May 29th, 2015
May 29, 2015
By Vikas Saini, MD
The press of several deadlines at the same time makes it difficult to respond adequately to the three-part series in the NEJM on conflicts of interest, but others have done a good job here, here and here.
The author, Lisa Rosenbaum, wrote a thoughtful piece on breast cancer screening that I admired because it captured the difficulty and uncertainty of making individualized, patient-centered clinical decisions in the face of the data, which is by its nature a statistical aggregate. So I was surprised to discover that her understanding of the Conflicts of Interest issue in medicine is so shallow.
I don’t have the time right now to do a proper line-by-line review of the rhetorical devices, the biggest one being her vigorous attack on straw men, that Rosenbaum uses to make a weak case. So a few impressionistic reactions will have to suffice.
Firstly, she does not seem to understand the concept of bias and how bias works. Bias is not providing a service in exchange for money. We have other words for that. Secondly, she doesn’t make much of a distinction between the process of production of new technology and medical knowledge (“innovation”) and the furious marketing efforts of approved products being sold. There are serious concerns and reasonable remedies for both, but they are very different.
Bias is the blindness to your own typos and elisions when you’ve been revising your document a hundred times. You are so close to it, you can’t see. You need an independent proofreader with fresh eyes. It happens to all of us – it’s in the nature of human cognition when we are thinking fast, because in addition to failing to connect the dots, we often connect dots that aren’t there. Moreover, humans are eusocial creatures, and relationships matter to us. The whole point of marketing efforts is to become a trusted friend of the customer. The WHOLE POINT of rubbing shoulders is to soften critical thinking – “benefit of the doubt” — and get what you want out of the relationship.
For an issue as deeply important to the public interest as innovation and the development of new drugs and devices, we need a much more robust understanding of how bias works and how we can minimize it than this series in NEJM manages.
To my pharmascold friends, I have a confession: I, like Rosenbaum, thought that your deep suspicion of industry studies was an over-the-top emotional reaction, and a knee-jerk anti-corporate bias. Until I starting reading the studies carefully. I mean really carefully – poring over the tables in the appendix, and calculating rates that weren’t presented but were discernible in the data. At which point my jaw dropped. Really – I couldn’t believe it at first. But if you posit that all our colleagues in industry and those in academia working for industry are good people, trying to do good things, you can only conclude, as I did, that there is a systematic way in which everyone wants to put their best foot forward, and where you sit in the system matters. You don’t have to assume malfeasance to see that the distortions that emerge can have serious consequences for the public.
So the claim, which Rosenbaum repeats, that industry and physicians are “all on the same side” in wanting what’s best for the patient doesn’t offer any insight into this problem.
I’ve been a principal in a number of start-ups. I have helped find academic clinicians willing to work with industry on products we were developing. Obviously, you can’t have innovation unless you are able to assess ideas in the real world. After reading Rosenbaum’s piece, I have to ask: Did someone ban industry-academic sponsored research while I wasn’t looking?
There’s no guarantee that because a company gets funded, its idea is good or its product will work. That’s a no brainer, right? Right. But once funded, there’s a huge drive on the part of everyone to figure out how to make it work, how to be successful, how to make money on the investment, no matter any serious limitations that might emerge. That injects a huge amount of bias in thinking about the design of a clinical trial, choosing the outcomes, interpreting the results and spinning the result in marketing drives. Do you think VCs could invest millions of dollars in a company with a crap product, do you think that such a company could find academic collaborators to do clinicals and find a reputable journal to publish them, get limited indications approved, start selling product and go public? You bet.
I’m sorry that Rosenbaum felt a twinge of social anxiety at the thought of consulting for industry, but she is at the start of her career, and there will no doubt be opportunities for her in the future, especially with the NEJM on her resume. But working for industry while having at the same time a regular public platform like the NEJM would have simply put the NEJM’s independence and integrity into question. Rosenbaum’s naivete about these matters is perhaps forgivable, but the willingness of the editors of NEJM and the Massachusetts Medical Society to give so much space to such an unsophisticated effort raises deeper, more troubling questions about their attitudes and their future role in American healthcare.
It’s been interesting to see the cheers of support for her series in the comments at NEJM and on Twitter. Clearly, there’s a belief system out there that in the healthcare sector industry, academia, and the public are all on the same side. Well, that belief system drove the repeal of Glass-Steagal and the subsequent poor regulation of the financial services industry and how did that work out for us? Clearly, there’s a lot of angst and resentment out there in the clinician community about being scolded for taking industry money. In a time of tumult in healthcare with so many changes, so many old business models under threat, and so much revenue at risk, that’s understandable. A discussion of that will have to wait for another day.