How should drug companies and physicians interact? And if something is wrong with the relationship, who's to blame?

Some of the best thinking I've seen on that important issue comes in the British Medical Journal, which devoted most of its May 31 issue to teasing out the complexities of how for-profit drug companies alter medicine's playing field. It's a project that could have fallen into pharma bashing-and some individual pieces come close. But it's saved by a willingness to look past the surface to figure out how the pharma-medical system really works and why.

"This relationship isn't a Manichean battle between good and evil," writes deputy editor Jane Smith, "but the entwinement of individuals from different backgrounds and value sets who get to know, and often to like each other and therefore want, as humans do, to reciprocate friendships and favors."

• An article on the use of third-party spokespersons zeroes in on the technique's double-bind: Pharma companies need physicians to speak on their behalf, because the companies' self-interests undermine virtually anything they could say. But because the third parties are working on behalf of drug companies, their own credibility erodes. One general practitioner says: "Either you are not aware that this is going on-in which case you trust everything and then people are harmed by prescribing medicine that has damaging side effects. Or, if we are aware of it, then we don't know who we can trust. This can lead to us being too sceptical so we don't prescribe good new drugs as soon as we should. Either way it's a no-win situation for patients and GPs."
• Several articles examine doctors' interactions with drug company sales reps. Some of the results are unsurprising: Doctors who interact weekly with them are more likely to prescribe new drugs-and to be isolated from colleagues and to work in poorer areas. For some of them, the authors argue, visits from drug reps "possibly fulfill a pastoral rather than an educative role." More troubling are data cited in another study that suggest that the more doctors accept gifts from drug companies, the more they (inaccurately) feel they are unaffected by them.
• Some of the most interesting analysis was devoted to medical journals. Several studies questioned whether there was something wrong with sponsored research or the articles that report on it. The situation is complex. Most sponsored research tends to be favorable to the sponsor's product, but not because such research uses inferior methodologies. Concerns include whether negative results are withheld and the comparisons experimenters choose to make in their studies. In general, the authors find problems that trouble them-and that threaten the ability of scientists to reach valid conclusions based on a review of the literature. But they have no easy solutions.