In my last post I raised the possibility that Steve Nissen, a highly influential cardiologist who has been an outspoken critic of industry influence in medicine, might have his own conflict of interest (COI) problem. In response, another cardiologist, James Stein, said that my post was unfair in its treatment of Nissen and failed to consider important distinctions and subtleties in the academic literature about conflict of interest.
Let me first of all confess that Stein– for whom I have the highest possible regard– makes some excellent points. And I further admit that my piece pretty much ignored some of these fine academic distinctions. But I also think it’s likely that we may have a forest and trees situation here and that by focusing on subtle COI distinctions it is possible to lose sight of the larger issue.
A Simple Test
Rather than focus on the subtle distinctions between different types of COI I’d like to propose a much simpler way to think about this problem as it exists in the real world. Here’s the test I would propose: can you imagine Nissen, or indeed anyone in a similar situation (running a large multimillion dollar trial of a drug) publicly saying something decidedly negative or critical about the drug?
Last week Nissen told NBC nightly news that “these drugs are breakthrough drugs, they are blockbuster drugs that are very likely going to have a big impact.” What is the likelihood that someone in his position would publicly state that there is a good chance that the drugs won’t be a breakthrough, that maybe they won’t be blockbusters, or that they shouldn’t be approved right away or broadly used?
The answer to this question is clear. I can’t recall ever seeing the leader of an ongoing trial express major doubts about the value of a drug or publicly differ with the sponsor about its potential significance. (And yet, somehow, so many of these promising breakthrough products have failed utterly.)
To be clear: this doesn’t mean that the opinions of trial leaders should be ignored or that they have no value. But it does mean that the audience for these opinions should be fully aware of the full context of the remarks, and that additional sources should be sought.
I think this could be a useful general test to evaluate the problem of COI: Given specific circumstances what is the likelihood that a perspective opposed to a financial supporter will receive fair play? This test could be used to assess the independence of news sources, teachers, guideline writers, regulators, and many others.
As an example, consider commercially supported continuing medical education (CME), another area where the problem of COI has been hotly debated. Supporters of commercial CME point to elaborate safeguards designed to insure the integrity of their programs. Many of us believe these safeguards are just window dressing or, less politely, lipstick on a pig. But instead of debating the details of the CME “safeguards,” let’s just look at the final product. Consider the hundreds of CME programs funded by Sanofi and Amgen in the last few years on the PCSK9 inhibitors — the same drugs that sparked this discussion. These programs have relentlessly promoted the need for and the importance of these drugs. Far less emphasis has been placed on the many caveats and concerns raised by FDA reviewers and the independent panel advisors. People following last week’s FDA panel would have been surprised by these issues if they had only relied on information derived from the ubiquitous commercial CME of the past few years.
Of course we will never completely remove COIs– financial or other– from medicine. But a first step to restore integrity is to make sure the system permits and encourages a wide diversity of views, including those without sponsorship.
Hi Larry – “lipstick on a pig” indeed. Here’s yet another way to think about this problem as it exists in the real world:
If referees, umpires and game officials in professional sports admitted that they regularly take money from team owners (or anybody else with a clear stake in the results of a game), the financial conflicts of interest would immediately render those on the take permanently unemployable in their chosen fields. But in medicine, we call people who take cash from drug or medical device companies “thought leaders” or “key opinion leaders” or “experts” – and then quote them in the media.
How does this make any sense?
Great comment! I love this.
Superb analogy!
Larry, let’s be honest about the pivot you just executed. The “lipstick on a pig” analogy might hold for some industry-sponsored CME, but not for a comment made by a trialist about a drug or trial. You pivoted from a much more obvious and well-documented COI to one that is less straightforward and based on your conjecture. Yes, there may be some pressure to be upbeat about a research project one is involved in or funded for (that is not unique to industry-sponsored trials, by the way), but it is not a sin along the lines of other COIs discussed in this forum previously. And yes, Larry, I have heard trialists speak negatively of study drugs when warranted. I even heard Dr. Nissen speak critically of a drug that he investigated – see http://www.ncbi.nlm.nih.gov/pubmed/16554527. Though the potential for COI from trialists is an interesting topic to discuss, it is not new territory and your tieing it to Steve Nissen is unfair.
James, I think it is very clear in my post that the pig and lipstick analogy applied only to my discussion of CME. Beyond that I’m not entirely sure exactly how and where we disagree, but let me address the question about the independence of trial leaders and the need for them to stand above the fray. After the last exchange I received several emails from readers with a common theme. They all wondered whether equipoise was still possible after Nissen’s public comments that the PCSK9 inhibitors were “blockbuster drugs” and “breakthrough drugs.” During the panel hearings Sidney Wolfe suggested that the ongoing cardiovascular outcomes trials with the drugs might be compromised by their approval, since subjects in the trial might well be tempted to start taking open label drugs if they are persuaded of their benefits. Nissen’s remarks, along with the remarks of other enthusiasts, can only add to this potential problem. Nissen’s role here is particularly ironic, since he has been extremely vocal in insisting on rigorous standards of integrity for clinical trials, most recently in the debacle surrounding the debacle of the cardiovascular outcomes trial with the diet drug Contrave. “High levels of patient retention are essential to maintain the integrity of randomization,” Nissen himself wrote in JAMA a few years ago, referring to yet another trial.
The phrase, “as an example” is hardly a connecting phrase that separates two points. To my read, the two discussion ran together. But thanks for clarifying. In regard to equipoise, I have a few points. First, Steve Nissen is not as powerful as some commentators like to think. One person or comment can’t remove “equipoise.” If he was that powerful he’d cancel his own study as would all the sponsors and just hope the FDA approves the drugs based on lab values. That concern is just noise. Indeed, though, as new drugs become standard care, conducting research becomes more difficult for many responses related to funding, recruitment, ethics, etc. But “remarks” are not major players in that phenomenon, at least compared to guidelines and the results of early trials. Indeed, are we not to comment on any trials because they might impact recruitment for future trials? Second, the media publishes quotes that are juicy so they are as much to blame if not more so. If everyone interviewed said “This is promising but we need more research,” there would be little interest in getting quotes. The media – indeed health care professionals and the public – want people to go out on a limb and make an assessment. Where we differ is in the focus of the story. You made it about Dr. Nissen. I would have preferred it be about the story. Sidney Wolfe’s comments are well taken but in his world view, nothing would be approved without hard outcomes. Though surrogates have limitations though also have an important role. Outside our little chamber of trialists, regulators, and commentators, there are real doctors and real patients who care what experts think.
How do you determine who is an “expert”? Does anyone who runs a clinical trial become an “expert”?