Commercial Influence on Research & on Publishing

Here's an interesting story about commercial influence on the world of scientific publishing. In a recent article in Harvard's Health Policy Review (HHPR), Donald W. Light and Rebecca N. Warburton detail the difficulties they had in trying to publish, in the Journal of Health Economics (JHE), a paper critical of another paper (by DiMasi et al) previously published by JHE.

Back in 2003, the JHE published a paper by DiMasi et al on the R&D costs involved in drug development (on average, $802 million). There's plenty of skepticism about that number. Light & Warburton argue, in particular (and among other criticisms) that that number was based on data that was not made public, and that the source of the funder for DiMasi et al's research was unclear.

In their HHPR paper, Light & Warburton detail the difficulties they had in dealing with the editors of JHE, who, it seemed to them (and they provide evidence of this in their paper) went to great lengths to soften criticism of the DiMasi et al paper, and in particular to eliminate any implication of commercial (i.e., industry) influence on the research. It's fascinating, thought-provoking reading.

Two important notes:
1) Light & Warburton don't provide direct evidence of commercial influence, either in the paper they wrote for JHE, or in their critical review of the JHE's process in the HHPR. But they do ask some very good questions, questions deserving of an answer. Direct proof of such things is hard to come by, but it behooves researchers and scholarly journals to do what they can to be sufficiently transparent to dispel any doubts.
2) There are no attacks, here, on anyone's character. Influence -- including corporate influence -- can happen in many subtle ways. To imply influence is not to imply malice.

---

Here's the website for the Health Policy Review.
The direct link to the article is here: Ethical Standards for Healthcare Journal Editors: A Case Report and Recommendations
Here's the 2003 paper by DiMasi et al: The price of innovation: new estimates of drug development costs

---

See also:

and

Pharmas Cap Money for Docs & Profs

From the Financial Times: GSK to publish level of doctors’ advisory fees

GlaxoSmithKline is to make public the level of advisory fees it offers to doctors and medical academics, and will strictly cap the payments they can receive in the US to $150,000 (£88,000) a year each.

Andrew Witty, chief executive of the UK-based pharmaceutical company, said he was introducing tougher new rules to impose a cap “without exception” on such payments and promised to publish the amounts.

His commitment comes at a time of growing concern that the widespread practice of payments by pharmaceutical companies may help unfairly influence “key opinion leaders” in the medical community, in a way that biases their judgments and recommendations for particular treatments.

Eli Lilly and Merck made similar moves last month.

It's worth reminding ourselves what this is all about: fear of conflict of interest. A conflict of interest is “a situation in which a person has a private or personal interest sufficient to appear to influence the objective exercise of his or her official duties as, say, a public official, an employee, or a professional.” (MacDonald et al “Charitable Conflicts of Interest.”  Journal of Business Ethics 39:1-2, 67-74, August 2002). The fear here is that pharma payments to docs & profs could be sufficient to bias, or to appear to bias, the prescribing habits of docs and the research done by profs. Part of the plan being announced here is not just a limit on payment but a process for making payments public. Disclosure is one of the "standard" mechanisms for dealing with conflict of interest.

What's the goal of disclosure? That's debatable. There's evidence that disclosure doesn't affect people's behaviour, at least not for the better. (See Cain, et al “Coming Clean but Playing Dirtier: The Shortcomings of Disclosure as a Solution to Conflicts of Interest”, in Moore et al, eds. in Conflict of Interest: Challenges and Solutions in Business, Law, Medicine and Public Policy. 2005) Patients may not know what to do in the face of a disclosure that their physician gets money from a drug company, and doctors may be more cavalier about their conflicts, which after all they've disclosed. On the other hand, ethics isn't all about direct outcomes. Regardless of whether disclosure is effective at achieving some goal, it may well be that patients and others have a right to that information.

---
Hat tip to the indispensable Ed Silverman @ Pharmalot.

Want to find out what makes you "stand out from the crowd"? Consider signing up for a research study!

Personalized genomics is a hot topic. Last week, an event in Toronto, Personalized Medicine: Are we ready? sponsored by the Canadian Institute for Advanced Research (CIFAR), had academics and industry representatives debating the issues in a public forum.

One company represented was 23andMe, a personalized genomics company from Silicon Valley, whose slogan is "genetics just got personal". I decided to check out their website further after finding a short article from earlier this year, in the May 2008 issue of Science, talking about how 23andMe are conducting research on Parkinson's with subjects who will participate in trials from the comfort of their homes, doing directed self assessments on motor abilities and tremors from their own computer. While this issue alone raises huge concerns about research validity and rigor, what I found on the website for 23andMe was more worrisome.

23andMe have a page on their website dedicated to, basically, recruiting people to take part in research. It is, for all intents and purposes, a recruitment flyer. And one, that I'd guess, would never be approved by most ethics review boards. Interestingly, they call the research area of their website, 23andWe, a very clever way of manipulating their slogan to imply collaboration. They promise direct benefits to individuals who choose to "leverage their data" by contributing it to research. These include:

"Direct research by participating in studies of conditions and traits you care about.
Learn new things about yourself - and what your genes may have to do with them.
Find out which traits make you stand out from the crowd."

They do also note that participating can lead to societal benefits by "discovering new genetic associations" and "improving health care for everyone"- which is, while very laudable, perhaps overstated in this context.

Not many research ethics boards would approve a recruitment flyer that promises the kind of direct benefits that 23andMe is promising (I doubt most "participants" can, as stated, actually "direct" research projects about what is important to them), overstates the potential societal benefits and makes no mention of potential for risk of harm, while never telling you exactly what you have to do to take part in their research (Do you have to provide extra samples? Can you decide how your samples are used?) or what tangibles, if any, you might actually get from participating (Do you get an extra report? Can you see the research results?). True, it's not a consent form - but we can agree that a recruitment flyer should provide basic information that is complete and transparent.

Should they just simply have asked: Would you like to allow us to do research on the sample you have provided for analysis? That would be more forthright. Far less sexy, perhaps. But as research ethicists, we're not concerned about ensuring that recruitment flyers are sexy. In fact, we're about ensuring that they're not - so that they're not more likely to induce highly suggestive potential target participants (like young enthusiastic students on Facebook who can become fans of 23andMe to take part in the company's research just because it sounds cool and will help them "stand out from the crowd".

Methodological Rigour as Ethical Requirement

Here's a good example of why those assessing clinical trials have an ethical obligation to care not just about protecting research subjects, but also about methodological rigour.

The FDA has just approved a process that uses magnetic pulses to treat depression. It's the first time noninvasive brain stimulator has been used to treat depression. It's a neat trick.

It's expensive, and doesn't work for most subjects, but it does have some effect. It's a promising addition to the range of tools available to treat depression. The FDA has asked for more data, so another trial is now recruiting. I'm not qualified to judge the design of this trial, but I know the details matter. Scientific rigour is all that stands between this magnetism-based form of therapy and the charlatans selling magnetic bracelets to cure any- and everything. The public needs to know there's a difference, and they need to know that both researchers and research ethics boards are working hard to maintain the difference.

Pfizer accused of manipulating publication of drug info

Reports from an expert panel examining how the pharmaceutical giant, Pfizer, has controlled the flow of research data on the epilepsy drug Neurontin (aka Gabapentin), posit that the company actively and purposefully manipulated the publication of clinical data in a number of ways, including:

"spinning negative data to place it in a more positive light, and bundling negative findings with positive studies to neutralize the results"

Essentially, what Pfizer did, according to the panel of experts, was control how information about the drug Neurontin was released, in order to convince prescribers that the drug was valuable for unapproved or "off-label" use for conditions such as treatment of bipolar disorder, pain control and prevention of migraine headaches. Many drugs are recommended for "off label" uses, which means use for indications not explicitly noted on the label or the monograph.

Neurontin, essentially a drug for adjunct use in the control of seizure disorders, has been prescribed for the treatment of neuopathic pain, migraine headaches and bipolar disorder. These "off-label" uses may have been encouraged through the publication strategy that the expert panel feels the company adopted. The drug's use as a treatment for bipolar disoder is contrversial at best. Additionally, there are studies, as noted in the article, that provide evidence that Neorontin is not an effective pain relief agent for common types of neuropathic pain. However these findings, it appears that Pfizer may have worked to delay publication of such data.

In the meantime, according to this article, Pfizer managed to accelerate sales of Neurontin to 3 billion USD annually, up until 2004.

Some might say, well, Pfizer has the right to control dissemination of findings as they wish. In this case, some might even go on to say, these studies were not focused on intended use but on "off-label" uses so what harm could result from suppression of negative data regarding something that no one is approved to do anyway?

Those who review clinical studies for ethical soundness usually ask researchers questions about how the results of their research will be disseminated. However there is little attention in the ethical review process to how and when data will be publicized. Many studies that are approved by research ethics boards (REBs) never end up being published in an academic journal, nor publicized in any other way. Alternatively, many studies that are approved end up being publicized beyond the initial expectations or aims of the researcher - as a result of unexpected, compelling or particularly useful findings across a number of contexts. Rarely, if ever, is an REB contacted by a researcher to ask if it is acceptable to wait to publish findings for a few years, or to publish data in a way that the researcher hadn't indicated during the prospective ethics review.

In the Neurontin case, however, the suppression and manipulation of data, as highlighted by the expert panel, which included the suppression of 8 out of 16 negative studies and the abbreviation of 2 additional negative studies, appears to be less of an incidental occurrence than a marketing strategy to, essentially, ensure continual "off-label" use and thereby increase profits. While Pfizer claims that it presented study results that are "accurate, balanced and complete", it appears that they may have strategized to hold back negative information that would, for most prescribers, compel them to question the use of Neurontin in a number of clinical situations.

Research ethics boards are constantly examining and renegotiating their jurisdiction. The question that this story leads to is: Should REBs also be involved in decisions about dissemination of data? Surely such an expansion of REB authority would meet significant resistance from researchers, and the mechanisms by which such oversight would operate are far from clear. But the Neurontin/Pfizer case demonstrates why the question is bound to keep arising.

Children's Cold Remedies: the Ethics of Not Doing Research

From Reuters: Don't use cold drugs in kids under 4, industry says

Oral cough and cold medicines sold over the counter should not be used in children younger than 4 years old because of the risk of rare complications linked to inappropriate use, manufacturers said on Tuesday.

The move seems a good one. Maybe it's step 1 towards action on the real issue:

Currently, the medicines are available under decades-old FDA rules that allow over-the-counter products to be sold without clinical trials showing their risks and benefits.

"The bottom line remains the same: that these products have never been proven to work in children," said Diana Zuckerman, president of the National Research Center for Women & Families.

Two quick, relevant, rules of thumb for business ethics, rules that seem entirely relevant here:
1) Information asymmetry matters. If customers don't understand what they're getting, they're not making fully voluntary purchases, and even the most zealous fans of free markets say that markets don't work properly in the absence of reliable information. Parents almost certainly do not know that the safety & efficacy of these products have not necessarily been proven in clinical trials.
2) All the standards get cranked up a notch where kids are concerned.

Double arm transplant: innovation or research?

This July, in Munich, the world’s first double arm transplant was successfully carried out on a Karl Merk, a farmer who lost both arms six years ago in a work related accident. While it's clear that the patient is making excellent progress three months later , it does raise some pretty compelling issues around what constitutes “research” versus “innovation”.

Transplant, as a surgical procedure, is no longer an innovation. The first kidney transplant was carried out in 1954 and the first heart was transplanted in 1967. So that makes transplant, as a surgical procedure, a pretty commonplace approach to a number of diseases and conditions.

But in even the most commonplace surgeries, innovations have to happen. Why? Well, first and foremost, they have to happen in order to further science, save more lives and make surgery easier, shorter and more efficient. However the burning issue is not why innovations have to happen but how they happen. Often, innovative approaches to common surgical procedures happen "on the fly" in the operating room, out of urgent need. Surgical teams have to make "on the spot" decisions if someone's anatomy is not as advertised, for example, or someone's physiology simply will not work with the planned surgical procedure. At times like this, there might be time to leave the operating room, approach a waiting family and get a quick consent from the designated substitute decision maker or proxy to change the procedure. In an emergency situation, however, sometimes innovations are the only way to save someone's life quickly and a new approach to an old problem is carried out, hopefully letting the patient know later on what went on while they were on the operating room table. Key life saving surgical procedures have been developed this way - as a "one off" at first, then afterwards, developed as a research protocol, studied and evaluated through a formal clinical trial.

Obviously, the double arm transplant in Munich was not an "on the fly" innovation - it was, it seems, a well planned and carefully executed surgical procedure that likely, underwent peer and ethics review. (Note that there is a standard treatment for tragic loss of limbs - the fitting of working prostheses - a treatment that is not without its problems, challenges and ongoing discomfort for patients). But it's important to realize that while the lines between innovation and standard treatment are quite clear, the lines dividing innovation and research are less well delineated and often involve significant risk of harm alongside yes, great potential for benefit. In Mr. Merk's case, two arms are definitely better than none. So the risk of having two non-prosthetic, fully human arms attached and then undergoing months of physiotherapy and years of anti-rejection therapy will likely be worth the potential benefit. However, it should be very clear to everyone, including and hopefully most of all, Karl Merk, that this innovative approach is still "research".

Therapeutic Misconception, Personalized Genetics, & Breast Cancer Research

Two of the hottest tech-related topics in the news today are genome sequencing and social networks. This story brings the two together.

According to this press release, the controversial personalized genetics company, 23andMe wants to build what is essentially a social network of women united by their experiences (first- or second-hand) with breast cancer.

Summer Johnson over at the excellent "blog.bioethics.net" worries that 23andMe isn't doing enough to avoid the 'therapeutic misconception' i.e., the misconception, common among participants in research, that the research they're participating in will somehow help them (which is typically false) rather than simply advancing knowledge in a way that may help people like them in the future.

According to Johnson:

What the company doesn't say to these women, though, is that this reality is certainly years if not decades away. The real benefits of this "research arm" of 23andMe is most certainly the "community" that this group provides--and these benefits can be quite real. The sense that there is a whole world of women out there going through what these women are, that can share information and experience, is a real benefit of this research--but telling these women that participation in this RESEARCH, not treatment, is nothing more than that.

The benefits of personalized genetics/genomics are certainly easily misunderstood. But it's not clear there's much risk of therapeutic misconception, here. Admittedly, the press release never states specifically that treatments are years or decades away, but it never implies they're closer than that, either, and there's an awful lot of very fair talk about community-building. In fact, although lots of studies promise the vague benefit of "helping others" and "being part of a community," the way 23andMe is approaching breast cancer research actually does stand to benefit participants (though clearly not by curing their breast cancer). In particular, participation in an actual social network could help individuals who take part by allowing them a place and space to articulate their experience, and to take part in a social network with empathetic or sympathetic persons who share their experiences and who can provide not only relevant but meaningful support that may otherwise be lacking that in their lives.

Canadian research ethics: Caught in a quagmire of privacy legislation?

"Trend to privacy seen as hurting research"

This article, from the National Post (September 24) revisits, again, the cry from researchers that overzealous privacy legislation, at both federal and provincial levels,  are preventing researchers from accessing individuals and databases. This quick overview of an article originally in the Canadian Journal of Public Health (July- August 2008, A. Harris et al) illuminates two important points that researchers and  REBs continue to agonize over. 

First, privacy legislation is making it difficult, if not impossible to access data that previously was available to researchers, even anonymized data stripped of all possible identifiers.

Second, and most importantly, the privacy legislation is difficult for many to understand and therefore, is applied and interpreted very differently by different agencies, data stewards and, I would add, REBs and researchers, who are obligated to understand the legislation as applied to specific situations, agencies, information across a variety of contexts.

True enough, the legislation is more than difficult for even smart people to interpret. The "basic reader" on Canadian privacy laws published by the Canadian government that I keep on my desk is thick and heavy with enough very small print to make it a daunting task to try to find an answer to the simplest question. It makes a great paperweight but is not overly helpful in my own interpretation and application of the Acts. 

According to the Canadian government, Canada has the most comprehensive privacy legislation in the world. Few would argue with that. At a federal level, there is PIPEDA and the Privacy Act. In addition, provinces and territories have their own privacy legislation statutes, and even the most educated persons are not always quite sure whether federal legislation trumps provincial or vice versa, on an ad hoc basis. 

Ontario, interestingly enough, has both PHIPA (The Personal Health Information Protection Act) and FIPPA (The Freedom of Information and Protection of Privacy Act). PHIPA allows for the protection of the privacy of persons and the confidentiality of their personal health information. FIPPA is designed in part to protect individuals' identifiable information held by institutions, guide institutions on how to handle information and allow individuals access to the information that an institution holds about them. 

Most REBs in Ontario are involved with applying both PHIPA and FIPPA when reviewing researchers' plans on accessing personal data and health information. Most researchers have to work within the constraints of FIPPA and/or PHIPA to access individuals and information. But it is a laborious task for all concerned. And many people aren't even sure which act is which if they aren't engaged with privacy issues on a daily basis. 

I was also amused to find that while everyone refers to FIPPA phonetically, some also pronounce the acronym for the Personal Health Information Act (PHIPA) exactly the same, correctly pronouncing the "ph" as a "f". Yet more confusion, no doubt, probably quite unknowingly, in many conversations.... 

While there are certainly clear distinctions between the Privacy Act, PIPEDA, PHIPA, FIPPA and other provincial statutes, it remains that the legislation created with the laudable aim of protecting Canadian citizens is unclear, the separate Acts can be difficult for both lay persons and academics to untangle and many institutions and agencies are either not applying the Acts at all, or interpreting and thereby applying them in an incorrect way. The valuable objectives and good intentions of protecting private information and those most vulnerable to a breach in privacy are obscured against a backdrop of confusion and a lack of guidance and clarity by governing bodies.