Physician Interpretation of Industry Funding

According to this article in the current New England Journal of Medicine, physicians are more apt to prescribe drugs supported by studies they perceive as rigorous, than those supported by studies perceived as less rigorous. More interestingly, though, physicians downgrade their perception of the rigor of a study, their confidence in its results, and their willingness to prescribe the studied drug, when the study discloses that it was industry-sponsored. Physicians surveyed were twice as willing to prescribe a drug supported by an NIH study than one supported by an industry-sponsored study.

In general I greet this as good news. Industry sponsorship injects bias into the literature in many ways: via unconscious biasing of sponsored researchers, via careful selection of the kinds of studies sponsored (e.g., avoidance of head-to-head drug comparisons that have any chance of turning out badly), and even (in a few documented cases) via the burial or abandonment of sponsored studies that don't work out. So it's fine if physicians greet industry-sponsored trials with some skepticism.

Two worries, though: first, there can be too much of a good thing. After all, if a company develops a drug that really works, it will want to sponsor studies to prove that fact; and too much skepticism of positive results could end up hurting patients. Second, if industry sponsorship gives rise to skepticism, then there's an incentive to bury the fact of industry sponsorship via ghost authorship or other means.

Study registers, publication of negative results, rigorous authorship standards, rules for disclosure of funding sources and of conflicting interests, public funding for independent research: these can all help.

But the facts are 1) that we rely on private, profit-motivated research to find our new drugs; 2) that such research has historically done an excellent job in discovering and publicizing useful new drugs; and 3) that the very same motives that drive the industry to do a good job in discovering and publicizing useful new drugs are those that drive it to pollute and bias the literature. So there will be no easy answers.

Doctor's Conflict of Interest

This is a wonderful, and subtle, discussion of what it was like for an MS patient to discover that her treating physician had long been on the payroll of the pharmaceutical firm which made the drugs he'd been asking her to take.

Conflict of Interest on NIH Guideline Panels

Here's a New York Times story from last week, reporting a very high prevalence of financial ties to industry among experts on NIH panels drafting guidelines on hypertension, obesity and cholesterol. (As the article notes, the panels were seated in 2008, before the current more rigorous guidelines on COI were put into place.) 20 members of the three panels, including some co-chairs, have been told they ought to recuse themselves from voting on crucial issues. 8 of 19 obesity panel members have ties to industry, as do 7 of 16 cholesterol panel members. Worryingly, 5 of 17 hypertension panel members have taken industry money since the panel was seated.

And government panels are among the least worrying on the COI issue. This piece from BMJ last month finds that, among 7 panels working on diabetes and 7 on hyperlipidaemia guidelines between 2000 and 2010, most panel members and half of all panel chairs had COI, and that the COI prevalence was lowest on government sponsored panels. It was highest on panels sponsored by specialty societies. It's time for a hard look, not only at standards but at remedies. Mere disclosure and self-recusal from voting doesn't help much, if the panelist has had the chance to frame the discussion. One panelist in the Times story is quoted as saying he severed his ties with industry in order to serve--but if he had the ties before, and will be free to have them again afterwards, what does the current severing really do to secure his independence? The best solution is probably the one used by NICE in the UK: no one who's received industry money in the last year is permitted to serve.