"Cancer at the Core of Evidence-Based Medicine"

Image Courtesy of Wikipedia

If you've been following this blog, you may enjoy this 13-minute TED Talk recorded last year.  It features Ben Goldacre (pictured left), a British physician and science writer who's on a mission to expose the public health dangers of clinical trials that never get published.

Many thanks to Mr. Gil Batio, a member of the Continuing Medical Education group on LinkedIn, who brought this item to my attention.

After Chicago: Raising a Stink

Irene Hames PhD: "The literature is polluted."

One of the most memorable moments during the peer review congress in Chicago came when Irene Hames of the Committee on Publication Ethics told Ivan Oransky of Retraction Watch that 'the literature is polluted.'

As a former environmental reporter who won a public service prize for exposing abuse of sewage sludge used as crop fertilizer by Delaware farmers, I found Hames' toxic metaphor to be amazingly powerful.

Think about it: if the scientific literature is, in fact, polluted, we could really raise a stink.

Who are the polluters?

Via Wikimedia Commons

How are they dumping toxic waste into our literature?

Via Wikimedia Commons

How extensive is the pollution?

Via Wikimedia Commons

Whose job is it to catch the polluters?

Public domain image

Polluting a waterway is a misdemeanor in the United States, punishable by jail time. Why are there no legal penalties against polluting the scientific literature?

Alcatraz photo by Don Harting

Cleaning up pollution is expensive. Do we need the equivalent of a Superfund to undertake literary remediation?

Photo courtesy NOAA via Wikimedia Commons

What patient wants to join a clinical trial that may never get published, or might really be just a sneaky way to sell more drugs?

Photo courtesy RayNata via Wikimedia Commons

And perhaps most important, what happens when your family doctor bases his or her treatment decision on a toxic study?

 

Photo by Bill Branson for National Cancer Institute

Now, it's time for readers of Peer Review Report to speak out. 

What are your thoughts about the literary pollution Irene Hames spoke about, and what, if anything, should be done about it?

For the purposes of discussion, I've lifted the requirement for readers to register before they comment on this blog. I've also removed word verification, which some people find annoying. However, I will still moderate all comments to prevent libel, slander, and defamation.

We've had more than 2,500 page views on this blog in less than 2 weeks, but only a few comments.  Now's the time to let the other visitors to Peer Review Report know how you feel and what you think.

BMJ's Godlee: Can We Trust These Results?

Fiona Godlee MD, Editor-in-Chief, BMJ

Just one day after an impassioned plea for more trust among members of the world's biomedical publishing community, one of the world's leading medical editors is wondering whether she can trust some of the information published in her own journal.

Clinicians are asking "Why are these studies appearing in these high impact journals?" Fiona Godlee, MD, editor-in-chief of the BMJ, said in an interview with Peer Review Report during the Seventh International Congress on Peer Review and Biomedical Publication. "Can we trust these results?"

Godlee was referring to what have been variously called 'marketing trials' or 'seeding trials,' namely, clinical trials that appear to be undertaken more with the intent of boosting drug sales than producing useful clinical or scientific knowledge. Results of a ground-breaking qualitative study aimed at hammering out a definition of these trials had just been presented from the podium in front of several hundred medical editors and journal publishers from around the world.

In her plenary presentation, Sara Schroter, PhD, -- Godlee's colleague at BMJ -- detailed the efforts of a small group of elite editor/clinicians from Oxford, Cambridge, and Yale to develop a consensus on the hallmarks of a marketing trial, absent internal company documents revealing intent. The study abstract is available here as a PDF. These hallmarks, Schroter said, include

• strong manufacturer support, 
• many medical centers in a large number of countries, 
• a small number of patients enrolled at each center, 
• a sophisticated statistical analysis paid for by the manufacturer, 
• drugs for common or prevalent diseases. 

The example Schroter showed was a trial for a COPD medication.

"Our findings suggest that a fifth of all drug trials published in the highest impact general medical journals in 2011 were designed primarily for marketing purposes," Schroter told the congress. "This study will highlight characteristics for editors, reviewers, and readers to be aware of when assessing published trials."

Some comments from the floor were highly critical of the  investigators' research approach, saying their arguments were circular, lacked logic, and indefensible absent some evidence of intent. Investigators themselves admitted their efforts to describe marketing trials using statistical methods had failed completely, and that they were all in positions of conflict of interest, as employees of the journals whose publications they were investigating. This conflict of interest was highlighted by Godlee in the followup interview, when she referred to the substantial, but unknown, sums that journals with high impact factors can earn by selling reprints of articles based on clinical study reports.

Other floor comments were more sympathetic. Some speakers encouraged the investigators to continue their work, while others suggested that the responsibility for detecting and preventing unethical or improper trials rests with institutional review boards, not journal editors. 

Howard Bauchner, editor-in-chief, JAMA

Speaking from a floor microphone, Howard Bauchner, MD, editor-in-chief of JAMA and The JAMA Network, told investigators and conferees that he 'struggles' with whether to publish these trials.

Ironically, the previous evening Kay Dickersin of the Johns Hopkins Bloomberg School of Public Health, while giving a lecture designed to enhance the quality and transparency of health research,  had exhorted the same group of listeners to trust and respect each other more, as members of the same biomedical publishing 'ecosystem.'

Tuesday was the final day of the seventh congress. Annette Flanagin of JAMA, congress coordinator, said a decision on where to convene the eighth congress would be made following a survey of attendees. Be sure to follow @BELS_Editors on Twitter for more information and followup on key issues raised in Chicago.

Can an Author be Too Prolific?

Elizabeth 'Liz' Wager

In the publish-or-perish world of academic and industrial research, it's easy to get caught up in the mindset that more publication credits are better than fewer. But is that always true?

Recent revelations of skullduggery -- including plagiarism and fabrication -- among authors writing articles for inclusion in the world's medical literature have induced researchers to take a closer look at some of these purportedly original contributions. One leading investigator in this area is Elizabeth 'Liz' Wager, a self-employed publications consultant from the UK who has become a fixture at quadrennial peer review congresses like the one taking place this week in Chicago.

For this congress, Wager teamed up with a systems analyst and an editor from The Lancet to investigate the seemingly superhuman feats of certain authors. Without naming names (and risking a defamation suit) Wager, Sanjay Singhvi, and Sabine Kleinert presented a study Sunday in which they identified 24 authors whose names were listed on at least one publication every 2 weeks for an entire year. These types of numbers, Wager says, fall well beyond the 99th percentile for the more than 160,000 authors measured, and should send a clear signal to university administrators or journal editors that something is amiss.

THE STUDY

Publications in Medline from a 5-year period (2008 - 2012) across 4 therapeutic areas were included in the study. A custom software package was used to disambiguate authors in PubMed by analyzing characteristics beyond first and last name, such as affiliation, past publication history, and coauthorships. The use of differing therapeutic areas was intended to reduce the chance of double counting publications from authors with names like Wang, Kim, and Smith. The maximum number of publications per year per author was 43, which might be possible if the majority of writings were letters to the editor or abstracts. But one person claimed to have authored 15 clinical trials in one year -- a feat Wager and her coauthors termed 'unfeasibly prolific.'

Wager's close look at prolific authors spawned a bit of light-hearted ribbing among researchers who have presented at past congresses and have become familiar with each others' work. For example, Chris Graf of Australia, editorial director for global research at Wiley, posted the chart below on his Exchanges blog:

For coverage of the meeting's final day, be sure to follow @BELS_Editors on Twitter.

Hames: Medical Literature "Polluted"

Peer Review Report is listening.

Unscrupulous authors have abused the peer review system, leaving it contaminated with articles that fall well short of the normally high standards for scientific publication, according to Irene Hames, PhD, of the Committee on Publication Ethics (COPE). "The literature is polluted," Hames told an audience of several hundred editors and publishers gathered Sunday night at a Chicago hotel.

Hames spoke near the end of a day that offered a succession of well-researched studies documenting the violation of publishing taboos. One study examined fraudulent authorship, one study looked at duplicate publication, one looked at retractions, while 2 looked at plagiarism. All studies were presented as part of the Seventh International Congress on Peer Review and Biomedical Publication, cosponsored by The JAMA Network and BMJ.

As an example of literary pollution, Hames mentioned a 2012 episode involving fake reviewers. Normally, science journal editors ask authors to suggest peer reviewers for their articles. In this case, the authors suggested the names of well-known colleagues, but supplied fictitious email addresses. The email addresses could be monitored by the authors, who intercepted the papers. "They made up the reviewers' names, and they reviewed the papers themselves," Hames explained. Fallout from this episode has spread "across countries, across journals, across disciplines," she continued. "One author has had 28 papers retracted. . . ."

Hames said afterward in an interview that COPE has no enforcement powers with which to punish polluters. However, in an effort to detoxify the literature, Hames unveiled a new and improved database of 500 publication ethics cases that, she says, can be used to provide much-needed training to newly hired life science editors. Every case within the database has been reclassified and recoded. "We've done this so we can characterise cases more comprehensively -- and to aid searching -- and look for trends that can help inform updating of COPE's current guidelines and resources, and the development of new ones," Hames wrote in an earlier email to Peer Review Report.

The 3-day congress continues Monday with 7 studies on bias, 3 on trial registration, and 4 on data sharing and accessibility. The day will conclude with the annual EQUATOR lecture by Kay Dickersin of the Department of Epidemiology at the Johns Hopkins Bloomberg School of Public Health in Baltimore, Maryland, USA. Be sure to follow @BELS_Editors on Twitter for congress updates. All tweets will be aggregated at #prc7.

Raking Muck via PubMed

Peer Review Report will be sniffing.

“Muck-raking with a clean end,” is how Elizabeth “Liz” Wager, a peer review veteran from the UK, describes an investigative project she undertook with coauthors Sanjay Singhvi and Sabine Kleinert for the Seventh International Congress on Peer Review and Biomedical Publication. The intriguing – and fragrant – results of their sleuthing will be presented Sunday morning during a session on authorship moderated by Drummond Rennie of JAMA.

Apparently a private company has developed software that can work in tandem with PubMed, the popular web portal developed by the U.S. government that has become familiar to many life science editors. The software is capable of identifying authors whose names appear frequently in connection with certain therapeutic areas in biomedical publications. Normally, the software is used to help identify key opinion leaders for marketing and promotional purposes. However, it can also be used to expose fraud.

“Too Much of a Good Thing? A Study of Prolific Authors” takes a deep dive into the numbers behind PubMed citations. Further aromatic details are embargoed for release until Sunday. Be sure to follow @BELS_Editors on Twitter for updates.

Liz Wager and Irene Hames Coming Soon

Peer Review Report has been listening.

Peer Review Report will soon feature posts based on exclusive interviews with 2 more thought leaders in the emerging field of peer review research. These should make ideal reading for congress attendees who are en route to Chicago, as well as life science editors who stayed home this weekend and who are looking for an excuse to avoid doing the laundry or cutting the grass.

Elizabeth "Liz" Wager, PhD, a peer review veteran and an authority on biomedical publishing, spoke with Peer Review Report from the UK via telephone on Friday and gave her take on the upcoming congress.

Irene Hames, PhD, another peer review veteran and an authority on publication ethics, sent Peer Review Report a detailed email offering her perspective on the most interesting sessions.

Be sure to follow @BELS_Editors on Twitter to be notified when these posts are published.