There is a fantastic post up at Slate Star Codex that I can't recommend enough (both the post and the blog as a whole). In it, the resident physician writer notes that it's unclear how new information is evaluated and adopted as standard-of-care. He gives an example of a now-poorly-supported medical theory (MS caused by poor circulation), a current case where the jury is still out, and a case where a new treatment seems to have very solid data - but is still anything but mainstream medicine. For my fellow psychiatrists, that last one would be the use of minocycline for negative symptoms of schizophrenia. (The writer says no psychiatrists he knows have heard of this but at my institution it's starting to be discussed; however I've never seen anyone started on it for negative symptoms.) The concern is really this: isn't it possible that a potentially valuable publication will languish in obscurity, never to be replicated and built into the evidence-based pantheon? My suspicion is that these things start getting to patients as soon as insurers' and institutions' formularies adopt them, and that medical education and dissemination by journals and conferences is only secondary (apart from the extent to which those things influence formularies.)
Of special importance, this article also makes a point of demolishing the slopping thinking that private sector drug development somehow "suppresses" new treatments that don't make money. That's false. What they do is un-suppress treatments that they think will make money. Before medical school I had a twelve-year career as a drug development consultant, running the studies that would un-suppress drugs, and I find it very frustrating to hear the bias in academia against an enterprise that has done so much good for so many patients. The reality is that drug companies do not suppress or distort information, but they do decide based on a profit motive what information to pursue in the first place. As with all science, each study is a move that decreases uncertainty - about efficacy and safety of each treatment. You have to decide what the marginal value of that uncertainty decrement is, based on some combination of patient suffering and money. And of course that value will be different if you're part of a for-profit company, or an academic institution. As with most things, any narrative that tries to reduce this to a more-neatly-worldview-fitting left-right political angle is at least oversimplifying to the point of incoherence, and more likely just flat out wrong. That is to say: if your claim is that big bad regulations are what make drug discovery difficult and the government is in the way of patients and profits, you're wrong, just like you're wrong if you think that big bad drug companies somehow suppress the truth.