Conflicts of Interest: What’s the Harm?

In May 2015, New England Journal of Medicine (NEJM) correspondent Lisa Rosenbaum published a three-part op-ed decrying the state of conflicts of interest policing in health and medicine. Unlike most of the available literature on the topic, she accused those in the medical community concerned with undue industry influence of falling victim to a sad combination of cognitive biases and excessive moralization. Many in health and medicine appropriately rejected Rosenbaum’s argument as a big step back not only for the NEJM, but also for medicine in general. As a publication that was once on the forefront of conflicts of interest protection, Rosenbaum’s op-ed was something of a slap in the face for many who had put their trust in the journal’s editorial board and its supposedly rigorous scrutiny of competing interests.

The second part of Rosenbaum’s argument focuses on a supposed lack of evidence regarding the harms of conflicts of interest. And while she acknowledges that many studies have found clear associations between industry relationships and physician behavior, she ultimately suggests that the effects of such relationships must be tied more closely to documented patient harm if we are to accept recommended limits on conflicts of interest:

Studies seeking evidence of industry influence usually find it, providing us with well-publicized associations. Some 94% of physicians have relationships with industry, though these interactions most often involve activities such as receiving drug samples or food in the workplace. Physicians who request a drug for hospital formulary are more likely than other physicians to have had drug company interactions. Industry-sponsored studies are more likely than government-sponsored ones to have positive results. Physicians who attend symposia funded by pharmaceutical companies subsequently prescribe the featured drugs at a higher rate. All these associations are probably valid. But they don’t answer the key question: Are any of these interactions, or efforts to curtail them, beneficial or harmful to patients?

As I (and many others) have argued, there’s already more than enough evidence to reject Rosenbaum’s arguments. However, if the collected research evidence isn’t enough to persuade you, reports from the U.S. District Court in Washington DC should provide significant cause for concern. Indeed, a recently filed lawsuit alleges that ex-FDA Commissioner Margaret Hamburg failed to disclose over $200 million in conflicts of interest related to her husband’s ties with Johnson & Johnson. Specifically, the suit suggests that,

During the confirmation process before Congress, Dr. Margaret A. Hamburg, acting in concert with her husband, Peter F. Brown, at all material times the Co-CEO of a hedge fund named Renaissance Technologies, L.L.C., failed to disclose to Congress and other relevant authorities, her and her husband’s clear-cut conflict of interest –specifically, that Renaissance Technologies, L.L.C. held hundreds of millions of dollars of Johnson & Johnson stock, the manufacturer of the deadly drug, Levaquin.

Levaquin is the key here, and it’s what specifically addresses Rosenbaum’s demand for clear evidence of patient harm. (Although, I’m not sure that any pile of data would be large enough to appease Rosenbaum.) Despite the fact that the antibiotic and related products have been tied to an estimated 300,000 deaths, the suit alleges that Commissioner Hamburg suppressed mounting evidence of harm.

Certainly, for now, we have only the complaint filed on behalf of the plaintiffs in the case. It remains to be seen how many of the allegations will be born out during discovery and litigation. Nevertheless, if found to be true, this would be entirely in keeping with the kinds of patient harms that are typically documented in cases of industry conflicts of interest. In point of fact, there is significant evidence already that conflicts of interest often result in the suppression of drug side effects and hazards data during testing and evaluation. This has been well documented during high profile cases like the Vioxx and Paxil scandals. However, more comprehensive analyses of industry-funded pharmaceuticals research as well as publications produced by authors with pharmaceuticals conflicts of interest indicate that drug hazards data are systematically suppressed. Closer to home, within the FDA, my own research [paywall, sorry] has documented how increases in the number of conflicts of interest at Oncologic Drugs Advisory Committee meetings correspond to a significant reduction in attention to evaluating drug side effects.

I wish I could be more optimistic that the Hamburg case will be the one that makes Rosenbaum, the NEJM editorial board, and everyone else finally sit up and pay attention. Unfortunately, despite an already large and steadily mounting pile of data on the harms (including patient harms) of conflicts of interest, we’ve seen more in the way of retreats than progress, of late. In 2009, the FDA banned conflicts of interest on its drug advisory committees, only to bring them back a year later. The NEJM set aside significant space in three successive issues to defend the role of drug money in medicine. And despite substantial and known ties to industry, Robert Califf’s confirmation as the new FDA commissioner sailed through the Senate 89 to 4. I suppose, if nothing else, we can take solace in having witnessed a rare act of bipartisanship.