Journal Should Keep Correcting Itself on Conflict of Interest

Nobody wants to show up in a correction.

It either means the publication said something wrong about you or that you were the one who erred. If the correction simply says that you “could not be reached or did not respond,” it leaves the impression that you are hiding from something.

That’s what the American Journal of Bioethics said this month about Drew Edmondson, Steve Passik, Will Rowe, Lisa Robin, and Evan Anderson, all authors on various articles related to pain medicine that were published in the journal’s November 2010 issue. They didn’t respond, the journal says, when they were asked whether they had conflicts of interest related to pain medicine.

No response to many in the public = guilty.

So some of these authors were understandably a little miffed when I contacted them to ask why they had not responded to the journal.

Here’s what Drew Edmondson had to say:

I have not been contacted by the Journal, nor, to my knowledge, has there been any attempt to contact me.  I am not that hard to find.  In November 2010 I was the Attorney General of Oklahoma, but have been in private practice at the firm where you have contacted me since January of 2011.  I expect and believe that I filled out a conflict of interest form at the time of publication but would have reported no conflicts since I had no conflicts.  They appear to be referring to monies received by the Center for Practical Bioethics from one or more pharmaceutical companies.  I did not and do not work for the Center and have received no such funding.  I do not know whether the disclosure process at AJOB differs from any other publications.

Passik had a similar response, which I shared on Friday.

I could not find Will Rowe. He was the CEO for the American Pain Foundation, which shut down shortly after an investigation into its connections to pharmaceutical companies was launched by the U.S. Senate Finance Committee. Rowe wrote a piece in the November 2010 edition of the journal that was critical of the agreements physicians have patients sign requiring a commitment not to abuse painkillers. He has been a vocal opponent of such contracts. In 2010, he told Kevin B. O’Reilly at American Medical News.

Just think about how it's presented — unless you sign this, you won't get your medications. That, to me, is crossing the line. … It's basically a document that says, 'You do this, or I've gotcha.' That's basically interfering with what should be a trusting relationship between the patient and the practitioner.

Rowe certainly would have had conflicts to disclose. ProPublica’s Charles Ornstein and Tracy Weber reported that the American Pain Foundation received 90% of its funding from the pharmaceutical industry. That includes more than $1 million from one drug company, Endo Pharmaceuticals, makers of Percocet and 10 other painkillers. Whether Rowe submitted a disclosure form that was overlooked or ignored remains a mystery. As for reaching him, I agree that he’s a hard guy to find right now.

Lisa Robin and Evan Anderson were not so hard to find, but they did not respond.

Robin is the chief advocacy officer at the Federation of State Medical Boards Foundation. Anderson is a senior legal fellow at Temple University Center for Health Law, Policy and Practice. Both were co-authors on “A Rose By Any Other Name,” an article about pain contracts that I discussed last week. It’s unclear to me whether either of them have any conflicts that should have been disclosed.

There were at least two other articles from November 2010 that had undisclosed conflicts.

In “The Patient-Centered Opioid Treatment Agreement,” Seddon Savage wrote:

Multidimensional approaches are clearly needed to reduce opioid-associated harm while maintaining their clinical availability. Provider education and practice improvement initiatives; patient and public education regarding safe use, storage, and disposal of opioids, and ongoing research to improve drug safety profiles and identify best practices are needed.

Hard to argue with that. Now the American Journal of Bioethics says that Savage, who works at Dartmouth, had prior ties to a pharmaceutical company and a company that makes drug screening tests:

Seddon Savage reported that she did not “have any relationships with industry at the time of the publication of this series of essays, November 2010.” Prior to the publication of this article, Seddon disclosed that “Up to May 2010, I served on an Ameritox (make urine drug screens) advisory board and on the Media Advisory Board for Risk Evaluation and Mitigation Strategies (REMS) for Onsolis.”

Then there is “The Opioid Emperor Has No Clothes,” in which Michael Keane compared withholding opioids from patients to torture:

Secondly, is there inordinate harm to any individual who might “abuse” opioids? Ignoring harm that comes from prohibition, what is the actual harm from the pharmacological action? From a functional perspective there is surprisingly little.

I’m sure that the family of Mark Allen Diggs in West Virginia would argue that point. Diggs jumped a median while driving high on oxycodone, hit a school bus, killed himself and injured 11 kids. The families of the kids would probably have a good argument, too.

Now the bioethics journal says that it should have made Keane’s ties to a drugmaker clear:

Michael Keane disclosed a financial interest in Biotech Pharmaceuticals, an Australian company developing an immediate release oral morphine product for acute pain.

Savage never responded to my questions, but Keane kindly did. He wrote:

The information that was in the recent correction was information that I provided BEFORE publication as part of the standard COI disclosure. As part of transparency I think it is important for that to be on record. As I can remember, I received an email from the Journal sometime around April or May this year. It was a very polite and matter of fact email. As one of the authors from that edition, they gave me the courtesy of asking me whether I had anything to disclose. I had already disclosed my COI prior to publication and I thus reaffirmed that COI.

So Keane did his job. He appears to fall into the category of disclosures that were overlooked by the journal, as co-editor-in-chief David Magnus explained to me last week.

Keane also had a response to my contrasting his work with Diggs’ violent death.

There is data about driving performance and opioid use. The question of whether we should limit people from receiving pain treatment in order to prevent ANY reckless use of opioids is a very vexing philosophical question. (There are people on the other side of the equation; that is, the impediments to pain relief for chronic pain can become prohibitive, especially if they don't "pain properly" as one person put it to me). I suspect they, and their families, would have something to say as well.

To sum up, let’s tally the various ties to pharmaceutical companies that went undisclosed by the journal in 2010 that are now being shown the light of day:

King Pharmaceuticals; Pfizer Pharmaceuticals; PriCara (part of Janssen Pharmaceuticals, which is owned by Johnson& Johnson); Purdue Pharma; TEVA Neuroscience; Genentech; Allergan; Endo Pharmaceuticals; NeurogesX; Cephalon; Covidian; Meda Pharmaceuticals; Biotech Pharmaceuticals; and Onsolis.

That’s 14.

If you add in the disclosures that Passik says he would have made, and that Rowe should have made, you would add:

Ligand Pharmaceuticals; Eli Lilly and Company; McNeil PPC; Merck & Co.; Medtronic; Forest Laboratories; Abbott Laboratories; Boston Scientific; St. Jude Medical; and Recro Pharma.

That brings the tally to at least 24 conflicts that definitely should have been disclosed by the journal in articles about pain management. There were 23 authors involved in the November 2010 issue on pain medicine. Of those, at least eight had conflicts that were not published in the journal.

But let’s be clear. The journal is correcting the record. It is coming forward with these disclosures even in a climate where many critics of the journal are ready to pounce on any perceived ethical slip. (A journal dedicated to ethics kind of sets itself up for this scrutiny.) So, while there are still some unanswered questions surrounding how the disclosure process has been working – or not – at the ethics journal, we at least are seeing signs that the American Journal of Bioethics wants to move forward and make things right.

Nobody at the journal is asking my opinion, but here are my three recommendations. First, make the conflict of interest policy clear by posting it online and outside of the paywall. Second, post a list of the editorial board members and the conflict of interest committee members. And third, don’t stop with the November 2010 issue. Make that same review of conflict of interest disclosures for all prior issues, and if it seems like too much work, ask for volunteers.

Image by Derek Gavey via Flickr