this week's JAMA, and they're outrageous:
Vioxx Documents Offer Glimpse Into Ghostwritten Manuscripts, "Hire-A-PI," and Data Manipulation
April 16, 2008 (Chicago, IL) - Two new studies comparing published research articles with insider documents obtained during the Vioxx (rofecoxib, Merck) lawsuits offer a glimpse into the behind-the-scenes influence of industry [1,2]. In one analysis, researchers used court documents released during the litigation process to show that many manuscripts were written by the sponsor or by third-party medical-publishing ghostwriters, then attributed to an academic who, in some cases, was paid to have his or her name used. The second paper also used documents obtained during litigation to demonstrate that mortality rates of patients taking Vioxx, reported in published manuscripts, were lower than those described in internal company documents. The studies are published in the April 16, 2008 issue of the Journal of the American Medical Association (JAMA).
An accompanying editorial by journal editors Drs Catherine D DeAngelis and Phil B Fontanarosa explores what steps can be taken to reduce the influence of for-profit companies on clinical research, making it clear that the issues detailed in this week's issue are likely not infrequent [3]. "Make no mistake--the manipulation of study results, authors, editors, and reviewers is not the sole purview of one company."
In a statement, Merck has said that the company believes many of the comments in the JAMA papers are "false, misleading, or lack context."
The hidden hands
In the first study, Dr Joseph Ross (Mount Sinai School of Medicine, New York, NY) and colleagues identified approximately 250 documents, ranging from internal and external letters and emails to reports, presentations, and receipts that were relevant to their study, then linked these to subsequently published clinical-trial results or scientific review papers. They found evidence of Merck employees preparing manuscripts, sometimes in collaboration with medical publishing companies, then recruiting academic physicians to be the first authors. For review papers, Ross et al found documents describing Merck marketing teams hatching plans for review papers, contracting ghostwriters to conduct and write the articles, then recruiting academics to be authors. Of 96 published articles, 92% of clinical-trial publications (22 out of a total of 24 papers) disclosed financial support from Merck, but only 50% of review articles disclosed Merck's backing or included any mention of the author's financial compensation.
In an interview with heartwire, Dr Harlan H Krumholz (Yale University, New Haven, CT), senior author on the paper, explained that many academic physicians know that ghostwriting goes on, but it has been difficult to prove.
"This is something that infects our profession and puts us in a very bad light. The idea that drug companies and marketing departments are developing strategies for papers, then finding medical writing companies to write them, then identifying authors to put their names on them without even citing the industry support, let alone the fact that it was drafted for them and conceived by others: it's a huge problem. Most of us in academics know about this practice, many of us are offended by it, but this is the first clear documentary evidence of what happens behind the scenes."
Ross told heartwire that the issue of "guest authorship" is likely a problem across all areas of clinical research: "Cardiology has so many more trials and so many more medications developed in the field that it's just as likely to be encountering these problems as any other field."
Indeed, Krumholz believes cardiovascular medicine may be "particularly vulnerable."
"There is so much money in cardiovascular medicine, it's such an important field, and there are so many blockbuster products," Krumholz said. "I don't think there's anything about our moral fiber that's less than anyone else's, I think it's just endemic in the entire field that people supplement their incomes and CVs in this way. But if they want to do this, they should just fully disclose that yes, this idea was developed by the company, the first draft was written by a medical writing company, and I edited it. If they have the guts to put that in a disclosure, let them submit it and see if they can get it published."
Ross says he hopes their paper will prompt physicians, particularly outside academic circles, to take more of a stand. "This has been suspected within the peer-review community but not necessarily well-known within the practicing doctor community," Ross said. "I'm hoping that the increased attention from our paper will lead to more of an outcry saying, this isn't right. This isn't the way our science should be conducted and disseminated."
Additional protections warranted
In the second paper, Drs Bruce M Psaty and Richard A Kronmal (University of Washington, Seattle) looked specifically at how mortality rates were reported in trials of Vioxx for the treatment of Alzheimer's relative to how mortality rates were identified in internal company analyses and in information Merck provided to the FDA. They found that two published articles reporting the results of these Alzheimer's trials, described deaths, including "non–drug-related deaths," in the trials but concluded that, in terms of safety, rofecoxib was "well-tolerated." But in-house analyses, including a pooled analysis from both trials, pointed to a significant increase in total mortality--something that was not provided to the FDA or published publicly "in a timely fashion." For the mortality data that were provided to the FDA, an "on-treatment" analysis, rather than an intention-to-treat analysis, was used that "minimized the appearance of any mortality risk," Psaty and Kronmal write. One of the studies, it came out, did not even have a data safety and monitoring board to alert the FDA of any presumed harm from the drug.
"The findings from this case study suggest that additional protections for human research participants, including new approaches for the conduct, oversight, and reporting of industry-sponsored trials, are necessary," they conclude.
The Merck statement explains that the company was given no chance to respond to the papers before their publication in JAMA and that an "unbiased review" of the Merck documents shows "significant errors in the conclusions put forward by the authors of the JAMA articles." The statement also emphasizes that the articles were written by people who were also paid consultants for plaintiff lawyers during the Vioxx lawsuits, a point also made by the editorialists.
Responding to Psaty and Kronmal's conclusions, Merck Research Laboratories president Dr Peter S Kim calls the notion that Merck misrepresented the mortality data in any way "just plain wrong," As for Ross et al's paper, Kim insists that the "outside authors" for the published clinical trials "were intimately involved" in the studies and also that the company has "explicit policies governing the authorship of papers related to Merck's medicines and vaccines," which the company takes "very seriously."
Improving transparency
In their editorial, DeAngelis and Fontanarosa point out that journal editors shoulder some of the blame for "enabling" companies to influence published information without mandating more thorough disclosure procedures, but the problem infects all parties. They propose 11 suggestions for "drastic action" to improve transparency in medical research:
1. Registration of all clinical trials, with names of principal investigator(s) included.
2. Disclosure by all authors of the extent of their involvement in manuscript preparation, including the names of individuals who helped but are not listed as authors.
3. Disclosure by journals of funding sources for articles and relationships between authors and for-profit companies
4. Consideration by journals of how funding sources influenced manuscripts during initial review for publication.
5. Sole or primary responsibility for research belonging to academic investigators, not companies.
6. Independent statistical review of clinical-trial data by someone not involved with a for-profit company.
7. Mandatory reporting of authors to appropriate authority (dean or department chair) if they do not report financial relationships or allow their names to be added to manuscripts.
8. Mandatory reporting to appropriate authority of peer reviewers who provide confidential information to third parties; banning of these reviewers from future reviewing or publishing in the journal.
9. Removal of a journal editor if he/she is found to have permitted for-profit companies to manipulate journal.
10. Prohibition of companies from having input into content of educational materials created by professional societies or medical-education programs.
11. Freedom from influence of companies over individuals relating to speakers' bureaus, acceptance of gifts.
Vioxx Documents Offer Glimpse Into Ghostwritten Manuscripts, "Hire-A-PI," and Data Manipulation
April 16, 2008 (Chicago, IL) - Two new studies comparing published research articles with insider documents obtained during the Vioxx (rofecoxib, Merck) lawsuits offer a glimpse into the behind-the-scenes influence of industry [1,2]. In one analysis, researchers used court documents released during the litigation process to show that many manuscripts were written by the sponsor or by third-party medical-publishing ghostwriters, then attributed to an academic who, in some cases, was paid to have his or her name used. The second paper also used documents obtained during litigation to demonstrate that mortality rates of patients taking Vioxx, reported in published manuscripts, were lower than those described in internal company documents. The studies are published in the April 16, 2008 issue of the Journal of the American Medical Association (JAMA).
An accompanying editorial by journal editors Drs Catherine D DeAngelis and Phil B Fontanarosa explores what steps can be taken to reduce the influence of for-profit companies on clinical research, making it clear that the issues detailed in this week's issue are likely not infrequent [3]. "Make no mistake--the manipulation of study results, authors, editors, and reviewers is not the sole purview of one company."
In a statement, Merck has said that the company believes many of the comments in the JAMA papers are "false, misleading, or lack context."
The hidden hands
In the first study, Dr Joseph Ross (Mount Sinai School of Medicine, New York, NY) and colleagues identified approximately 250 documents, ranging from internal and external letters and emails to reports, presentations, and receipts that were relevant to their study, then linked these to subsequently published clinical-trial results or scientific review papers. They found evidence of Merck employees preparing manuscripts, sometimes in collaboration with medical publishing companies, then recruiting academic physicians to be the first authors. For review papers, Ross et al found documents describing Merck marketing teams hatching plans for review papers, contracting ghostwriters to conduct and write the articles, then recruiting academics to be authors. Of 96 published articles, 92% of clinical-trial publications (22 out of a total of 24 papers) disclosed financial support from Merck, but only 50% of review articles disclosed Merck's backing or included any mention of the author's financial compensation.
In an interview with heartwire, Dr Harlan H Krumholz (Yale University, New Haven, CT), senior author on the paper, explained that many academic physicians know that ghostwriting goes on, but it has been difficult to prove.
"This is something that infects our profession and puts us in a very bad light. The idea that drug companies and marketing departments are developing strategies for papers, then finding medical writing companies to write them, then identifying authors to put their names on them without even citing the industry support, let alone the fact that it was drafted for them and conceived by others: it's a huge problem. Most of us in academics know about this practice, many of us are offended by it, but this is the first clear documentary evidence of what happens behind the scenes."
Ross told heartwire that the issue of "guest authorship" is likely a problem across all areas of clinical research: "Cardiology has so many more trials and so many more medications developed in the field that it's just as likely to be encountering these problems as any other field."
Indeed, Krumholz believes cardiovascular medicine may be "particularly vulnerable."
"There is so much money in cardiovascular medicine, it's such an important field, and there are so many blockbuster products," Krumholz said. "I don't think there's anything about our moral fiber that's less than anyone else's, I think it's just endemic in the entire field that people supplement their incomes and CVs in this way. But if they want to do this, they should just fully disclose that yes, this idea was developed by the company, the first draft was written by a medical writing company, and I edited it. If they have the guts to put that in a disclosure, let them submit it and see if they can get it published."
Ross says he hopes their paper will prompt physicians, particularly outside academic circles, to take more of a stand. "This has been suspected within the peer-review community but not necessarily well-known within the practicing doctor community," Ross said. "I'm hoping that the increased attention from our paper will lead to more of an outcry saying, this isn't right. This isn't the way our science should be conducted and disseminated."
Additional protections warranted
In the second paper, Drs Bruce M Psaty and Richard A Kronmal (University of Washington, Seattle) looked specifically at how mortality rates were reported in trials of Vioxx for the treatment of Alzheimer's relative to how mortality rates were identified in internal company analyses and in information Merck provided to the FDA. They found that two published articles reporting the results of these Alzheimer's trials, described deaths, including "non–drug-related deaths," in the trials but concluded that, in terms of safety, rofecoxib was "well-tolerated." But in-house analyses, including a pooled analysis from both trials, pointed to a significant increase in total mortality--something that was not provided to the FDA or published publicly "in a timely fashion." For the mortality data that were provided to the FDA, an "on-treatment" analysis, rather than an intention-to-treat analysis, was used that "minimized the appearance of any mortality risk," Psaty and Kronmal write. One of the studies, it came out, did not even have a data safety and monitoring board to alert the FDA of any presumed harm from the drug.
"The findings from this case study suggest that additional protections for human research participants, including new approaches for the conduct, oversight, and reporting of industry-sponsored trials, are necessary," they conclude.
The Merck statement explains that the company was given no chance to respond to the papers before their publication in JAMA and that an "unbiased review" of the Merck documents shows "significant errors in the conclusions put forward by the authors of the JAMA articles." The statement also emphasizes that the articles were written by people who were also paid consultants for plaintiff lawyers during the Vioxx lawsuits, a point also made by the editorialists.
Responding to Psaty and Kronmal's conclusions, Merck Research Laboratories president Dr Peter S Kim calls the notion that Merck misrepresented the mortality data in any way "just plain wrong," As for Ross et al's paper, Kim insists that the "outside authors" for the published clinical trials "were intimately involved" in the studies and also that the company has "explicit policies governing the authorship of papers related to Merck's medicines and vaccines," which the company takes "very seriously."
Improving transparency
In their editorial, DeAngelis and Fontanarosa point out that journal editors shoulder some of the blame for "enabling" companies to influence published information without mandating more thorough disclosure procedures, but the problem infects all parties. They propose 11 suggestions for "drastic action" to improve transparency in medical research:
1. Registration of all clinical trials, with names of principal investigator(s) included.
2. Disclosure by all authors of the extent of their involvement in manuscript preparation, including the names of individuals who helped but are not listed as authors.
3. Disclosure by journals of funding sources for articles and relationships between authors and for-profit companies
4. Consideration by journals of how funding sources influenced manuscripts during initial review for publication.
5. Sole or primary responsibility for research belonging to academic investigators, not companies.
6. Independent statistical review of clinical-trial data by someone not involved with a for-profit company.
7. Mandatory reporting of authors to appropriate authority (dean or department chair) if they do not report financial relationships or allow their names to be added to manuscripts.
8. Mandatory reporting to appropriate authority of peer reviewers who provide confidential information to third parties; banning of these reviewers from future reviewing or publishing in the journal.
9. Removal of a journal editor if he/she is found to have permitted for-profit companies to manipulate journal.
10. Prohibition of companies from having input into content of educational materials created by professional societies or medical-education programs.
11. Freedom from influence of companies over individuals relating to speakers' bureaus, acceptance of gifts.